[HEAD]: initial commit

.Rprofile

@@ -0,0 +1 @@
+source("renv/activate.R")

.Rproj.user/461AD266/pcs/files-pane.pper

@@ -0,0 +1,9 @@
+{
+    "path": "~/football_competitiveness",
+    "sortOrder": [
+        {
+            "columnIndex": 2,
+            "ascending": true
+        }
+    ]
+}

.Rproj.user/461AD266/pcs/source-pane.pper

@@ -0,0 +1,3 @@
+{
+    "activeTab": 0
+}

.Rproj.user/461AD266/pcs/windowlayoutstate.pper

@@ -0,0 +1,14 @@
+{
+    "left": {
+        "splitterpos": 312,
+        "topwindowstate": "NORMAL",
+        "panelheight": 744,
+        "windowheight": 782
+    },
+    "right": {
+        "splitterpos": 469,
+        "topwindowstate": "NORMAL",
+        "panelheight": 744,
+        "windowheight": 782
+    }
+}

.Rproj.user/461AD266/pcs/workbench-pane.pper

@@ -0,0 +1,5 @@
+{
+    "TabSet1": 0,
+    "TabSet2": 0,
+    "TabZoom": {}
+}

.Rproj.user/461AD266/sources/prop/9DEA2FB2

@@ -0,0 +1,4 @@
+{
+    "source_window_id": "",
+    "Source": "Source"
+}

.Rproj.user/461AD266/sources/prop/INDEX

@@ -0,0 +1 @@
+~%2Ffootball_competitiveness%2Fdata_manipulation.R="9DEA2FB2"

.Rproj.user/461AD266/sources/session-9bca3639/2E827BA4

@@ -0,0 +1,24 @@
+{
+    "id": "2E827BA4",
+    "path": "~/football_competitiveness/data_manipulation.R",
+    "project_path": "data_manipulation.R",
+    "type": "r_source",
+    "hash": "0",
+    "contents": "",
+    "dirty": false,
+    "created": 1785865647905.0,
+    "source_on_save": false,
+    "relative_order": 1,
+    "properties": {
+        "source_window_id": "",
+        "Source": "Source"
+    },
+    "folds": "",
+    "lastKnownWriteTime": 1785865410,
+    "encoding": "UTF-8",
+    "collab_server": "",
+    "source_window": "",
+    "last_content_update": 1785865410,
+    "read_only": false,
+    "read_only_alternatives": []
+}

.Rproj.user/461AD266/sources/session-9bca3639/2E827BA4-contents

@@ -0,0 +1,126 @@
+European_performance_data = read_csv('Data/European performance historical data.csv')
+Points_difference_data <- read_csv("Data/Points difference.csv")
+Players_used_data <- read_csv("Data/European clubs number of players used.csv")
+Top_goalscorers_data <- read_csv("Data/Average top scorers.csv")
+Market_value_data <- read_csv("Data/Market values.csv")
+gini_coeffecient_data <- Market_value_data %>%
+  select(-`Average market value of clubs`)
+all_seasons <- unique(European_performance_data$Season)
+
+overall_country_performance <- European_performance_data %>%
+  group_by(Country) %>%
+  summarise(total_performance = sum(European_performance, na.rm = TRUE)) %>%
+  arrange(desc(total_performance))
+
+#season by season performance
+country_performance_by_season <- European_performance_data %>%
+  filter(Country %in% c("Germany", "Spain", "England", "Portugal", "Italy", "France")) %>%
+  group_by(Country, Season) %>%
+  summarise(
+    European_performance = sum(European_performance, na.rm = TRUE),
+    .groups = "drop"
+  ) %>%
+  complete(Country, Season = all_seasons, fill = list(European_performance = 0))
+#season by season performance by competition
+country_performance_by_season_comp_breakdown <- European_performance_data %>%
+  filter(Country %in% c("Germany", "Spain", "England", "Portugal", "Italy", "France")) %>%
+  group_by(Country, Season) %>%
+  summarise(
+    CL = sum(European_performance[Competition == "CL"], na.rm = TRUE),
+    EL = sum(European_performance[Competition == "EL"], na.rm = TRUE),
+    .groups = "drop"
+  ) %>%
+  complete(Country, Season = all_seasons, fill = list(CL = 0, EL = 0)) %>%
+  pivot_longer(cols = c("CL", "EL"), names_to = "competition", values_to = "European_performance") %>%
+  mutate(European_performance = ifelse(competition == "CL", European_performance/2, European_performance)) #halving score for CL as had double weighting
+
+#join all metrics data together
+#impute missing Gini coefficient with average for league - assumed the same over time
+season_league_metrics_dataset <- Points_difference_data %>%
+  left_join(Players_used_data, by = c("Country", "Season")) %>%
+  left_join(Top_goalscorers_data, by = c("Country", "Season")) %>%
+  left_join(gini_coeffecient_data, by = c("Country", "Season")) %>%
+  mutate(season_start = as.integer(str_extract(Season, "^\\d{4}"))) %>%
+  group_by(Country) %>%
+  mutate(Gini_coefficient = coalesce(Gini_coefficient, mean(Gini_coefficient, na.rm = TRUE))) %>%
+  ungroup()
+
+#average country performance
+average_country_metrics <- season_league_metrics_dataset %>%
+  group_by(Country) %>%
+  summarise(across(!contains("eason"), mean, na.rm = TRUE))
+
+#scaling to min-max of variables
+average_country_metrics_scaled <- average_country_metrics
+average_country_metrics_scaled[,-1] <- lapply(average_country_metrics[,-1], function(x) {
+  (x - min(x)) / (max(x) - min(x))
+})
+
+#scaling metrics normally
+scaled_metrics_data <- season_league_metrics_dataset %>%
+  mutate(across(
+    c(First_and_second, First_and_CL, First_and_relegated, CL_and_relegated,
+      Winners_GD, Top_4_total_GD, Average_number_of_players,
+      Average_goals_by_top_3_players, Gini_coefficient),
+    ~ as.numeric(scale(.))
+  ))
+
+#join metrics data to European performance data
+#calculate European performance in next season as lead
+season_league_full_dataset <- season_league_metrics_dataset %>%
+  left_join(country_performance_by_season, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#join metrics data to European performance data by competition
+#calculate European performance in next season as lead
+season_league_full_dataset_comp_breakdown <- season_league_metrics_dataset %>%
+  left_join(country_performance_by_season_comp_breakdown, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country, competition) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#full data scaled
+scaled_full_data <- scaled_metrics_data %>%
+  left_join(country_performance_by_season, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#data by competition scaled
+scaled_full_data_comp_breakdown <- scaled_metrics_data %>%
+  left_join(country_performance_by_season_comp_breakdown, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country, competition) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#full dataset with competition breakdown and full
+pivoted_scaled_full_data_comp_breakdown <- scaled_full_data_comp_breakdown %>%
+  pivot_wider(names_from = competition, values_from = c(European_performance, lead_european_performance)) %>%
+  mutate(European_performance = 2*European_performance_CL + European_performance_EL,
+         lead_european_performance = 2*lead_european_performance_CL + lead_european_performance_EL)
+
+#basic profiling based on European performance
+quantile_splits <- scaled_full_data %>%
+  mutate(
+    performance_group = ntile(European_performance, 3)
+  )
+european_performance_groups_characteristics <- quantile_splits %>%
+  group_by(performance_group) %>%
+  summarise(across(where(is.numeric), mean, na.rm = TRUE)) %>%
+  select(-season_start) %>%
+  mutate(label = case_when(
+    performance_group == 1 ~ "Worst performing",
+    performance_group == 2 ~ "Middle performing",
+    performance_group == 3 ~ "Best performing"
+  )) %>%
+  mutate(performance_group = case_when(
+    label == "Worst performing" ~ 1,
+    label == "Middle performing" ~ 3,
+    label == "Best performing" ~ 2
+  ))

.dockerignore

@@ -0,0 +1,10 @@
+.Rproj.user
+.Rhistory
+.RData
+.git
+.gitignore
+
+renv/library
+renv/staging
+
+*.Rproj

Data/Average top scorers.csv

@@ -0,0 +1,163 @@
+Season,Average goals by top 3 players,Country
+1999–20,24.333333333333332,England
+2000–01,19.666666666666668,England
+2001–02,23.333333333333332,England
+2002–03,24.0,England
+2003–04,24.0,England
+2004–05,20.0,England
+2005–06,22.0,England
+2006–07,18.333333333333332,England
+2007–08,26.333333333333332,England
+2008–09,17.666666666666668,England
+2009–10,26.333333333333332,England
+2010–11,19.333333333333332,England
+2011–12,26.666666666666668,England
+2012–13,23.333333333333332,England
+2013–14,24.333333333333332,England
+2014–15,22.333333333333332,England
+2015–16,24.333333333333332,England
+2016–17,26.0,England
+2017–18,27.666666666666668,England
+2018–19,22.0,England
+2019–20,22.333333333333332,England
+2020–21,21.0,England
+2021–22,21.333333333333332,England
+2022–23,28.666666666666668,England
+2023–24,23.333333333333332,England
+2024–25,24.666666666666668,England
+2025–26,22.0,England
+1999–20,25.0,Spain
+2000–01,23.0,Spain
+2001–02,19.0,Spain
+2002–03,25.0,Spain
+2003–04,21.0,Spain
+2004–05,24.0,Spain
+2005–06,22.666666666666668,Spain
+2006–07,22.666666666666668,Spain
+2007–08,23.333333333333332,Spain
+2008–09,30.0,Spain
+2009–10,29.0,Spain
+2010–11,30.333333333333332,Spain
+2011–12,40.0,Spain
+2012–13,36.0,Spain
+2013–14,28.666666666666668,Spain
+2014–15,37.666666666666664,Spain
+2015–16,33.666666666666664,Spain
+2016–17,30.0,Spain
+2017–18,28.333333333333332,Spain
+2018–19,25.666666666666668,Spain
+2019–20,21.333333333333332,Spain
+2020–21,25.333333333333332,Spain
+2021–22,20.333333333333332,Spain
+2022–23,19.333333333333332,Spain
+2023–24,22.0,Spain
+2024–25,26.333333333333332,Spain
+2025–26,21.666666666666668,Spain
+1999–20,16.666666666666668,Germany
+2000–01,21.0,Germany
+2001–02,17.666666666666668,Germany
+2002–03,19.333333333333332,Germany
+2003–04,23.666666666666668,Germany
+2004–05,22.0,Germany
+2005–06,22.0,Germany
+2006–07,17.333333333333332,Germany
+2007–08,19.333333333333332,Germany
+2008–09,26.0,Germany
+2009–10,20.666666666666668,Germany
+2010–11,22.333333333333332,Germany
+2011–12,25.666666666666668,Germany
+2012–13,21.666666666666668,Germany
+2013–14,18.333333333333332,Germany
+2014–15,17.666666666666668,Germany
+2015–16,25.0,Germany
+2016–17,28.666666666666668,Germany
+2017–18,19.333333333333332,Germany
+2018–19,19.0,Germany
+2019–20,26.333333333333332,Germany
+2020–21,32.0,Germany
+2021–22,27.0,Germany
+2022–23,15.666666666666666,Germany
+2023–24,29.333333333333332,Germany
+2024–25,22.666666666666668,Germany
+2025–26,24.0,Germany
+1999–20,23.0,Italy
+2000–01,24.0,Italy
+2001–02,23.333333333333332,Italy
+2002–03,19.666666666666668,Italy
+2003–04,22.333333333333332,Italy
+2004–05,22.666666666666668,Italy
+2005–06,25.333333333333332,Italy
+2006–07,21.666666666666668,Italy
+2007–08,20.0,Italy
+2008–09,24.333333333333332,Italy
+2009–10,23.333333333333332,Italy
+2010–11,25.0,Italy
+2011–12,25.0,Italy
+2012–13,22.666666666666668,Italy
+2013–14,20.333333333333332,Italy
+2014–15,21.333333333333332,Italy
+2015–16,24.333333333333332,Italy
+2016–17,27.666666666666668,Italy
+2017–18,26.666666666666668,Italy
+2018–19,23.666666666666668,Italy
+2019–20,30.0,Italy
+2020–21,25.0,Italy
+2021–22,24.0,Italy
+2022–23,21.0,Italy
+2023–24,18.333333333333332,Italy
+2024–25,19.666666666666668,Italy
+2025–26,15.0,Italy
+1999–20,22.0,France
+2000–01,19.333333333333332,France
+2001–02,21.0,France
+2002–03,22.0,France
+2003–04,21.666666666666668,France
+2004–05,16.333333333333332,France
+2005–06,16.333333333333332,France
+2006–07,13.333333333333334,France
+2007–08,18.0,France
+2008–09,19.333333333333332,France
+2009–10,16.666666666666668,France
+2010–11,21.333333333333332,France
+2011–12,20.666666666666668,France
+2012–13,22.666666666666668,France
+2013–14,19.333333333333332,France
+2014–15,22.333333333333332,France
+2015–16,26.0,France
+2016–17,28.0,France
+2017–18,23.0,France
+2018–19,24.333333333333332,France
+2019–20,17.333333333333332,France
+2020–21,22.333333333333332,France
+2021–22,24.666666666666668,France
+2022–23,26.666666666666668,France
+2023–24,21.666666666666668,France
+2024–25,19.666666666666668,France
+2025–26,17.666666666666668,France
+1999–20,27.0,Portugal
+2000–01,19.666666666666668,Portugal
+2001–02,27.333333333333332,Portugal
+2002–03,17.333333333333332,Portugal
+2003–04,17.666666666666668,Portugal
+2004–05,18.0,Portugal
+2005–06,15.666666666666666,Portugal
+2006–07,12.666666666666666,Portugal
+2007–08,16.333333333333332,Portugal
+2008–09,18.0,Portugal
+2009–10,21.333333333333332,Portugal
+2010–11,18.333333333333332,Portugal
+2011–12,18.666666666666668,Portugal
+2012–13,21.0,Portugal
+2013–14,17.0,Portugal
+2014–15,20.0,Portugal
+2015–16,26.333333333333332,Portugal
+2016–17,23.0,Portugal
+2017–18,27.666666666666668,Portugal
+2018–19,20.0,Portugal
+2019–20,18.0,Portugal
+2020–21,20.333333333333332,Portugal
+2021–22,21.666666666666668,Portugal
+2022–23,19.333333333333332,Portugal
+2023–24,23.333333333333332,Portugal
+2024–25,25.666666666666668,Portugal
+2025–26,23.333333333333332,Portugal

Data/European clubs number of players used.csv

@@ -0,0 +1,163 @@
+Season,Average_number_of_players,Country
+1999–20,27.71428571,England
+2000–01,27.16666667,England
+2001–02,24.85714286,England
+2002–03,26.25,England
+2003–04,25.25,England
+2004–05,27.33333333,England
+2005–06,27.28571429,England
+2006–07,27.375,England
+2007–08,26.25,England
+2008–09,27.88888889,England
+2009–10,28.57142857,England
+2010–11,28.14285714,England
+2011–12,28,England
+2012–13,27.14285714,England
+2013–14,27.33333333,England
+2014–15,26.71428571,England
+2015–16,27.875,England
+2016–17,27,England
+2017–18,27.14285714,England
+2018–19,25,England
+2019–20,26.14285714,England
+2020–21,26.85714286,England
+2021–22,26.57142857,England
+2022–23,27.14285714,England
+2023–24,28.75,England
+2024–25,28.28571429,England
+2025–26,28.22222222,England
+1999–20,26,Spain
+2000–01,24,Spain
+2001–02,26.8,Spain
+2002–03,28.2,Spain
+2003–04,25.5,Spain
+2004–05,26.85714286,Spain
+2005–06,28.75,Spain
+2006–07,25.4,Spain
+2007–08,27.28571429,Spain
+2008–09,28.5,Spain
+2009–10,26.83333333,Spain
+2010–11,28.33333333,Spain
+2011–12,26.33333333,Spain
+2012–13,26.5,Spain
+2013–14,27.8,Spain
+2014–15,25.66666667,Spain
+2015–16,27.16666667,Spain
+2016–17,26,Spain
+2017–18,27.5,Spain
+2018–19,27,Spain
+2019–20,27.6,Spain
+2020–21,29,Spain
+2021–22,32.6,Spain
+2022–23,30.4,Spain
+2023–24,30.8,Spain
+2024–25,30,Spain
+2025–26,30.83333333,Spain
+1999–20,26,Germany
+2000–01,28,Germany
+2001–02,24.5,Germany
+2002–03,25.5,Germany
+2003–04,24.33333333,Germany
+2004–05,24.66666667,Germany
+2005–06,24.5,Germany
+2006–07,28,Germany
+2007–08,26.33333333,Germany
+2008–09,26.16666667,Germany
+2009–10,27,Germany
+2010–11,27,Germany
+2011–12,24.66666667,Germany
+2012–13,24.6,Germany
+2013–14,28,Germany
+2014–15,24.5,Germany
+2015–16,27.2,Germany
+2016–17,26.5,Germany
+2017–18,27.6,Germany
+2018–19,27,Germany
+2019–20,26.8,Germany
+2020–21,25,Germany
+2021–22,29.75,Germany
+2022–23,27.4,Germany
+2023–24,29.25,Germany
+2024–25,28,Germany
+2025–26,27,Germany
+1999–20,24.33333333,Italy
+2000–01,25,Italy
+2001–02,24,Italy
+2002–03,29.75,Italy
+2003–04,24.33333333,Italy
+2004–05,27.66666667,Italy
+2005–06,25,Italy
+2006–07,28,Italy
+2007–08,24.75,Italy
+2008–09,29,Italy
+2009–10,29,Italy
+2010–11,30.5,Italy
+2011–12,30.66666667,Italy
+2012–13,28.5,Italy
+2013–14,29,Italy
+2014–15,30.33333333,Italy
+2015–16,30,Italy
+2016–17,24.5,Italy
+2017–18,24.66666667,Italy
+2018–19,26.66666667,Italy
+2019–20,29.33333333,Italy
+2020–21,30,Italy
+2021–22,29.66666667,Italy
+2022–23,29.66666667,Italy
+2023–24,32.5,Italy
+2024–25,30.66666667,Italy
+2025–26,28,Italy
+1999–20,26.33333333,France
+2000–01,26.8,France
+2001–02,27.5,France
+2002–03,25.8,France
+2003–04,27.57142857,France
+2004–05,25.2,France
+2005–06,27.33333333,France
+2006–07,25.14285714,France
+2007–08,27.28571429,France
+2008–09,25.71428571,France
+2009–10,26.25,France
+2010–11,25,France
+2011–12,26.8,France
+2012–13,28.25,France
+2013–14,27,France
+2014–15,25,France
+2015–16,30.5,France
+2016–17,29,France
+2017–18,29,France
+2018–19,30.33333333,France
+2019–20,28.75,France
+2020–21,31.2,France
+2021–22,29.6,France
+2022–23,29.66666667,France
+2023–24,28.8,France
+2024–25,30.2,France
+2025–26,33.2,France
+1999–20,29,Portugal
+2000–01,26,Portugal
+2001–02,29,Portugal
+2002–03,29.5,Portugal
+2003–04,32,Portugal
+2004–05,32,Portugal
+2005–06,29,Portugal
+2006–07,31,Portugal
+2007–08,28,Portugal
+2008–09,27,Portugal
+2009–10,28,Portugal
+2010–11,26,Portugal
+2011–12,30,Portugal
+2012–13,26.5,Portugal
+2013–14,31,Portugal
+2014–15,26.5,Portugal
+2015–16,29.5,Portugal
+2016–17,28.33333333,Portugal
+2017–18,30,Portugal
+2018–19,30,Portugal
+2019–20,28,Portugal
+2020–21,30.5,Portugal
+2021–22,31.66666667,Portugal
+2022–23,30,Portugal
+2023–24,31.5,Portugal
+2024–25,32,Portugal
+2025–26,33.5,Portugal

Data/European performance historical data.csv

@@ -0,0 +1,433 @@
+Season,Team,European_performance,Country,Competition
+2025–26,Paris Saint-Germain,20,France,CL
+2025–26,Arsenal,16,England,CL
+2025–26,Atlético Madrid,12,Spain,CL
+2025–26,Bayern Munich,12,Germany,CL
+2025–26,Aston Villa,10,England,EL
+2025–26,Freiburg,8,Germany,EL
+2025–26,Barcelona,8,Spain,CL
+2025–26,Liverpool,8,England,CL
+2025–26,Real Madrid,8,Spain,CL
+2025–26,Sporting CP,8,Portugal,CL
+2025–26,Nottingham Forest,6,England,EL
+2025–26,Braga,6,Portugal,EL
+2025–26,Celta Vigo,4,Spain,EL
+2025–26,Real Betis,4,Spain,EL
+2025–26,Bologna,4,Italy,EL
+2025–26,Porto,4,Portugal,EL
+2024–25,Paris Saint-Germain,20,France,CL
+2024–25,Internazionale,16,Italy,CL
+2024–25,Arsenal,12,England,CL
+2024–25,Barcelona,12,Spain,CL
+2024–25,Tottenham Hotspur,10,England,EL
+2024–25,Manchester United,8,England,EL
+2024–25,Aston Villa,8,England,CL
+2024–25,Real Madrid,8,Spain,CL
+2024–25,Borussia Dortmund,8,Germany,CL
+2024–25,Bayern Munich,8,Germany,CL
+2024–25,Bodo/Glimt,6,Norway,EL
+2024–25,Athletic Bilbao,6,Spain,EL
+2024–25,Lazio,4,Italy,EL
+2024–25,Lyon,4,France,EL
+2024–25,Rangers,4,Scotland,EL
+2024–25,Eintracht Frankfurt,4,Germany,EL
+2023–24,Real Madrid,20,Spain,CL
+2023–24,Borussia Dortmund,16,Germany,CL
+2023–24,Paris Saint-Germain,12,France,CL
+2023–24,Bayern Munich,12,Germany,CL
+2023–24,Atalanta,10,Italy,EL
+2023–24,Arsenal,8,England,CL
+2023–24,Atlético Madrid,8,Spain,CL
+2023–24,Manchester City,8,England,CL
+2023–24,Barcelona,8,Spain,CL
+2023–24,Bayer Leverkusen,8,Germany,EL
+2023–24,Marseille,6,France,EL
+2023–24,Roma,6,Italy,EL
+2023–24,Milan,4,Italy,EL
+2023–24,Liverpool,4,England,EL
+2023–24,West Ham United,4,England,EL
+2023–24,Benfica,4,Portugal,EL
+2022–23,Manchester City,20,England,CL
+2022–23,Internazionale,16,Italy,CL
+2022–23,Milan,12,Italy,CL
+2022–23,Real Madrid,12,Spain,CL
+2022–23,Sevilla,10,Spain,EL
+2022–23,Chelsea,8,England,CL
+2022–23,Benfica,8,Portugal,CL
+2022–23,Bayern Munich,8,Germany,CL
+2022–23,Napoli,8,Italy,CL
+2022–23,Roma,8,Italy,EL
+2022–23,Juventus,6,Italy,EL
+2022–23,Bayer Leverkusen,6,Germany,EL
+2022–23,Manchester United,4,England,EL
+2022–23,Sporting CP,4,Portugal,EL
+2022–23,Union Saint-Gilloise,4,Belgium,EL
+2022–23,Feyenoord,4,Netherlands,EL
+2021–22,Real Madrid,20,Spain,CL
+2021–22,Liverpool,16,England,CL
+2021–22,Manchester City,12,England,CL
+2021–22,Villarreal,12,Spain,CL
+2021–22,Eintracht Frankfurt,10,Germany,EL
+2021–22,Chelsea,8,England,CL
+2021–22,Atlético Madrid,8,Spain,CL
+2021–22,Bayern Munich,8,Germany,CL
+2021–22,Benfica,8,Portugal,CL
+2021–22,Rangers,8,Scotland,EL
+2021–22,RB Leipzig,6,Germany,EL
+2021–22,West Ham United,6,England,EL
+2021–22,Atalanta,4,Italy,EL
+2021–22,Barcelona,4,Spain,EL
+2021–22,Lyon,4,France,EL
+2021–22,Braga,4,Portugal,EL
+2020–21,Chelsea,20,England,CL
+2020–21,Manchester City,16,England,CL
+2020–21,Paris Saint-Germain,12,France,CL
+2020–21,Real Madrid,12,Spain,CL
+2020–21,Villarreal,10,Spain,EL
+2020–21,Borussia Dortmund,8,Germany,CL
+2020–21,Porto,8,Portugal,CL
+2020–21,Bayern Munich,8,Germany,CL
+2020–21,Liverpool,8,England,CL
+2020–21,Manchester United,8,England,EL
+2020–21,Roma,6,Italy,EL
+2020–21,Arsenal,6,England,EL
+2020–21,Granada,4,Spain,EL
+2020–21,Slavia Prague,4,Czech Republic,EL
+2020–21,Ajax,4,Netherlands,EL
+2020–21,Dinamo Zagreb,4,Croatia,EL
+2019–20,Bayern Munich,20,Germany,CL
+2019–20,Paris Saint-Germain,16,France,CL
+2019–20,Lyon,12,France,CL
+2019–20,RB Leipzig,12,Germany,CL
+2019–20,Sevilla,10,Spain,EL
+2019–20,Manchester City,8,England,CL
+2019–20,Atlético Madrid,8,Spain,CL
+2019–20,Barcelona,8,Spain,CL
+2019–20,Atalanta,8,Italy,CL
+2019–20,Internazionale,8,Italy,EL
+2019–20,Manchester United,6,England,EL
+2019–20,Shakhtar Donetsk,6,Ukraine,EL
+2019–20,Basel,4,Switzerland,EL
+2019–20,Copenhagen,4,Denmark,EL
+2019–20,Bayer Leverkusen,4,Germany,EL
+2019–20,Wolverhampton Wanderers,4,England,EL
+2018–19,Liverpool,20,England,CL
+2018–19,Tottenham Hotspur,16,England,CL
+2018–19,Ajax,12,Netherlands,CL
+2018–19,Barcelona,12,Spain,CL
+2018–19,Chelsea,10,England,EL
+2018–19,Juventus,8,Italy,CL
+2018–19,Porto,8,Portugal,CL
+2018–19,Manchester City,8,England,CL
+2018–19,Manchester United,8,England,CL
+2018–19,Arsenal,8,England,EL
+2018–19,Valencia,6,Spain,EL
+2018–19,Eintracht Frankfurt,6,Germany,EL
+2018–19,Napoli,4,Italy,EL
+2018–19,Villarreal,4,Spain,EL
+2018–19,Benfica,4,Portugal,EL
+2018–19,Slavia Prague,4,Czech Republic,EL
+2017–18,Real Madrid,20,Spain,CL
+2017–18,Liverpool,16,England,CL
+2017–18,Bayern Munich,12,Germany,CL
+2017–18,Roma,12,Italy,CL
+2017–18,Atlético Madrid,10,Spain,EL
+2017–18,Barcelona,8,Spain,CL
+2017–18,Sevilla,8,Spain,CL
+2017–18,Juventus,8,Italy,CL
+2017–18,Manchester City,8,England,CL
+2017–18,Marseille,8,France,EL
+2017–18,Red Bull Salzburg,6,Austria,EL
+2017–18,Arsenal,6,England,EL
+2017–18,RB Leipzig,4,Germany,EL
+2017–18,CSKA Moscow,4,Russia,EL
+2017–18,Sporting CP,4,Portugal,EL
+2017–18,Lazio,4,Italy,EL
+2016–17,Real Madrid,20,Spain,CL
+2016–17,Juventus,16,Italy,CL
+2016–17,Atlético Madrid,12,Spain,CL
+2016–17,Monaco,12,France,CL
+2016–17,Manchester United,10,England,EL
+2016–17,Leicester City,8,England,CL
+2016–17,Borussia Dortmund,8,Germany,CL
+2016–17,Bayern Munich,8,Germany,CL
+2016–17,Barcelona,8,Spain,CL
+2016–17,Ajax,8,Netherlands,EL
+2016–17,Lyon,6,France,EL
+2016–17,Celta Vigo,6,Spain,EL
+2016–17,Anderlecht,4,Belgium,EL
+2016–17,Genk,4,Belgium,EL
+2016–17,Schalke 04,4,Germany,EL
+2016–17,Beşiktaş,4,Turkey,EL
+2015–16,Real Madrid,20,Spain,CL
+2015–16,Atlético Madrid,16,Spain,CL
+2015–16,Manchester City,12,England,CL
+2015–16,Bayern Munich,12,Germany,CL
+2015–16,Sevilla,10,Spain,EL
+2015–16,VfL Wolfsburg,8,Germany,CL
+2015–16,Benfica,8,Portugal,CL
+2015–16,Barcelona,8,Spain,CL
+2015–16,Paris Saint-Germain,8,France,CL
+2015–16,Liverpool,8,England,EL
+2015–16,Shakhtar Donetsk,6,Ukraine,EL
+2015–16,Villarreal,6,Spain,EL
+2015–16,Braga,4,Portugal,EL
+2015–16,Sparta Prague,4,Czech Republic,EL
+2015–16,Athletic Bilbao,4,Spain,EL
+2015–16,Borussia Dortmund,4,Germany,EL
+2014–15,Barcelona,20,Spain,CL
+2014–15,Juventus,16,Italy,CL
+2014–15,Bayern Munich,12,Germany,CL
+2014–15,Real Madrid,12,Spain,CL
+2014–15,Sevilla,10,Spain,EL
+2014–15,Paris Saint-Germain,8,France,CL
+2014–15,Atlético Madrid,8,Spain,CL
+2014–15,Porto,8,Portugal,CL
+2014–15,Monaco,8,France,CL
+2014–15,Dnipro Dnipropetrovsk,8,Ukraine,EL
+2014–15,Napoli,6,Italy,EL
+2014–15,Fiorentina,6,Italy,EL
+2014–15,Zenit Saint Petersburg,4,Russia,EL
+2014–15,Club Brugge,4,Belgium,EL
+2014–15,Dynamo Kyiv,4,Ukraine,EL
+2014–15,VfL Wolfsburg,4,Germany,EL
+2013–14,Real Madrid,20,Spain,CL
+2013–14,Atlético Madrid,16,Spain,CL
+2013–14,Bayern Munich,12,Germany,CL
+2013–14,Chelsea,12,England,CL
+2013–14,Sevilla,10,Spain,EL
+2013–14,Barcelona,8,Spain,CL
+2013–14,Borussia Dortmund,8,Germany,CL
+2013–14,Paris Saint-Germain,8,France,CL
+2013–14,Manchester United,8,England,CL
+2013–14,Benfica,8,Portugal,EL
+2013–14,Valencia,6,Spain,EL
+2013–14,Juventus,6,Italy,EL
+2013–14,AZ,4,Netherlands,EL
+2013–14,Lyon,4,France,EL
+2013–14,Basel,4,Switzerland,EL
+2013–14,Porto,4,Portugal,EL
+2012–13,Bayern Munich,20,Germany,CL
+2012–13,Borussia Dortmund,16,Germany,CL
+2012–13,Barcelona,12,Spain,CL
+2012–13,Real Madrid,12,Spain,CL
+2012–13,Chelsea,10,England,EL
+2012–13,Málaga,8,Spain,CL
+2012–13,Galatasaray,8,Turkey,CL
+2012–13,Paris Saint-Germain,8,France,CL
+2012–13,Juventus,8,Italy,CL
+2012–13,Benfica,8,Portugal,EL
+2012–13,Fenerbahçe,6,Turkey,EL
+2012–13,Basel,6,Switzerland,EL
+2012–13,Rubin Kazan,4,Russia,EL
+2012–13,Tottenham Hotspur,4,England,EL
+2012–13,Lazio,4,Italy,EL
+2012–13,Newcastle United,4,England,EL
+2011–12,Chelsea,20,England,CL
+2011–12,Bayern Munich,16,Germany,CL
+2011–12,Real Madrid,12,Spain,CL
+2011–12,Barcelona,12,Spain,CL
+2011–12,Atlético Madrid,10,Spain,EL
+2011–12,APOEL,8,Cyprus,CL
+2011–12,Marseille,8,France,CL
+2011–12,Benfica,8,Portugal,CL
+2011–12,Milan,8,Italy,CL
+2011–12,Athletic Bilbao,8,Spain,EL
+2011–12,Valencia,6,Spain,EL
+2011–12,Sporting CP,6,Portugal,EL
+2011–12,AZ,4,Netherlands,EL
+2011–12,Schalke 04,4,Germany,EL
+2011–12,Metalist Kharkiv,4,Ukraine,EL
+2011–12,Hannover 96,4,Germany,EL
+2010–11,Barcelona,20,Spain,CL
+2010–11,Manchester United,16,England,CL
+2010–11,Schalke 04,12,Germany,CL
+2010–11,Real Madrid,12,Spain,CL
+2010–11,Porto,10,Portugal,EL
+2010–11,Tottenham Hotspur,8,England,CL
+2010–11,Chelsea,8,England,CL
+2010–11,Shakhtar Donetsk,8,Ukraine,CL
+2010–11,Internazionale,8,Italy,CL
+2010–11,Braga,8,Portugal,EL
+2010–11,Benfica,6,Portugal,EL
+2010–11,Villarreal,6,Spain,EL
+2010–11,Spartak Moscow,4,Russia,EL
+2010–11,PSV Eindhoven,4,Netherlands,EL
+2010–11,Twente,4,Netherlands,EL
+2010–11,Dynamo Kyiv,4,Ukraine,EL
+2009–10,Internazionale,20,Italy,CL
+2009–10,Bayern Munich,16,Germany,CL
+2009–10,Lyon,12,France,CL
+2009–10,Barcelona,12,Spain,CL
+2009–10,Atlético Madrid,10,Spain,EL
+2009–10,Bordeaux,8,France,CL
+2009–10,Manchester United,8,England,CL
+2009–10,Arsenal,8,England,CL
+2009–10,CSKA Moscow,8,Russia,CL
+2009–10,Fulham,8,England,EL
+2009–10,Hamburg,6,Germany,EL
+2009–10,Liverpool,6,England,EL
+2009–10,VfL Wolfsburg,4,Germany,EL
+2009–10,Standard Liège,4,Belgium,EL
+2009–10,Valencia,4,Spain,EL
+2009–10,Benfica,4,Portugal,EL
+2008–09,Barcelona,20,Spain,CL
+2008–09,Manchester United,16,England,CL
+2008–09,Arsenal,12,England,CL
+2008–09,Chelsea,12,England,CL
+2008–09,Shakhtar Donetsk,10,Ukraine,EL
+2008–09,Villarreal,8,Spain,CL
+2008–09,Porto,8,Portugal,CL
+2008–09,Liverpool,8,England,CL
+2008–09,Bayern Munich,8,Germany,CL
+2008–09,Werder Bremen,8,Germany,EL
+2008–09,Hamburg,6,Germany,EL
+2008–09,Dynamo Kyiv,6,Ukraine,EL
+2008–09,Manchester City,4,England,EL
+2008–09,Paris Saint-Germain,4,France,EL
+2008–09,Marseille,4,France,EL
+2008–09,Udinese,4,Italy,EL
+2007–08,Manchester United,20,England,CL
+2007–08,Chelsea,16,England,CL
+2007–08,Liverpool,12,England,CL
+2007–08,Barcelona,12,Spain,CL
+2007–08,Zenit Saint Petersburg,10,Russia,EL
+2007–08,Arsenal,8,England,CL
+2007–08,Roma,8,Italy,CL
+2007–08,Schalke 04,8,Germany,CL
+2007–08,Fenerbahçe,8,Turkey,CL
+2007–08,Rangers,8,Scotland,EL
+2007–08,Bayern Munich,6,Germany,EL
+2007–08,Fiorentina,6,Italy,EL
+2007–08,Bayer Leverkusen,4,Germany,EL
+2007–08,Sporting CP,4,Portugal,EL
+2007–08,Getafe,4,Spain,EL
+2007–08,PSV Eindhoven,4,Netherlands,EL
+2006–07,Milan,20,Italy,CL
+2006–07,Liverpool,16,England,CL
+2006–07,Chelsea,12,England,CL
+2006–07,Manchester United,12,England,CL
+2006–07,Sevilla,10,Spain,EL
+2006–07,Bayern Munich,8,Germany,CL
+2006–07,PSV Eindhoven,8,Netherlands,CL
+2006–07,Roma,8,Italy,CL
+2006–07,Valencia,8,Spain,CL
+2006–07,Espanyol,8,Spain,EL
+2006–07,Werder Bremen,6,Germany,EL
+2006–07,Osasuna,6,Spain,EL
+2006–07,AZ,4,Netherlands,EL
+2006–07,Bayer Leverkusen,4,Germany,EL
+2006–07,Tottenham Hotspur,4,England,EL
+2006–07,Benfica,4,Portugal,EL
+2005–06,Barcelona,20,Spain,CL
+2005–06,Arsenal,16,England,CL
+2005–06,Villarreal,12,Spain,CL
+2005–06,Milan,12,Italy,CL
+2005–06,Sevilla,10,Spain,EL
+2005–06,Juventus,8,Italy,CL
+2005–06,Lyon,8,France,CL
+2005–06,Internazionale,8,Italy,CL
+2005–06,Benfica,8,Portugal,CL
+2005–06,Middlesbrough,8,England,EL
+2005–06,Schalke 04,6,Germany,EL
+2005–06,Steaua București,6,Romania,EL
+2005–06,Zenit Saint Petersburg,4,Russia,EL
+2005–06,Basel,4,Switzerland,EL
+2005–06,Rapid București,4,Romania,EL
+2005–06,Levski Sofia,4,Bulgaria,EL
+2004–05,Liverpool,20,England,CL
+2004–05,Milan,16,Italy,CL
+2004–05,Chelsea,12,England,CL
+2004–05,PSV Eindhoven,12,Netherlands,CL
+2004–05,CSKA Moscow,10,Russia,EL
+2004–05,Juventus,8,Italy,CL
+2004–05,Lyon,8,France,CL
+2004–05,Bayern Munich,8,Germany,CL
+2004–05,Internazionale,8,Italy,CL
+2004–05,Sporting CP,8,Portugal,EL
+2004–05,AZ,6,Netherlands,EL
+2004–05,Parma,6,Italy,EL
+2004–05,Newcastle United,4,England,EL
+2004–05,Villarreal,4,Spain,EL
+2004–05,Austria Wien,4,Austria,EL
+2004–05,Auxerre,4,France,EL
+2003–04,Porto,20,Portugal,CL
+2003–04,Monaco,16,France,CL
+2003–04,Chelsea,12,England,CL
+2003–04,Deportivo La Coruña,12,Spain,CL
+2003–04,Valencia,10,Spain,EL
+2003–04,Arsenal,8,England,CL
+2003–04,Milan,8,Italy,CL
+2003–04,Lyon,8,France,CL
+2003–04,Real Madrid,8,Spain,CL
+2003–04,Marseille,8,France,EL
+2003–04,Newcastle United,6,England,EL
+2003–04,Villarreal,6,Spain,EL
+2003–04,Bordeaux,4,France,EL
+2003–04,Internazionale,4,Italy,EL
+2003–04,Celtic,4,Scotland,EL
+2003–04,PSV Eindhoven,4,Netherlands,EL
+2002–03,Milan,20,Italy,CL
+2002–03,Juventus,16,Italy,CL
+2002–03,Real Madrid,12,Spain,CL
+2002–03,Internazionale,12,Italy,CL
+2002–03,Porto,10,Portugal,EL
+2002–03,Manchester United,8,England,CL
+2002–03,Ajax,8,Netherlands,CL
+2002–03,Valencia,8,Spain,CL
+2002–03,Barcelona,8,Spain,CL
+2002–03,Celtic,8,Scotland,EL
+2002–03,Lazio,6,Italy,EL
+2002–03,Boavista,6,Portugal,EL
+2002–03,Panathinaikos,4,Greece,EL
+2002–03,Beşiktaş,4,Turkey,EL
+2002–03,Liverpool,4,England,EL
+2002–03,Málaga,4,Spain,EL
+2001–02,Real Madrid,20,Spain,CL
+2001–02,Bayer Leverkusen,16,Germany,CL
+2001–02,Barcelona,12,Spain,CL
+2001–02,Manchester United,12,England,CL
+2001–02,Feyenoord,10,Netherlands,EL
+2001–02,Panathinaikos,8,Greece,CL
+2001–02,Bayern Munich,8,Germany,CL
+2001–02,Deportivo La Coruña,8,Spain,CL
+2001–02,Liverpool,8,England,CL
+2001–02,Borussia Dortmund,8,Germany,EL
+2001–02,Internazionale,6,Italy,EL
+2001–02,Milan,6,Italy,EL
+2001–02,Valencia,4,Spain,EL
+2001–02,PSV Eindhoven,4,Netherlands,EL
+2001–02,Slovan Liberec,4,Czech Republic,EL
+2001–02,Hapoel Tel Aviv,4,Israel,EL
+2000–01,Bayern Munich,20,Germany,CL
+2000–01,Valencia,16,Spain,CL
+2000–01,Leeds United,12,England,CL
+2000–01,Real Madrid,12,Spain,CL
+2000–01,Liverpool,10,England,EL
+2000–01,Deportivo La Coruña,8,Spain,CL
+2000–01,Arsenal,8,England,CL
+2000–01,Galatasaray,8,Turkey,CL
+2000–01,Manchester United,8,England,CL
+2000–01,Deportivo Alavés,8,Spain,EL
+2000–01,Kaiserslautern,6,Germany,EL
+2000–01,Barcelona,6,Spain,EL
+2000–01,Celta Vigo,4,Spain,EL
+2000–01,Porto,4,Portugal,EL
+2000–01,Rayo Vallecano,4,Spain,EL
+2000–01,PSV Eindhoven,4,Netherlands,EL
+1999–20,Real Madrid,20,Spain,CL
+1999–20,Valencia,16,Spain,CL
+1999–20,Barcelona,12,Spain,CL
+1999–20,Bayern Munich,12,Germany,CL
+1999–20,Galatasaray,10,Turkey,EL
+1999–20,Manchester United,8,England,CL
+1999–20,Porto,8,Portugal,CL
+1999–20,Chelsea,8,England,CL
+1999–20,Lazio,8,Italy,CL
+1999–20,Arsenal,8,England,EL
+1999–20,Leeds United,6,England,EL
+1999–20,Lens,6,France,EL
+1999–20,Slavia Prague,4,Czech Republic,EL
+1999–20,Werder Bremen,4,Germany,EL
+1999–20,Mallorca,4,Spain,EL
+1999–20,Celta Vigo,4,Spain,EL

Data/Market values.csv

@@ -0,0 +1,132 @@
+Season,Gini coefficient,Average market value of clubs,Country
+2004–05,0.360018498852386,116763999.95,England
+2005–06,0.3442440170940171,117000000.0,England
+2006–07,0.37966666666666665,125249999.9,England
+2007–08,0.3519417957722006,151071500.0,England
+2008–09,0.37122676914762665,166535000.0,England
+2009–10,0.36977061182601817,174420499.95,England
+2010–11,0.3476851851851852,188676000.0,England
+2011–12,0.385829420048362,178239000.0,England
+2012–13,0.3682488958864264,187367500.0,England
+2013–14,0.37329755805691633,200557499.95,England
+2014–15,0.3786803024727264,210597500.0,England
+2015–16,0.3183357245632489,242901500.0,England
+2016–17,0.3324523769830874,293923000.0,England
+2017–18,0.39222764569225443,404620000.0,England
+2018–19,0.380701799118799,490240000.0,England
+2019–20,0.35927350571500166,415048000.0,England
+2020–21,0.3383646115529144,457720000.0,England
+2021–22,0.3088707777325594,466869999.95,England
+2022–23,0.2379889748754373,565979999.95,England
+2023–24,0.2943039490285971,605516000.0,England
+2024–25,0.28445084329176856,591165500.0,England
+2025–26,0.24882569255851744,650468500.0,England
+2004–05,0.48341967819470966,83000500.0,Spain
+2005–06,0.4349132168859722,90512999.95,Spain
+2006–07,0.41443248867167715,109570500.0,Spain
+2007–08,0.3898532139052204,126715000.0,Spain
+2008–09,0.40867937105438573,127515000.0,Spain
+2009–10,0.4496773655226122,134827499.95,Spain
+2010–11,0.45834173618113483,136362500.0,Spain
+2011–12,0.4736133893325162,138692500.0,Spain
+2012–13,0.49595993904566066,136167500.0,Spain
+2013–14,0.5224207917235434,139189999.95,Spain
+2014–15,0.5430249292418888,156342499.95,Spain
+2015–16,0.5248095677765595,173290000.0,Spain
+2016–17,0.499748389569604,202197499.95,Spain
+2017–18,0.5347268401602979,256397499.95,Spain
+2018–19,0.48223915117990523,309134999.95,Spain
+2019–20,0.4799536796864471,270831500.0,Spain
+2020–21,0.4776222471071295,261202499.95,Spain
+2021–22,0.44673797312770364,261570499.95,Spain
+2022–23,0.43895120412800465,263177999.9,Spain
+2023–24,0.45606337058230506,281708000.0,Spain
+2024–25,0.5408563121624481,271717499.9,Spain
+2025–26,0.5066320158848527,278252500.0,Spain
+2004–05,0.32341981351909727,60265000.0,Germany
+2005–06,0.2789416254489784,67093333.333333336,Germany
+2006–07,0.3090923172242875,71733333.33333333,Germany
+2007–08,0.33890211084667193,79366666.66666667,Germany
+2008–09,0.3118161071448985,88556666.66666667,Germany
+2009–10,0.30818810283251885,98640555.55555555,Germany
+2010–11,0.3106611606332398,107446666.6111111,Germany
+2011–12,0.3382094918058693,106248888.8888889,Germany
+2012–13,0.412129874895549,114723888.8888889,Germany
+2013–14,0.40688376911217505,132167222.22222222,Germany
+2014–15,0.4228109848839647,138394444.44444445,Germany
+2015–16,0.44632989517068217,148783888.8888889,Germany
+2016–17,0.40110039109009477,160251666.55555555,Germany
+2017–18,0.4084188563742467,221476666.6111111,Germany
+2018–19,0.37264934156985474,271240000.0,Germany
+2019–20,0.42824720935129273,248310555.55555555,Germany
+2020–21,0.3991102398876678,266405555.5,Germany
+2021–22,0.4215972188716651,241806111.05555555,Germany
+2022–23,0.41630635649677783,258607777.7222222,Germany
+2023–24,0.4428287408678364,271983333.3333333,Germany
+2024–25,0.4436581724640299,260410000.0,Germany
+2025–26,0.3733180378765145,292510000.0,Germany
+2004–05,0.44144154142093134,94614000.0,Italy
+2005–06,0.43461945910016203,96284000.0,Italy
+2006–07,0.38563841667056215,96265500.0,Italy
+2007–08,0.3832586436229568,112019000.0,Italy
+2008–09,0.40188776845700325,120340500.0,Italy
+2009–10,0.36207897021072266,135367500.0,Italy
+2010–11,0.351433925495717,130341500.0,Italy
+2011–12,0.3761044534344282,117908999.95,Italy
+2012–13,0.3293413780469743,128283000.0,Italy
+2013–14,0.33876816377095625,147718500.0,Italy
+2014–15,0.34808548211106644,139303999.95,Italy
+2015–16,0.39253230390872257,148821000.0,Italy
+2016–17,0.4088623997373235,170552000.0,Italy
+2017–18,0.4264639426373497,229696500.0,Italy
+2018–19,0.4014885661837348,296073500.0,Italy
+2019–20,0.379941242052611,248902500.0,Italy
+2020–21,0.3849354176959003,272249499.95,Italy
+2021–22,0.3598889347810097,265114500.0,Italy
+2022–23,0.35907910742611354,256897500.0,Italy
+2023–24,0.35090251172285253,270412000.0,Italy
+2024–25,0.3784043169013681,261715500.0,Italy
+2025–26,0.33513392632615085,278362000.0,Italy
+2004–05,0.2766737410765966,51830000.0,France
+2005–06,0.2705703151825855,62614500.0,France
+2006–07,0.32013975578609044,69693000.0,France
+2007–08,0.24847617780143733,75140000.0,France
+2008–09,0.2891482124764799,69090000.0,France
+2009–10,0.27391808830908176,79720000.0,France
+2010–11,0.2889899318977778,79806500.0,France
+2011–12,0.3258695447261759,79567500.0,France
+2012–13,0.3787823567890851,80257500.0,France
+2013–14,0.4248340493127178,86245500.0,France
+2014–15,0.4098437690711583,81930000.0,France
+2015–16,0.46182494823437237,89345000.0,France
+2016–17,0.45867096220732356,107243000.0,France
+2017–18,0.4986615772275405,173823999.95,France
+2018–19,0.44668007798124076,201074000.0,France
+2019–20,0.46338912869048804,170743000.0,France
+2020–21,0.40368277702674105,200833499.95,France
+2021–22,0.40885929715133307,210344000.0,France
+2022–23,0.4320352161412733,215299000.0,France
+2023–24,0.3983899007790753,242761666.66666666,France
+2024–25,0.4628263378361487,212350000.0,France
+2025–26,0.44695848532221644,266527777.7777778,France
+2004–05,0.5638919437159857,32836111.0,Portugal
+2006–07,0.5314655998466845,32612500.0,Portugal
+2007–08,0.44544708061078303,35773749.9375,Portugal
+2008–09,0.4063953488372093,40312500.0,Portugal
+2009–10,0.46722537891257687,47463125.0,Portugal
+2010–11,0.5165052816901409,53249999.9375,Portugal
+2011–12,0.5210959922099221,60975000.0,Portugal
+2012–13,0.5345414159258595,66158125.0,Portugal
+2013–14,0.4982536812529268,60059374.9375,Portugal
+2014–15,0.4960187847296924,56810000.0,Portugal
+2015–16,0.5264034577922594,58311666.666666664,Portugal
+2016–17,0.5495889590377779,62653333.333333336,Portugal
+2017–18,0.5739048970689851,66285555.5,Portugal
+2018–19,0.6245746519586375,76535000.0,Portugal
+2019–20,0.6045455184308435,61487777.777777776,Portugal
+2020–21,0.5976790618273584,83166666.6111111,Portugal
+2021–22,0.5904649510943162,92032777.77777778,Portugal
+2022–23,0.6132456987830703,94880555.55555555,Portugal
+2023–24,0.5920322374736268,101346666.66666667,Portugal
+2024–25,0.6401313407032617,115400555.55555555,Portugal
+2025–26,0.6073487609491816,124692222.22222222,Portugal

Data/Number of players used.csv

@@ -0,0 +1,163 @@
+Season,Average number of players,Country
+1999–20,27.45,England
+2000–01,28.05,England
+2001–02,27.05,England
+2002–03,26.9,England
+2003–04,26.6,England
+2004–05,26.65,England
+2005–06,27.7,England
+2006–07,27.55,England
+2007–08,27.7,England
+2008–09,28.15,England
+2009–10,28.25,England
+2010–11,28.15,England
+2011–12,28.05,England
+2012–13,26.85,England
+2013–14,28.05,England
+2014–15,27.45,England
+2015–16,28.05,England
+2016–17,27.15,England
+2017–18,26.45,England
+2018–19,25.4,England
+2019–20,26.1,England
+2020–21,26.6,England
+2021–22,27.3,England
+2022–23,28.45,England
+2023–24,29.0,England
+2024–25,28.7,England
+2025–26,27.55,England
+1999–20,25.85,Spain
+2000–01,25.8,Spain
+2001–02,26.2,Spain
+2002–03,27.4,Spain
+2003–04,26.95,Spain
+2004–05,27.15,Spain
+2005–06,27.45,Spain
+2006–07,26.2,Spain
+2007–08,27.25,Spain
+2008–09,26.55,Spain
+2009–10,27.25,Spain
+2010–11,27.75,Spain
+2011–12,27.6,Spain
+2012–13,27.55,Spain
+2013–14,27.2,Spain
+2014–15,26.6,Spain
+2015–16,27.3,Spain
+2016–17,27.75,Spain
+2017–18,28.65,Spain
+2018–19,27.2,Spain
+2019–20,28.5,Spain
+2020–21,29.1,Spain
+2021–22,30.85,Spain
+2022–23,29.8,Spain
+2023–24,30.45,Spain
+2024–25,30.05,Spain
+2025–26,30.55,Spain
+1999–20,25.22222222222222,Germany
+2000–01,26.666666666666668,Germany
+2001–02,26.22222222222222,Germany
+2002–03,26.22222222222222,Germany
+2003–04,26.38888888888889,Germany
+2004–05,25.77777777777778,Germany
+2005–06,26.11111111111111,Germany
+2006–07,26.22222222222222,Germany
+2007–08,26.833333333333332,Germany
+2008–09,26.61111111111111,Germany
+2009–10,27.166666666666668,Germany
+2010–11,26.5,Germany
+2011–12,27.166666666666668,Germany
+2012–13,26.61111111111111,Germany
+2013–14,26.444444444444443,Germany
+2014–15,25.666666666666668,Germany
+2015–16,27.333333333333332,Germany
+2016–17,26.444444444444443,Germany
+2017–18,26.88888888888889,Germany
+2018–19,26.333333333333332,Germany
+2019–20,27.666666666666668,Germany
+2020–21,28.055555555555557,Germany
+2021–22,29.055555555555557,Germany
+2022–23,28.61111111111111,Germany
+2023–24,28.166666666666668,Germany
+2024–25,27.333333333333332,Germany
+2025–26,28.166666666666668,Germany
+1999–20,26.5,Italy
+2000–01,27.444444444444443,Italy
+2001–02,27.72222222222222,Italy
+2002–03,27.38888888888889,Italy
+2003–04,27.5,Italy
+2004–05,27.5,Italy
+2005–06,27.9,Italy
+2006–07,27.55,Italy
+2007–08,27.15,Italy
+2008–09,28.5,Italy
+2009–10,28.8,Italy
+2010–11,28.55,Italy
+2011–12,29.3,Italy
+2012–13,31.3,Italy
+2013–14,30.5,Italy
+2014–15,30.45,Italy
+2015–16,29.2,Italy
+2016–17,29.1,Italy
+2017–18,27.6,Italy
+2018–19,28.45,Italy
+2019–20,30.0,Italy
+2020–21,30.95,Italy
+2021–22,31.6,Italy
+2022–23,30.15,Italy
+2023–24,30.8,Italy
+2024–25,31.7,Italy
+2025–26,30.45,Italy
+1999–20,26.055555555555557,France
+2000–01,26.833333333333332,France
+2001–02,28.88888888888889,France
+2002–03,27.7,France
+2003–04,27.6,France
+2004–05,26.05,France
+2005–06,27.7,France
+2006–07,26.15,France
+2007–08,27.0,France
+2008–09,27.3,France
+2009–10,26.45,France
+2010–11,27.15,France
+2011–12,27.4,France
+2012–13,28.1,France
+2013–14,27.9,France
+2014–15,27.4,France
+2015–16,29.45,France
+2016–17,28.75,France
+2017–18,27.65,France
+2018–19,28.0,France
+2019–20,27.1,France
+2020–21,29.25,France
+2021–22,30.2,France
+2022–23,30.3,France
+2023–24,30.0,France
+2024–25,30.72222222222222,France
+2025–26,31.166666666666668,France
+1999–20,27.166666666666668,Portugal
+2000–01,26.944444444444443,Portugal
+2001–02,27.833333333333332,Portugal
+2002–03,27.61111111111111,Portugal
+2003–04,28.055555555555557,Portugal
+2004–05,27.27777777777778,Portugal
+2005–06,27.77777777777778,Portugal
+2006–07,28.6875,Portugal
+2007–08,27.875,Portugal
+2008–09,27.8125,Portugal
+2009–10,28.8125,Portugal
+2010–11,27.625,Portugal
+2011–12,29.4375,Portugal
+2012–13,28.25,Portugal
+2013–14,29.4375,Portugal
+2014–15,29.61111111111111,Portugal
+2015–16,30.27777777777778,Portugal
+2016–17,30.333333333333332,Portugal
+2017–18,28.88888888888889,Portugal
+2018–19,30.055555555555557,Portugal
+2019–20,31.22222222222222,Portugal
+2020–21,30.166666666666668,Portugal
+2021–22,32.27777777777778,Portugal
+2022–23,32.22222222222222,Portugal
+2023–24,29.77777777777778,Portugal
+2024–25,32.5,Portugal
+2025–26,32.166666666666664,Portugal

Data/Points difference.csv

@@ -0,0 +1,163 @@
+Season,First_and_second,First_and_CL,First_and_relegated,CL_and_relegated,Winners_GD,Top_4_total_GD,Country
+1999–20,0.9,1.2,2.9,1.7,2.6,5.9,England
+2000–01,0.5,0.6,2.3,1.7,2.4,6.3,England
+2001–02,0.35,0.8,2.55,1.75,2.15,7.2,England
+2002–03,0.25,0.8,2.05,1.25,2,6.4,England
+2003–04,0.55,1.5,2.85,1.35,2.35,6.55,England
+2004–05,0.6,1.7,3.1,1.4,2.85,6.95,England
+2005–06,0.4,1.2,2.85,1.65,2.5,7.85,England
+2006–07,0.3,1.05,2.55,1.5,2.8,7.7,England
+2007–08,0.1,0.55,2.55,2,2.9,8.95,England
+2008–09,0.2,0.9,2.8,1.9,2.2,8.45,England
+2009–10,0.05,0.8,2.8,2,3.55,9.85,England
+2010–11,0.45,0.6,2.05,1.45,2.05,6.65,England
+2011–12,0,1,2.65,1.65,3.2,8.5,England
+2012–13,0.55,0.8,2.65,1.85,2.15,7.3,England
+2013–14,0.1,0.35,2.65,2.3,3.25,9.35,England
+2014–15,0.4,0.85,2.6,1.75,2.05,7.3,England
+2015–16,0.5,0.75,2.2,1.45,1.6,6.25,England
+2016–17,0.35,0.85,2.95,2.1,2.6,9.45,England
+2017–18,0.95,1.25,3.35,2.1,3.95,10.15,England
+2018–19,0.05,1.35,3.2,1.85,3.6,9.55,England
+2019–20,0.9,1.65,3.25,1.6,2.6,8.2,England
+2020–21,0.6,0.95,2.9,1.95,2.55,6.4,England
+2021–22,0.05,1.1,2.9,1.8,3.65,10.65,England
+2022–23,0.25,0.9,2.75,1.85,3.05,7.8,England
+2023–24,0.1,1.15,3.25,2.1,3.1,9.2,England
+2024–25,0.5,0.75,2.95,2.2,2.25,6.45,England
+2025–26,0.35,1,2.3,1.3,2.2,5.6,England
+1999–20,0.25,0.3,1.35,1.05,1.1,4.3,Spain
+2000–01,0.35,0.85,1.95,1.1,2.05,5.55,Spain
+2001–02,0.35,0.55,1.75,1.2,1.2,5.05,Spain
+2002–03,0.1,0.85,2.1,1.25,2.2,4.95,Spain
+2003–04,0.25,0.35,1.8,1.45,2.2,5.6,Spain
+2004–05,0.2,1.1,2.35,1.25,2.2,6.35,Spain
+2005–06,0.6,0.7,2.15,1.45,2.25,5.3,Spain
+2006–07,0,0.5,1.85,1.35,1.3,5.75,Spain
+2007–08,0.4,1.05,2.15,1.1,2.4,6.15,Spain
+2008–09,0.45,1,2.25,1.25,3.5,6.95,Spain
+2009–10,0.15,1.8,3.15,1.35,3.7,8.8,Spain
+2010–11,0.2,1.7,2.65,0.95,3.7,8.65,Spain
+2011–12,0.45,2.1,2.95,0.85,4.45,9.5,Spain
+2012–13,0.75,1.7,3.2,1.5,3.75,9.55,Spain
+2013–14,0.15,1,2.55,1.55,2.55,10.55,Spain
+2014–15,0.1,0.85,2.95,2.1,4.45,12.25,Spain
+2015–16,0.05,1.35,2.65,1.3,4.15,10.65,Spain
+2016–17,0.15,1.05,3.1,2.05,3.25,10.35,Spain
+2017–18,0.7,1,3.2,2.2,3.5,9.15,Spain
+2018–19,0.55,1.3,2.5,1.2,2.7,5.65,Spain
+2019–20,0.25,0.85,2.55,1.7,2.25,6.85,Spain
+2020–21,0.1,0.45,2.6,2.15,2.1,7.4,Spain
+2021–22,0.65,0.8,2.4,1.6,2.45,6.2,Spain
+2022–23,0.5,0.85,2.4,1.55,2.5,7.1,Spain
+2023–24,0.5,0.95,3.1,2.15,3.05,8.1,Spain
+2024–25,0.2,0.9,2.4,1.5,3.15,8.3,Spain
+2025–26,0.4,1.25,2.6,1.35,2.95,7.25,Spain
+1999–20,0,1.111111111,2.111111111,1,2.5,6.333333333,Germany
+2000–01,0.055555556,0.333333333,1.555555556,1.222222222,1.388888889,4.944444444,Germany
+2001–02,0.055555556,0.5,2.222222222,1.722222222,1.611111111,7.277777778,Germany
+2002–03,0.888888889,1.055555556,2.166666667,1.111111111,2.5,5.166666667,Germany
+2003–04,0.333333333,0.555555556,2.333333333,1.777777778,2.277777778,7.444444444,Germany
+2004–05,0.777777778,1.055555556,2.333333333,1.277777778,2.333333333,6.166666667,Germany
+2005–06,0.277777778,0.777777778,2.333333333,1.555555556,1.944444444,6.444444444,Germany
+2006–07,0.111111111,0.555555556,2,1.444444444,1.333333333,5.333333333,Germany
+2007–08,0.555555556,1.222222222,2.5,1.277777778,2.611111111,6.722222222,Germany
+2008–09,0.111111111,0.333333333,2.166666667,1.833333333,2.166666667,5.277777778,Germany
+2009–10,0.277777778,0.611111111,2.166666667,1.555555556,2.277777778,6.722222222,Germany
+2010–11,0.388888889,0.833333333,2.166666667,1.333333333,2.5,6.111111111,Germany
+2011–12,0.444444444,1.166666667,2.777777778,1.611111111,3.055555556,9.166666667,Germany
+2012–13,1.388888889,2,3.333333333,1.333333333,4.444444444,8.5,Germany
+2013–14,1.055555556,1.611111111,3.5,1.888888889,3.944444444,8.444444444,Germany
+2014–15,0.555555556,1,2.444444444,1.444444444,3.444444444,8.222222222,Germany
+2015–16,0.555555556,1.833333333,2.888888889,1.055555556,3.5,8,Germany
+2016–17,0.833333333,1.111111111,2.5,1.388888889,3.722222222,8.5,Germany
+2017–18,1.166666667,1.611111111,2.833333333,1.222222222,3.555555556,6.388888889,Germany
+2018–19,0.111111111,1.111111111,2.777777778,1.666666667,3.111111111,8,Germany
+2019–20,0.722222222,0.944444444,2.833333333,1.888888889,3.777777778,10.05555556,Germany
+2020–21,0.722222222,0.944444444,2.5,1.555555556,3.055555556,7.555555556,Germany
+2021–22,0.444444444,1.055555556,2.444444444,1.388888889,3.333333333,8.944444444,Germany
+2022–23,0,0.5,2.111111111,1.611111111,3,7.166666667,Germany
+2023–24,0.944444444,1.388888889,3.166666667,1.777777778,3.611111111,10.61111111,Germany
+2024–25,0.8125,1.5625,3.3125,1.75,4.1875,8.625,Germany
+2025–26,1,1.6875,3.75,2.0625,5.375,10.1875,Germany
+1999–20,0.055555556,0.777777778,2.555555556,1.777777778,1.722222222,5.777777778,Italy
+2000–01,0.111111111,1.055555556,2.166666667,1.111111111,1.944444444,6.555555556,Italy
+2001–02,0.055555556,0.888888889,2.388888889,1.5,2.277777778,6.444444444,Italy
+2002–03,0.388888889,0.666666667,2.333333333,1.666666667,1.944444444,6.166666667,Italy
+2003–04,0.611111111,1.277777778,2.888888889,1.611111111,2.277777778,7.611111111,Italy
+2004–05,0.35,1.2,2.2,1,2,5.95,Italy
+2005–06,0.35,1.1,2.35,1.25,1.9,6.25,Italy
+2006–07,1.1,1.8,2.9,1.1,2.3,6.65,Italy
+2007–08,0.15,0.95,2.45,1.5,2.15,6.45,Italy
+2008–09,0.5,0.8,2.5,1.7,1.9,6,Italy
+2009–10,0.1,0.75,2.35,1.6,2.05,4.85,Italy
+2010–11,0.3,0.8,2.3,1.5,2.05,5.5,Italy
+2011–12,0.2,1.1,2.4,1.3,2.4,5.75,Italy
+2012–13,0.45,0.85,2.75,1.9,2.35,7,Italy
+2013–14,0.85,1.85,3.5,1.65,2.85,8.15,Italy
+2014–15,0.85,1.15,2.65,1.5,2.4,5.95,Italy
+2015–16,0.45,1.2,2.65,1.45,2.75,7.85,Italy
+2016–17,0.2,0.95,2.95,2,2.5,8.9,Italy
+2017–18,0.2,1.15,3,1.85,3.1,8.95,Italy
+2018–19,0.55,1.05,2.6,1.55,2,6.65,Italy
+2019–20,0.05,0.25,2.4,2.15,1.65,8.25,Italy
+2020–21,0.6,0.65,2.9,2.25,2.7,8.45,Italy
+2021–22,0.1,0.8,2.8,2,1.9,7.65,Italy
+2022–23,0.8,1,2.95,1.95,2.45,6.45,Italy
+2023–24,0.95,1.25,2.95,1.7,3.35,7.35,Italy
+2024–25,0.05,0.6,2.55,1.95,1.6,7,Italy
+2025–26,0.55,0.8,2.65,1.85,2.7,7,Italy
+1999–20,0.388888889,0.611111111,1.277777778,0.666666667,1.722222222,3.333333333,France
+2000–01,0.222222222,0.611111111,1.722222222,1.111111111,1.222222222,4.444444444,France
+2001–02,0.111111111,0.444444444,1.722222222,1.277777778,1.666666667,4.666666667,France
+2002–03,0.05,0.2,1.5,1.3,1.1,4.05,France
+2003–04,0.15,0.7,2.05,1.35,1.9,5.75,France
+2004–05,0.6,1.2,1.85,0.65,1.7,4.05,France
+2005–06,0.75,1.2,2.55,1.35,2.1,4.95,France
+2006–07,0.85,1.2,2.1,0.9,1.85,3.05,France
+2007–08,0.2,0.95,1.95,1,1.85,4.55,France
+2008–09,0.15,0.8,2.15,1.35,1.5,5.15,France
+2009–10,0.3,0.4,2.3,1.9,1.65,5.2,France
+2010–11,0.4,0.8,1.6,0.8,1.6,4.55,France
+2011–12,0.15,0.9,2.2,1.3,1.7,5.7,France
+2012–13,0.6,0.95,2.25,1.3,2.3,4.3,France
+2013–14,0.45,1,2.45,1.45,3.05,6.75,France
+2014–15,0.4,0.7,2.3,1.6,2.35,7.25,France
+2015–16,1.55,1.65,2.85,1.2,4.15,6.55,France
+2016–17,0.4,1.4,2.95,1.55,3.8,9.4,France
+2017–18,0.65,0.8,2.8,2,3.95,9.8,France
+2018–19,0.8,1.25,2.85,1.6,3.5,7.3,France
+2019–20,0.6,0.95,2.05,1.1,2.55,4.25,France
+2020–21,0.05,0.35,2.15,1.8,2.05,8.55,France
+2021–22,0.75,1,2.7,1.7,2.7,7.3,France
+2022–23,0.05,0.85,2.95,2.1,2.45,7.25,France
+2024–25,1.1875,1.5,3.1875,1.6875,3.5625,8.1875,France
+2025–26,0.375,1,2.75,1.75,2.8125,6.5,France
+2023–24,0.5,0.944444444,2.611111111,1.666666667,2.666666667,6.166666667,France
+1999–20,0.222222222,1.222222222,2.444444444,1.222222222,1.944444444,6.055555556,Portugal
+2000–01,0.055555556,1.111111111,2.5,1.388888889,2.277777778,6.444444444,Portugal
+2001–02,0.277777778,0.666666667,2.5,1.833333333,2.722222222,7.944444444,Portugal
+2002–03,0.611111111,2,2.777777778,0.777777778,2.611111111,6.055555556,Portugal
+2003–04,0.444444444,1.444444444,2.611111111,1.166666667,2.444444444,7,Portugal
+2004–05,0.166666667,0.388888889,1.722222222,1.333333333,1.111111111,4.444444444,Portugal
+2005–06,0.388888889,1.166666667,2.5,1.333333333,2.111111111,5.666666667,Portugal
+2006–07,0.0625,1.1875,2.9375,1.75,2.8125,7.75,Portugal
+2007–08,1.25,1.4375,3.6875,2.25,2.9375,5.8125,Portugal
+2008–09,0.25,1.125,2.9375,1.8125,2.6875,6.5625,Portugal
+2009–10,0.3125,1.75,3.4375,1.6875,3.625,9.125,Portugal
+2010–11,1.3125,2.375,3.8125,1.4375,3.5625,6.8125,Portugal
+2011–12,0.375,1,3.5,2.5,3.125,8.75,Portugal
+2012–13,0.0625,1.625,3.4375,1.8125,3.5,8.875,Portugal
+2013–14,0.4375,1.25,3.125,1.875,2.5,7.625,Portugal
+2014–15,0.166666667,1.5,3.166666667,1.666666667,3.888888889,10.88888889,Portugal
+2015–16,0.111111111,1.666666667,3.222222222,1.555555556,3.666666667,10,Portugal
+2016–17,0.333333333,1.111111111,2.777777778,1.666666667,3,8.277777778,Portugal
+2017–18,0.388888889,0.722222222,3.166666667,2.444444444,3.555555556,11.44444444,Portugal
+2018–19,0.111111111,1.111111111,3.055555556,1.944444444,4,10.22222222,Portugal
+2019–20,0.277777778,1.222222222,2.666666667,1.444444444,2.888888889,7.388888889,Portugal
+2020–21,0.277777778,1.166666667,2.833333333,1.666666667,2.5,8.444444444,Portugal
+2021–22,0.333333333,1.444444444,3.444444444,2,3.555555556,10.16666667,Portugal
+2022–23,0.111111111,0.722222222,3.388888889,2.666666667,3.444444444,10.94444444,Portugal
+2023–24,0.555555556,1.222222222,3.222222222,2,3.722222222,9.611111111,Portugal
+2024–25,0.125,1,3.4375,2.4375,3.8125,11.0625,Portugal
+2025–26,0.375,1.8125,3.625,1.8125,3,11.875,Portugal

Data/Variation in leagues performance.csv

@@ -0,0 +1,25 @@
+Stage,Unique teams,Country
+Champions League,17,Spain
+European,25,Spain
+Relegated,42,Spain
+Winners,5,Spain
+Champions League,11,England
+European,21,England
+Relegated,39,England
+Winners,6,England
+Champions League,15,Germany
+European,23,Germany
+Relegated,32,Germany
+Winners,6,Germany
+Champions League,17,Italy
+European,26,Italy
+Relegated,40,Italy
+Winners,6,Italy
+Champions League,17,France
+European,25,France
+Relegated,35,France
+Winners,8,France
+Champions League,10,Portugal
+European,23,Portugal
+Relegated,36,Portugal
+Winners,4,Portugal

Dockerfile

@@ -0,0 +1,36 @@
+FROM rocker/shiny:4.4.0
+
+# System dependencies commonly required by R packages
+RUN apt-get update && apt-get install -y \
+    libcurl4-openssl-dev \
+    libssl-dev \
+    libxml2-dev \
+    libgit2-dev \
+    libfontconfig1-dev \
+    libharfbuzz-dev \
+    libfribidi-dev \
+    libfreetype6-dev \
+    libpng-dev \
+    libjpeg-dev \
+    libtiff5-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+WORKDIR /srv/shiny-server
+
+# Copy the application
+COPY . /srv/shiny-server/
+
+# Prevent renv attempting to use pak
+ENV RENV_CONFIG_PAK_ENABLED=FALSE
+
+# Install renv and restore project packages
+RUN R -e "install.packages('renv', repos='https://cran.rstudio.com'); \
+          renv::init(bare = TRUE); \
+          renv::restore(prompt = FALSE)"
+
+# Permissions
+RUN chown -R shiny:shiny /srv/shiny-server
+
+EXPOSE 3838
+
+CMD ["/usr/bin/shiny-server"]

README.md

@@ -0,0 +1,3 @@
+# football_competitiveness
+
+Exploration into the impact of competitiveness within a domestic football league on the teams performance in European competition

clustering.R

@@ -0,0 +1,200 @@
+##k-means clustering----
+#using all data
+clustering_data_all <- scaled_full_data %>%
+  select(-c(Country, Season, season_start, European_performance, lead_european_performance))
+#considering optimal clusters
+n <- 10
+wss <- numeric(n)
+set.seed(123)
+for (i in 1:n) {
+  km.out <- kmeans(clustering_data_all, centers = i, nstart = 20)
+  wss[i] <- km.out$tot.withinss
+}
+wss_df <- tibble(clusters = 1:n, wss = wss)
+ggplot(wss_df, aes(x = clusters, y = wss, group = 1)) +
+  geom_point(size = 4)+
+  geom_line() +
+  scale_x_continuous(breaks = c(2, 4, 6, 8, 10)) +
+  xlab('Number of clusters')
+#perform clustering
+set.seed(123)
+kmeans_model_all <- kmeans(clustering_data_all, centers = 3, nstart = 50)
+#joining to data and considering characteristics of clusters
+clustered_data_all <- pivoted_scaled_full_data_comp_breakdown %>%
+  mutate(cluster = as.factor(kmeans_model_all$cluster))
+cluster_characteristics_all <- clustered_data_all %>%
+  group_by(cluster) %>%
+  summarise(across(where(is.numeric), mean, na.rm = TRUE)) %>%
+  select(-season_start)
+
+
+#using principal components
+clustering_data_pc <- pca_results %>%
+  select(contains("PC"))
+#considering optimal clusters
+n <- 10
+wss <- numeric(n)
+set.seed(123)
+for (i in 1:n) {
+  km.out <- kmeans(clustering_data_pc, centers = i, nstart = 20)
+  wss[i] <- km.out$tot.withinss
+}
+wss_df <- tibble(clusters = 1:n, wss = wss)
+ggplot(wss_df, aes(x = clusters, y = wss, group = 1)) +
+  geom_point(size = 4)+
+  geom_line() +
+  scale_x_continuous(breaks = c(2, 4, 6, 8, 10)) +
+  xlab('Number of clusters')
+#perform clustering
+set.seed(123)
+kmeans_model_pc <- kmeans(clustering_data_pc, centers = 3, nstart = 50)
+#joining to data and considering characteristics of clusters
+clustered_data_pc <- pivoted_scaled_full_data_comp_breakdown %>%
+  mutate(cluster = as.factor(kmeans_model_pc$cluster))
+cluster_characteristics_pc <- clustered_data_pc %>%
+  group_by(cluster) %>%
+  summarise(across(where(is.numeric), mean, na.rm = TRUE)) %>%
+  select(-season_start)
+
+
+#using competitiveness indices
+clustering_data_comp_index <- scaled_full_data_competitiveness_index %>%
+  select(contains("index")) %>%
+  select(1:3)
+#considering optimal clusters
+n <- 10
+wss <- numeric(n)
+set.seed(123)
+for (i in 1:n) {
+  km.out <- kmeans(clustering_data_comp_index, centers = i, nstart = 20)
+  wss[i] <- km.out$tot.withinss
+}
+wss_df <- tibble(clusters = 1:n, wss = wss)
+ggplot(wss_df, aes(x = clusters, y = wss, group = 1)) +
+  geom_point(size = 4)+
+  geom_line() +
+  scale_x_continuous(breaks = c(2, 4, 6, 8, 10)) +
+  xlab('Number of clusters')
+#perform clustering
+set.seed(123)
+kmeans_model_comp_index <- kmeans(clustering_data_comp_index, centers = 3, nstart = 50)
+#joining to data and considering characteristics of clusters
+clustered_data_comp_index <- pivoted_scaled_full_data_comp_breakdown_competitiveness_index %>%
+  mutate(cluster = as.factor(kmeans_model_comp_index$cluster))
+cluster_characteristics_comp_index <- clustered_data_comp_index %>%
+  group_by(cluster) %>%
+  summarise(across(where(is.numeric), mean, na.rm = TRUE)) %>%
+  select(-season_start)
+plotting_cluster_data <- cluster_characteristics_comp_index %>% select(-11:-19)
+
+
+##Evaluating clusters----
+compare_clusters <- function(data, km_model, name) {
+  dist_matrix <- dist(data)
+  sil <- silhouette(km_model$cluster, dist_matrix)
+  stats <- cluster.stats(dist_matrix, km_model$cluster)
+  
+  tibble(
+    model = name,
+    avg_silhouette = mean(sil[, 3]),
+    dunn_index = stats$dunn,
+    within_ss = km_model$tot.withinss,
+    between_ss = km_model$betweenss,
+    total_ss = km_model$totss,
+    variance_explained = km_model$betweenss / km_model$totss
+  )
+}
+
+comparison <- bind_rows(
+  compare_clusters(clustering_data_all, kmeans_model_all, "All variables"),
+  compare_clusters(clustering_data_pc, kmeans_model_pc, "Principal components"),
+  compare_clusters(clustering_data_comp_index, kmeans_model_comp_index, "Theory-based indices")
+) %>%
+  arrange(desc(avg_silhouette), desc(dunn_index), desc(variance_explained))
+
+
+##evaluating 'best' (indices) model
+#full European performance
+anova_model <- aov(European_performance ~ cluster, data = clustered_data_comp_index)
+summary(anova_model)
+
+kruskal.test(European_performance ~ cluster, data = clustered_data_comp_index)
+
+panel_clusters <- plm(
+  European_performance ~ cluster,
+  data = clustered_data_comp_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_clusters)
+coeftest(panel_clusters, vcov = vcovHC, type = "HC1")
+
+#lead European performance
+anova_model <- aov(lead_european_performance ~ cluster, data = clustered_data_comp_index)
+summary(anova_model)
+
+kruskal.test(lead_european_performance ~ cluster, data = clustered_data_comp_index)
+
+panel_clusters_lead <- plm(
+  lead_european_performance ~ cluster,
+  data = clustered_data_comp_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_clusters_lead)
+coeftest(panel_clusters_lead, vcov = vcovHC, type = "HC1")
+
+#CL performance
+anova_model <- aov(European_performance_CL ~ cluster, data = clustered_data_comp_index)
+summary(anova_model)
+
+kruskal.test(European_performance_CL ~ cluster, data = clustered_data_comp_index)
+
+panel_clusters_CL <- plm(
+  European_performance_CL ~ cluster,
+  data = clustered_data_comp_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_clusters_CL)
+coeftest(panel_clusters_CL, vcov = vcovHC, type = "HC1")
+
+#EL performance
+anova_model <- aov(European_performance_EL ~ cluster, data = clustered_data_comp_index)
+summary(anova_model)
+
+kruskal.test(European_performance_EL ~ cluster, data = clustered_data_comp_index)
+
+panel_clusters_EL <- plm(
+  European_performance_EL ~ cluster,
+  data = clustered_data_comp_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_clusters_EL)
+coeftest(panel_clusters_EL, vcov = vcovHC, type = "HC1")
+
+
+##finding seasons closest to cluster centre
+closest_n_per_cluster <- clustered_data_comp_index %>%
+  group_by(cluster) %>%
+  group_modify(~ {
+    cluster_id <- as.character(.y$cluster)
+    center_coords <- kmeans_model_comp_index$centers[cluster_id, , drop = FALSE]
+    
+    matrix_data <- .x %>% select(contains("index")) %>% as.matrix()
+    
+    center_matrix <- matrix(center_coords, nrow = nrow(matrix_data), ncol = ncol(matrix_data), byrow = TRUE)
+    
+    .x %>% mutate(
+      Distance = sqrt(rowSums((matrix_data - center_matrix)^2))
+    )
+  }) %>%
+  ungroup() %>%
+  group_by(cluster) %>%
+  slice_min(order_by = Distance, n = 10, with_ties = FALSE) %>%
+  relocate(c(cluster, Season, Country, European_performance))

data_manipulation.R

@@ -0,0 +1,126 @@
+European_performance_data = read_csv('Data/European performance historical data.csv')
+Points_difference_data <- read_csv("Data/Points difference.csv")
+Players_used_data <- read_csv("Data/European clubs number of players used.csv")
+Top_goalscorers_data <- read_csv("Data/Average top scorers.csv")
+Market_value_data <- read_csv("Data/Market values.csv")
+gini_coeffecient_data <- Market_value_data %>%
+  select(-`Average market value of clubs`)
+all_seasons <- unique(European_performance_data$Season)
+
+overall_country_performance <- European_performance_data %>%
+  group_by(Country) %>%
+  summarise(total_performance = sum(European_performance, na.rm = TRUE)) %>%
+  arrange(desc(total_performance))
+
+#season by season performance
+country_performance_by_season <- European_performance_data %>%
+  filter(Country %in% c("Germany", "Spain", "England", "Portugal", "Italy", "France")) %>%
+  group_by(Country, Season) %>%
+  summarise(
+    European_performance = sum(European_performance, na.rm = TRUE),
+    .groups = "drop"
+  ) %>%
+  complete(Country, Season = all_seasons, fill = list(European_performance = 0))
+#season by season performance by competition
+country_performance_by_season_comp_breakdown <- European_performance_data %>%
+  filter(Country %in% c("Germany", "Spain", "England", "Portugal", "Italy", "France")) %>%
+  group_by(Country, Season) %>%
+  summarise(
+    CL = sum(European_performance[Competition == "CL"], na.rm = TRUE),
+    EL = sum(European_performance[Competition == "EL"], na.rm = TRUE),
+    .groups = "drop"
+  ) %>%
+  complete(Country, Season = all_seasons, fill = list(CL = 0, EL = 0)) %>%
+  pivot_longer(cols = c("CL", "EL"), names_to = "competition", values_to = "European_performance") %>%
+  mutate(European_performance = ifelse(competition == "CL", European_performance/2, European_performance)) #halving score for CL as had double weighting
+
+#join all metrics data together
+#impute missing Gini coefficient with average for league - assumed the same over time
+season_league_metrics_dataset <- Points_difference_data %>%
+  left_join(Players_used_data, by = c("Country", "Season")) %>%
+  left_join(Top_goalscorers_data, by = c("Country", "Season")) %>%
+  left_join(gini_coeffecient_data, by = c("Country", "Season")) %>%
+  mutate(season_start = as.integer(str_extract(Season, "^\\d{4}"))) %>%
+  group_by(Country) %>%
+  mutate(Gini_coefficient = coalesce(Gini_coefficient, mean(Gini_coefficient, na.rm = TRUE))) %>%
+  ungroup()
+
+#average country performance
+average_country_metrics <- season_league_metrics_dataset %>%
+  group_by(Country) %>%
+  summarise(across(!contains("eason"), mean, na.rm = TRUE))
+
+#scaling to min-max of variables
+average_country_metrics_scaled <- average_country_metrics
+average_country_metrics_scaled[,-1] <- lapply(average_country_metrics[,-1], function(x) {
+  (x - min(x)) / (max(x) - min(x))
+})
+
+#scaling metrics normally
+scaled_metrics_data <- season_league_metrics_dataset %>%
+  mutate(across(
+    c(First_and_second, First_and_CL, First_and_relegated, CL_and_relegated,
+      Winners_GD, Top_4_total_GD, Average_number_of_players,
+      Average_goals_by_top_3_players, Gini_coefficient),
+    ~ as.numeric(scale(.))
+  ))
+
+#join metrics data to European performance data
+#calculate European performance in next season as lead
+season_league_full_dataset <- season_league_metrics_dataset %>%
+  left_join(country_performance_by_season, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#join metrics data to European performance data by competition
+#calculate European performance in next season as lead
+season_league_full_dataset_comp_breakdown <- season_league_metrics_dataset %>%
+  left_join(country_performance_by_season_comp_breakdown, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country, competition) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#full data scaled
+scaled_full_data <- scaled_metrics_data %>%
+  left_join(country_performance_by_season, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#data by competition scaled
+scaled_full_data_comp_breakdown <- scaled_metrics_data %>%
+  left_join(country_performance_by_season_comp_breakdown, by = c("Country", "Season")) %>%
+  arrange(Season, Country) %>%
+  group_by(Country, competition) %>%
+  mutate(lead_european_performance = dplyr::lead(European_performance)) %>%
+  ungroup()
+
+#full dataset with competition breakdown and full
+pivoted_scaled_full_data_comp_breakdown <- scaled_full_data_comp_breakdown %>%
+  pivot_wider(names_from = competition, values_from = c(European_performance, lead_european_performance)) %>%
+  mutate(European_performance = 2*European_performance_CL + European_performance_EL,
+         lead_european_performance = 2*lead_european_performance_CL + lead_european_performance_EL)
+
+#basic profiling based on European performance
+quantile_splits <- scaled_full_data %>%
+  mutate(
+    performance_group = ntile(European_performance, 3)
+  )
+european_performance_groups_characteristics <- quantile_splits %>%
+  group_by(performance_group) %>%
+  summarise(across(where(is.numeric), mean, na.rm = TRUE)) %>%
+  select(-season_start) %>%
+  mutate(label = case_when(
+    performance_group == 1 ~ "Worst performing",
+    performance_group == 2 ~ "Middle performing",
+    performance_group == 3 ~ "Best performing"
+  )) %>%
+  mutate(performance_group = case_when(
+    label == "Worst performing" ~ 1,
+    label == "Middle performing" ~ 3,
+    label == "Best performing" ~ 2
+  ))

feature_engineering.R

@@ -0,0 +1,126 @@
+##correlation coefficient----
+cor_matrix <- season_league_full_dataset %>%
+  select(-c(Country, Season, season_start)) %>%
+  cor(use = "pairwise.complete.obs")
+
+plot_ly(
+  x = colnames(cor_matrix),
+  y = rownames(cor_matrix),
+  z = cor_matrix,
+  type = "heatmap",
+  colorscale = "RdBu",
+  zmin = -1,
+  zmax = 1,
+  showscale = TRUE
+) %>%
+  add_trace(
+    x = rep(colnames(cor_matrix), each = nrow(cor_matrix)),
+    y = rep(rownames(cor_matrix), times = ncol(cor_matrix)),
+    text = sprintf("%.2f", as.vector(cor_matrix)),
+    type = "scatter",
+    mode = "text"      ,
+    textfont = list(color = "black"),
+    hoverinfo = "none",
+    showlegend = FALSE
+  ) %>%
+  layout(
+    xaxis = list(tickangle = 45),
+    yaxis = list(autorange = "reversed")
+  )
+
+#drop multicollinearity variables
+season_league_full_dataset_no_multicollinearity <- season_league_full_dataset %>%
+  select(-c(Winners_GD, First_and_relegated))
+
+
+##competitiveness index----
+scaled_full_data_competitiveness_index <- scaled_full_data %>%
+  mutate(competitive_title_index =
+           rowMeans(cbind(-First_and_second, -First_and_CL)),
+         elite_dominance_index =
+           rowMeans(cbind(CL_and_relegated, Top_4_total_GD, Gini_coefficient,
+                          Average_goals_by_top_3_players)),
+         competitive_balance_index =
+           rowMeans(cbind(-First_and_second, -First_and_CL, -CL_and_relegated,
+                          -Top_4_total_GD, -Gini_coefficient)),
+         player_rotation_index = Average_number_of_players,
+         multilevel_competitiveness_index =
+           rowMeans(cbind(-First_and_second, -First_and_CL, -CL_and_relegated))
+  )
+#by competition breakdown
+scaled_full_data_comp_breakdown_competitiveness_index <- scaled_full_data_comp_breakdown %>%
+  mutate(competitive_title_index =
+           rowMeans(cbind(-First_and_second, -First_and_CL)),
+         elite_dominance_index =
+           rowMeans(cbind(CL_and_relegated, Top_4_total_GD, Gini_coefficient,
+                          Average_goals_by_top_3_players)),
+         competitive_balance_index =
+           rowMeans(cbind(-First_and_second, -First_and_CL, -CL_and_relegated,
+                          -Top_4_total_GD, -Gini_coefficient)),
+         player_rotation_index = Average_number_of_players,
+         multilevel_competitiveness_index =
+           rowMeans(cbind(-First_and_second, -First_and_CL, -CL_and_relegated))
+  )
+#for clustering
+pivoted_scaled_full_data_comp_breakdown_competitiveness_index <- scaled_full_data_comp_breakdown_competitiveness_index %>%
+  pivot_wider(names_from = competition, values_from = c(European_performance, lead_european_performance)) %>%
+  mutate(European_performance = 2*European_performance_CL + European_performance_EL,
+         lead_european_performance = 2*lead_european_performance_CL + lead_european_performance_EL)
+
+##principal component analysis----
+pca_data <- season_league_full_dataset %>%
+  select(-c(Season, Country, season_start, European_performance, lead_european_performance))
+pca <- prcomp(pca_data, center = TRUE, scale. = TRUE)
+#weights of each variable in the principal components
+loadings <- as.data.frame(pca$rotation) %>%
+  rownames_to_column("Variable")
+#calculating proportion of variance explained by each component
+pca.var <- pca$sdev^2
+propve <- pca.var / sum(pca.var)
+plot(cumsum(propve), xlab = "Principal Component",
+     ylab = "Cumulative Proportion of Variance Explained",
+     ylim = c(0, 1), type = "b")
+abline(h=0.9, lty = "dashed")
+which(cumsum(propve) >= 0.9)[1] #number of principal components that explain at least 90% of variance
+#values of the observations on each principal component
+pc_scores <- as.data.frame(pca$x[, 1:5])
+pc_scores["PC1"] <- pc_scores["PC1"] * -1
+pca_results <- bind_cols(
+  scaled_full_data,
+  pc_scores
+)
+#by competition breakdown
+comp_breakdown_pc_scores <- pc_scores %>%
+  slice(rep(1:n(), each = 2))
+pca_results_comp_breakdown <- bind_cols(
+  scaled_full_data_comp_breakdown,
+  comp_breakdown_pc_scores
+)
+#reformatting data for plotting composition of components
+loadings_plotting <- loadings %>%
+  select(Variable, PC1:PC5) %>%
+  pivot_longer(
+    cols = starts_with("PC"),
+    names_to = "PC",
+    values_to = "Loading"
+  ) %>%
+  mutate(Loading = ifelse(PC == "PC1", Loading*-1, Loading))
+#plotting
+ggplot(loadings_plotting %>% filter(PC == "PC4"),
+       aes(x = reorder(Variable, abs(Loading)),
+           y = Loading,
+           fill = Loading > 0)) +
+  geom_col() +
+  coord_flip() +
+  scale_fill_manual(values = c("TRUE" = "steelblue",
+                               "FALSE" = "firebrick")) +
+  labs(
+    x = NULL,
+    y = "Loading",
+    title = "PCA Loadings for Principal Components explaining 90% of variance"
+  ) +
+  theme_minimal() +
+  theme(
+    legend.position = "none",
+    strip.text = element_text(face = "bold")
+  )

football_competitiveness.Rproj

@@ -0,0 +1,13 @@
+Version: 1.0
+
+RestoreWorkspace: Default
+SaveWorkspace: Default
+AlwaysSaveHistory: Default
+
+EnableCodeIndexing: Yes
+UseSpacesForTab: Yes
+NumSpacesForTab: 2
+Encoding: UTF-8
+
+RnwWeave: Sweave
+LaTeX: pdfLaTeX

functions.R

@@ -0,0 +1,181 @@
+wrap_plotly_title <- function(title, width, type = "normal") {
+  if(type == "yaxis"){
+    max_chars <- ifelse(width <= 767, 30, 50)
+  } else {
+    max_chars <- ifelse(width <= 767, 40, 100)
+  }
+  line_break <- "<br>"
+  
+  words <- strsplit(title, " ")[[1]]
+  
+  new_title <- ""
+  current_line <- ""
+  
+  for (word in words) {
+    if (nchar(current_line) + nchar(word) + 1 > max_chars) {
+      new_title <- paste0(new_title, trimws(current_line), line_break)
+      current_line <- word
+    } else {
+      current_line <- paste(current_line, word)
+    }
+  }
+  
+  new_title <- paste0(new_title, trimws(current_line))
+  
+  return(new_title)
+}
+
+custom_plotly <- function(..., xaxis_title, yaxis_title, width, hover_info = TRUE) {
+  p <- plot_ly(...) %>%
+    layout(
+      xaxis = list(
+        title = list(text = wrap_plotly_title(xaxis_title, width), standoff = 10),
+        fixedrange = TRUE,
+        automargin = TRUE,
+        tickangle = ifelse(width <= 767, 90, "auto")
+      ),
+      yaxis = list(
+        title = list(text = wrap_plotly_title(yaxis_title, width, "yaxis"), standoff = 10),
+        rangemode = "tozero",
+        tickformat = ",d",
+        fixedrange = TRUE
+      ),
+      legend = list(
+        orientation = "h",
+        xanchor = "center",
+        x = 0.5,
+        y = ifelse(width <= 767, -0.4, -0.25)
+      ),
+      autosize = TRUE,
+      dragmode = FALSE
+    ) %>%
+    config(
+      displayModeBar = FALSE,
+      responsive = TRUE,
+      scrollZoom = FALSE,
+      doubleClick = FALSE
+    )
+  if (hover_info == TRUE){
+    p <- p %>%
+      style(hovertemplate = "%{x}<br>%{y}<extra></extra>", legendgroup = NULL)
+  }
+  
+  return(p)
+}
+
+regression_coefficient_data <- function(regression_result){
+  coef_plot <- tidy(coeftest(regression_result, vcov = vcovHC, type = "HC1")) %>%
+    mutate(conf.low = estimate - 1.96 * std.error,
+           conf.high = estimate + 1.96 * std.error) %>%
+    filter(p.value < 0.1)
+  
+  return(coef_plot)
+}
+
+coefficient_plot_function <- function(regression_result, coefficient_plotting_data = NULL, width, range_val = 10) {
+  
+  if (!is.null(regression_result)) {
+    coefficient_plotting_data <- regression_coefficient_data(regression_result)
+  }
+  
+  coefficient_plotting_data <- coefficient_plotting_data %>%
+    mutate(term = gsub("_", " ", term))
+  
+  custom_plotly(
+    data = coefficient_plotting_data,
+    x = ~estimate,
+    y = ~reorder(term, estimate),
+    type = "scatter",
+    mode = "markers",
+    marker = list(size = 10),
+    error_x = list(
+      type = "data",
+      symmetric = FALSE,
+      array = ~conf.high - estimate,
+      arrayminus = ~estimate - conf.low
+    ),
+    hovertemplate = paste(
+      "<b>%{y}</b><br>",
+      "Estimate: %{x:.3f}<br>",
+      "<extra></extra>"
+    ),
+    xaxis_title = "Estimated coefficient",
+    yaxis_title = "",
+    width = width,
+    hover_info = FALSE
+  ) %>%
+    layout(autosize = FALSE,
+           margin = list(
+             l = 225
+           ),
+      xaxis = list(
+        range = c(-range_val, range_val),
+        zeroline = FALSE,
+        fixedrange = TRUE
+      ),
+      yaxis = list(
+        title = ""
+      ),
+      shapes = list(
+        list(
+          type = "line",
+          x0 = 0, x1 = 0,
+          y0 = 0, y1 = 1,
+          yref = "paper",
+          line = list(
+            color = "grey50",
+            dash = "dash"
+          )
+        )
+      ),
+      showlegend = FALSE
+    )
+}
+
+# coefficient_plot_function <- function(regression_result, coefficient_plotting_data = NULL){
+#   if (!is.null(regression_result)){
+#     coefficient_plotting_data <- regression_coefficient_data(regression_result)
+#   }
+# 
+#   plot_ly(
+#     data = coefficient_plotting_data,
+#     x = ~estimate,
+#     y = ~reorder(term, estimate),
+#     type = "scatter",
+#     mode = "markers",
+#     marker = list(size = 10),
+#     error_x = list(
+#       type = "data",
+#       symmetric = FALSE,
+#       array = ~conf.high - estimate,
+#       arrayminus = ~estimate - conf.low
+#     ),
+#     hovertemplate = paste(
+#       "<b>%{y}</b><br>",
+#       "Estimate: %{x:.3f}<br>",
+#       "<extra></extra>"
+#     )
+#   ) %>%
+#     layout(
+#       xaxis = list(
+#         title = "Estimated coefficient",
+#         range = c(-10, 10),
+#         zeroline = FALSE
+#       ),
+#       yaxis = list(
+#         title = ""
+#       ),
+#       shapes = list(
+#         list(
+#           type = "line",
+#           x0 = 0, x1 = 0,
+#           y0 = 0, y1 = 1,
+#           yref = "paper",
+#           line = list(
+#             color = "grey50",
+#             dash = "dash"
+#           )
+#         )
+#       )
+#     )
+# }

global.R

@@ -0,0 +1,19 @@
+library(plotly)
+library(shiny)
+library(shinyjs)
+library(htmlwidgets)
+library(tidyverse)
+library(DT)
+library(plm)
+library(lmtest)
+library(cluster)
+library(fpc)
+library(broom)
+
+source("scrolly_functions.R")
+source("text.R")
+source("functions.R")
+source("data_manipulation.R")
+source("feature_engineering.R")
+source("regression.R")
+source("clustering.R")

regression.R

@@ -0,0 +1,570 @@
+##panel regression----
+##all data with multicollinearity accounted for
+# panel_all <- plm(
+#   European_performance ~
+#     First_and_second +
+#     First_and_CL +
+#     CL_and_relegated +
+#     Top_4_total_GD +
+#     Average_number_of_players +
+#     Average_goals_by_top_3_players +
+#     Gini_coefficient,
+#   data = scaled_full_data,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_all)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_all, vcov = vcovHC, type = "HC1")
+
+# #rerun with only significant vars
+# panel_all_robust <- plm(
+#   European_performance ~
+#     Top_4_total_GD +
+#     Average_goals_by_top_3_players +
+#     Gini_coefficient,
+#   data = scaled_full_data,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_all_robust)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_all_robust, vcov = vcovHC, type = "HC1")
+
+#rerun with only significant vars & theory included
+panel_all_robust_theory <- plm(
+  European_performance ~
+    First_and_second +
+    CL_and_relegated +
+    Top_4_total_GD +
+    Average_goals_by_top_3_players +
+    Gini_coefficient,
+  data = scaled_full_data,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_all_robust_theory)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_all_robust_theory, vcov = vcovHC, type = "HC1")
+
+
+##using principal components
+panel_pc <- plm(
+  European_performance ~
+    PC1 +
+    PC2 +
+    PC3 +
+    PC4 +
+    PC5,
+  data = pca_results,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_pc)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_pc, vcov = vcovHC, type = "HC1")
+
+# #rerun with only significant pcs
+# panel_pc_robust <- plm(
+#   European_performance ~
+#     PC1 +
+#     PC4,
+#   data = pca_results,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_pc_robust)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_pc_robust, vcov = vcovHC, type = "HC1")
+
+
+##using competitive indices
+panel_indices_multi <- plm(
+  European_performance ~
+    competitive_title_index +
+    elite_dominance_index +
+    player_rotation_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_multi)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_multi, vcov = vcovHC, type = "HC1")
+
+panel_indices_title <- plm(
+  European_performance ~ competitive_title_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_title)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_title, vcov = vcovHC, type = "HC1")
+
+panel_indices_dominance <- plm(
+  European_performance ~ elite_dominance_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_dominance)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_dominance, vcov = vcovHC, type = "HC1")
+
+panel_indices_competitiveness <- plm(
+  European_performance ~ competitive_balance_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_competitiveness)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_competitiveness, vcov = vcovHC, type = "HC1")
+
+panel_indices_points <- plm(
+  European_performance ~ multilevel_competitiveness_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_points)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_points, vcov = vcovHC, type = "HC1")
+
+# panel_indices_players <- plm(
+#   European_performance ~ player_rotation_index,
+#   data = scaled_full_data_competitiveness_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_players)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_players, vcov = vcovHC, type = "HC1")
+
+indices_all_plotting_data <- regression_coefficient_data(panel_indices_multi) %>%
+  mutate(term = paste0(term, " (multi)")) %>%
+  rbind(regression_coefficient_data(panel_indices_title)) %>%
+  rbind(regression_coefficient_data(panel_indices_dominance)) %>%
+  rbind(regression_coefficient_data(panel_indices_competitiveness)) %>%
+  rbind(regression_coefficient_data(panel_indices_points))
+
+
+##panel regression using next years European performance value----
+##all data with multicollinearity accounted for
+panel_all_lead <- plm(
+  lead_european_performance ~
+    First_and_second +
+    First_and_CL +
+    CL_and_relegated +
+    Top_4_total_GD +
+    Average_number_of_players +
+    Average_goals_by_top_3_players +
+    Gini_coefficient,
+  data = scaled_full_data,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_all_lead)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_all_lead, vcov = vcovHC, type = "HC1")
+
+# #rerun with only significant vars & theory included
+# panel_all_lead_robust_theory <- plm(
+#   lead_european_performance ~
+#     First_and_CL +
+#     CL_and_relegated +
+#     Top_4_total_GD +
+#     Average_goals_by_top_3_players +
+#     Gini_coefficient,
+#   data = scaled_full_data,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_all_lead_robust_theory)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_all_lead_robust_theory, vcov = vcovHC, type = "HC1")
+
+
+##using principal components
+panel_lead_pc <- plm(
+  lead_european_performance ~
+    PC1 +
+    PC2 +
+    PC3 +
+    PC4 +
+    PC5,
+  data = pca_results,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_lead_pc)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_lead_pc, vcov = vcovHC, type = "HC1")
+
+#rerun with only significant pcs
+# panel_lead_pc_robust <- plm(
+#   lead_european_performance ~
+#     PC1 +
+#     PC3 +
+#     PC4,
+#   data = pca_results,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_lead_pc_robust)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_lead_pc_robust, vcov = vcovHC, type = "HC1")
+
+
+##using competitive indices
+panel_indices_multi_lead <- plm(
+  lead_european_performance ~
+    competitive_title_index +
+    elite_dominance_index +
+    player_rotation_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_multi_lead)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_multi_lead, vcov = vcovHC, type = "HC1")
+
+# panel_indices_title_lead <- plm(
+#   lead_european_performance ~ competitive_title_index,
+#   data = scaled_full_data_competitiveness_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_title_lead)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_title_lead, vcov = vcovHC, type = "HC1")
+
+panel_indices_dominance_lead <- plm(
+  lead_european_performance ~ elite_dominance_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_dominance_lead)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_dominance_lead, vcov = vcovHC, type = "HC1")
+
+panel_indices_competitiveness_lead <- plm(
+  lead_european_performance ~ competitive_balance_index,
+  data = scaled_full_data_competitiveness_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_competitiveness_lead)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_competitiveness_lead, vcov = vcovHC, type = "HC1")
+
+# panel_indices_points_lead <- plm(
+#   lead_european_performance ~ multilevel_competitiveness_index,
+#   data = scaled_full_data_competitiveness_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_points_lead)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_points_lead, vcov = vcovHC, type = "HC1")
+
+# panel_indices_players_lead <- plm(
+#   lead_european_performance ~ player_rotation_index,
+#   data = scaled_full_data_competitiveness_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_players_lead)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_players_lead, vcov = vcovHC, type = "HC1")
+
+indices_lead_plotting_data <- regression_coefficient_data(panel_indices_multi_lead) %>%
+  mutate(term = paste0(term, " (multi)")) %>%
+  rbind(regression_coefficient_data(panel_indices_dominance_lead)) %>%
+  rbind(regression_coefficient_data(panel_indices_competitiveness_lead))
+
+
+##panel regression for CL----
+##all data with multicollinearity accounted for
+CL_scaled_data <- scaled_full_data_comp_breakdown %>% filter(competition == "CL")
+panel_all_CL <- plm(
+  European_performance ~
+    First_and_second +
+    First_and_CL +
+    CL_and_relegated +
+    Top_4_total_GD +
+    Average_number_of_players +
+    Average_goals_by_top_3_players +
+    Gini_coefficient,
+  data = CL_scaled_data,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_all_CL)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_all_CL, vcov = vcovHC, type = "HC1")
+
+#rerun with only significant vars
+# panel_all_CL_robust <- plm(
+#   European_performance ~
+#     CL_and_relegated +
+#     Top_4_total_GD +
+#     Average_goals_by_top_3_players +
+#     Gini_coefficient,
+#   data = CL_scaled_data,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_all_CL_robust)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_all_CL_robust, vcov = vcovHC, type = "HC1")
+
+
+##using principal components
+CL_pca_results <- pca_results_comp_breakdown %>% filter(competition == "CL")
+panel_pc_CL <- plm(
+  European_performance ~
+    PC1 +
+    PC2 +
+    PC3 +
+    PC4 +
+    PC5,
+  data = CL_pca_results,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_pc_CL)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_pc_CL, vcov = vcovHC, type = "HC1")
+
+#rerun with only significant pcs
+# panel_pc_CL_robust <- plm(
+#   European_performance ~
+#     PC1 +
+#     PC3 +
+#     PC4,
+#   data = CL_pca_results,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_pc_CL_robust)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_pc_CL_robust, vcov = vcovHC, type = "HC1")
+
+
+##using competitive indices
+CL_competitive_index <- scaled_full_data_comp_breakdown_competitiveness_index %>% filter(competition == "CL")
+panel_indices_multi_CL <- plm(
+  European_performance ~
+    competitive_title_index +
+    elite_dominance_index +
+    player_rotation_index,
+  data = CL_competitive_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_multi_CL)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_multi_CL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_title_CL <- plm(
+#   European_performance ~ competitive_title_index,
+#   data = CL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_title_CL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_title_CL, vcov = vcovHC, type = "HC1")
+
+panel_indices_dominance_CL <- plm(
+  European_performance ~ elite_dominance_index,
+  data = CL_competitive_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_dominance_CL)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_dominance_CL, vcov = vcovHC, type = "HC1")
+
+panel_indices_competitiveness_CL <- plm(
+  European_performance ~ competitive_balance_index,
+  data = CL_competitive_index,
+  index = c("Country","season_start"),
+  effect = "twoways",
+  model = "within"
+)
+summary(panel_indices_competitiveness_CL)
+#compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+coeftest(panel_indices_competitiveness_CL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_points_CL <- plm(
+#   European_performance ~ multilevel_competitiveness_index,
+#   data = CL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_points_CL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_points_CL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_players_CL <- plm(
+#   European_performance ~ player_rotation_index,
+#   data = CL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_players_CL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_players_CL, vcov = vcovHC, type = "HC1")
+
+
+##panel regression for EL----
+##all data with multicollinearity accounted for
+# EL_scaled_data <- scaled_full_data_comp_breakdown %>% filter(competition == "EL")
+# panel_all_EL <- plm(
+#   European_performance ~
+#     First_and_second +
+#     First_and_CL +
+#     CL_and_relegated +
+#     Top_4_total_GD +
+#     Average_number_of_players +
+#     Average_goals_by_top_3_players +
+#     Gini_coefficient,
+#   data = EL_scaled_data,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_all_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_all_EL, vcov = vcovHC, type = "HC1")
+
+
+##using principal components
+# EL_pca_results <- pca_results_comp_breakdown %>% filter(competition == "EL")
+# panel_pc_EL <- plm(
+#   European_performance ~
+#     PC1 +
+#     PC2 +
+#     PC3 +
+#     PC4 +
+#     PC5,
+#   data = EL_pca_results,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_pc_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_pc_EL, vcov = vcovHC, type = "HC1")
+
+
+##using competitive indices
+# EL_competitive_index <- scaled_full_data_comp_breakdown_competitiveness_index %>% filter(competition == "EL")
+# panel_indices_multi_EL <- plm(
+#   European_performance ~
+#     competitive_title_index +
+#     elite_dominance_index +
+#     player_rotation_index,
+#   data = EL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_multi_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_multi_EL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_title_EL <- plm(
+#   European_performance ~ competitive_title_index,
+#   data = EL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_title_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_title_EL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_dominance_EL <- plm(
+#   European_performance ~ elite_dominance_index,
+#   data = EL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_dominance_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_dominance_EL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_competitiveness_EL <- plm(
+#   European_performance ~ competitive_balance_index,
+#   data = EL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_competitiveness_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_competitiveness_EL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_points_EL <- plm(
+#   European_performance ~ multilevel_competitiveness_index,
+#   data = EL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_points_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_points_EL, vcov = vcovHC, type = "HC1")
+
+# panel_indices_players_EL <- plm(
+#   European_performance ~ player_rotation_index,
+#   data = EL_competitive_index,
+#   index = c("Country","season_start"),
+#   effect = "twoways",
+#   model = "within"
+# )
+# summary(panel_indices_players_EL)
+# #compute clustered standard errors (adjustment for autocorrelation + heteroskedasticity)
+# coeftest(panel_indices_players_EL, vcov = vcovHC, type = "HC1")
+
+indices_comp_breakdown_plotting_data <- regression_coefficient_data(panel_indices_multi_CL) %>%
+  mutate(term = paste0(term, " (CL multi)")) %>%
+  rbind(regression_coefficient_data(panel_indices_dominance_CL) %>% mutate(term = paste0(term, " (CL)"))) %>%
+  rbind(regression_coefficient_data(panel_indices_competitiveness_CL) %>% mutate(term = paste0(term, " (CL)")))

scrolly_functions.R

@@ -0,0 +1,322 @@
+
+create_scrolly_side <- function(name,css
+    
+) {
+name<-name
+css<-paste0(
+
+"
+
+
+#scrolly-side .scrolly article .scrolly-side {
+min-height: 105vh;
+margin-bottom: 1rem;
+
+}
+
+#scrolly-side .scrolly article .scrolly-side.is-active p {
+background-color: #f4f4f4;
+opacity:1;
+color:black;
+border-radius: 25px;
+
+
+}
+#scrolly-side .scrolly article .scrolly-side p {
+margin: 0;
+padding: 1rem;
+text-align: left;
+font-weight: 400;
+opacity:0.3;
+transition: background-color 250ms ease-in-out;
+color:black;
+border-radius: 25px;
+
+
+
+}
+
+
+
+/* css when screen is larger than 768px */
+
+
+@media screen and (min-width: 768px){
+
+#scrolly-side .scrolly article {
+padding: 0;
+margin: 0 auto;
+max-width:40%;
+position: relative;
+text-align: left;
+font-size: 20px;
+}
+
+#scrolly-side .scrolly {
+display: flex;           /* this is what stops content overlapping */
+max-width:100%;
+margin: 3rem auto;
+background-color:#ffffff;
+padding: 1rem;
+border-radius:0;
+gap:10px;
+}
+
+#scrolly-side .scrolly figure.sticky {
+position: sticky;
+width: 100%;
+height: 80vh;
+background: #ffffff;
+margin: 10;
+top: 10vh;
+left: 0;
+border-radius: 25px;
+order: 1;
+}
+}
+
+@media screen and (max-width: 767px){
+
+#scrolly-side .scrolly article {
+padding: 0;
+margin: 0 auto;
+max-width:100%;
+position: relative;
+text-align: left;
+font-size: 15px;
+pointer-events: none !important;
+}
+
+
+#scrolly-side .scrolly {
+max-width:90%;
+margin: 3rem auto;
+background-color: #ffffff;
+opacity:1;
+padding: 1rem;
+border-radius: 25px;
+}
+
+#scrolly-side .scrolly article .scrolly-side p {
+border-radius: 25px;
+background-color: #F4F4F4;
+opacity:1;
+color:black;
+max-width:100%;
+}
+
+#scrolly-side .scrolly figure.sticky {
+position: sticky;
+width: 100%;
+height: 40vh;
+background: #ffffff;
+margin: 10;
+top: 30vh;
+left: 0;
+border-radius: 25px;
+order: 1;
+}
+
+#scrolly-side .sticky{
+position: relative;
+z-index: 0;
+pointer-events: auto;
+}
+
+#scrolly-side .text-article{
+pointer-events: none !important;
+}
+#scrolly-side .text-article .detail-toggle-btn {
+  pointer-events: auto !important;
+}
+
+}
+
+
+
+")
+
+  
+  #### script #####
+  js<-"<script>
+const container = d3.select('#scrolly-side');
+const stepSel = container.selectAll('.step');
+
+function updateChart(index) {
+  const sel = container.select(`[data-index='${index}']`);
+  const width = sel.attr('data-width');
+  stepSel.classed('is-active', (d, i) => i === index);
+  container.select('.bar-inner').style('width', width);
+  Shiny.setInputValue(\"input1\", index)
+
+}
+
+function init() {
+  Stickyfill.add(d3.select('.sticky').node());
+
+  enterView({
+    selector: stepSel.nodes(),
+    offset: 0.5,
+    enter: el => {
+      const index = +d3.select(el).attr('data-index');
+      updateChart(index);
+    },
+    exit: el => {
+      let index = +d3.select(el).attr('data-index');
+      index = Math.max(0, index - 1);
+      updateChart(index);
+    } });
+
+}
+
+init();
+</script>"
+
+css<-gsub("scrolly-side",name,css)
+css<-gsub("step",name,css)
+js<-gsub("step",name,js)
+js<-gsub("scrolly-side",name,js)
+js<-gsub("container",paste0("container",name),js)
+js<-gsub("updateChart",paste0("updateChart",name),js)
+js<-gsub("init",paste0("init",name),js)
+js<-gsub("input1",paste0(name),js)
+
+list<-list(css,js)
+
+return(list)
+}
+
+
+##### create_scrolly_overlay ####
+
+
+create_scrolly_overlay <- function(name) {
+  name<-name
+  css<-
+    "
+#scrolly-overlay .scrolly {
+max-width: 100%;
+margin: 3rem auto;
+background: transparent;
+padding: 0;
+border-radius: 0;
+}
+#scrolly-overlay .scrolly article {
+padding: 0;
+max-width: 100rem;
+margin: 0 auto;
+color:black;
+position: relative;
+text-align: left;
+font-size: 20px;
+border-radius: 25px;
+pointer-events: none !important;
+}
+#scrolly-overlay .scrolly article .scrolly-overlay {
+min-height: 105vh;
+margin-bottom: 1rem;
+border-radius: 25px;
+
+}
+#scrolly-overlay .scrolly article .scrolly-overlay:last-of-type {
+margin-bottom: 0;
+border-radius: 25px;
+
+}
+#scrolly-overlay .scrolly article .scrolly-overlay.is-active p {
+border-radius: 25px;
+color:white;
+background-color: #e7e7e7  ;
+opacity:0.8;
+
+color:black;
+
+}
+#scrolly-overlay .scrolly article .scrolly-overlay p {
+margin: 0;
+padding: 1rem;
+text-align: center;
+font-weight: 400;
+background-color: #e7e7e7 ;
+opacity:0.3;
+
+transition: background-color 250ms ease-in-out;
+border-radius: 25px;
+
+  }
+#scrolly-overlay .scrolly figure.sticky {
+position: sticky;
+width: 100%;
+height: 98vh;
+background:white;
+margin: 0;
+top: 1vh;
+left: 0;
+border-radius: 25px;
+
+}
+
+#scrolly-overlay .sticky{
+position: relative;
+z-index: 0;
+pointer-events: auto;
+}
+
+#scrolly-overlay .text-article{
+pointer-events: none !important;
+}
+#scrolly-overlay .text-article .detail-toggle-btn {
+  pointer-events: auto !important;
+}
+"
+
+#### script #####
+js<-"<script>
+const container = d3.select('#scrolly-overlay');
+const stepSel = container.selectAll('.step');
+
+function updateChart(index) {
+  const sel = container.select(`[data-index='${index}']`);
+  const width = sel.attr('data-width');
+  stepSel.classed('is-active', (d, i) => i === index);
+  container.select('.bar-inner').style('width', width);
+  Shiny.setInputValue(\"input1\", index)
+
+}
+
+function init() {
+  Stickyfill.add(d3.select('.sticky').node());
+
+  enterView({
+    selector: stepSel.nodes(),
+    offset: 0.5,
+    enter: el => {
+      const index = +d3.select(el).attr('data-index');
+      updateChart(index);
+    },
+    exit: el => {
+      let index = +d3.select(el).attr('data-index');
+      index = Math.max(0, index - 1);
+      updateChart(index);
+    } });
+
+}
+
+init();
+</script>"
+
+css<-gsub("scrolly-overlay",name,css)
+css<-gsub("step",name,css)
+js<-gsub("step",name,js)
+
+js<-gsub("scrolly-overlay",name,js)
+js<-gsub("container",paste0("container",name),js)
+js<-gsub("updateChart",paste0("updateChart",name),js)
+
+js<-gsub("init",paste0("init",name),js)
+js<-gsub("input1",paste0(name),js)
+
+list<-list(css,js)
+
+return(list)
+}

server.R

@@ -0,0 +1,189 @@
+server <- function(input, output, session) {
+  
+  countries <- c("Spain", "England", "Germany", "Italy", "France", "Portugal")
+  
+  #### European performance over time ####
+  time_plot_debounced <- debounce(reactive(input$time_plot), millis = 200)
+  time_plot_level <- reactive({
+    val <- time_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+  
+  time_plot_selected_data <- reactive({
+    season_league_full_dataset %>% filter(Country == countries[time_plot_level() + 1])
+  })
+  
+  output$european_performance_over_time_plot <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+    
+    custom_plotly(data = time_plot_selected_data(), x = ~Season, y = ~European_performance, type = "scatter", mode = "lines", hoverinfo = "none",
+                  xaxis_title = "Season", yaxis_title = "European performance", width = width) %>%
+      layout(yaxis = list(range = c(0, 70)))
+  })
+  
+  
+  #### League characteristics ####
+  league_averages_plot_debounced <- debounce(reactive(input$league_averages_plot), millis = 200)
+  league_averages_plot_level <- reactive({
+    val <- league_averages_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+  
+  league_averages_plot_selected_data <- reactive({
+    average_country_metrics_scaled %>% filter(Country == countries[league_averages_plot_level() + 1])
+  })
+  
+  #plotting radar chart
+  output$league_characteristics_plot <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+    
+    country_data <- league_averages_plot_selected_data()
+    values <- as.numeric(country_data[1, -1])
+    
+    custom_plotly(type = 'scatterpolar', fill = 'toself', hoverinfo = "none", xaxis_title = "", yaxis_title = "", width = width) %>%
+      add_trace(
+        r = c(values, values[1]),
+        theta = c(gsub("_", " ", names(country_data)[-1]), gsub("_", " ", names(country_data)[-1][1]))
+      ) %>%
+      layout(
+        polar = list(radialaxis = list(range = c(0,1))),
+        showlegend = FALSE
+      )
+  })
+  
+  
+  #### Basic group profiling ####
+  basic_profiless_plot_debounced <- debounce(reactive(input$basic_profiling_plot), millis = 200)
+  basic_profiless_plot_level <- reactive({
+    val <- basic_profiless_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+  
+  basic_profiless_plot_selected_data <- reactive({
+    european_performance_groups_characteristics %>% filter(performance_group == basic_profiless_plot_level() + 1)
+  })
+  
+  #plotting radar chart
+  output$basic_profiles_characteristic_plot  <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+    
+    group_data <- basic_profiless_plot_selected_data()
+    values <- as.numeric(group_data[1, 2:10])
+    
+    custom_plotly(type = 'scatterpolar', fill = 'toself', hoverinfo = "none", xaxis_title = "", yaxis_title = "", width = width) %>%
+      add_trace(
+        r = c(values, values[1]),
+        theta = c(gsub("_", " ", names(group_data)[2:10]), gsub("_", " ", names(group_data)[2:10][1]))
+      ) %>%
+      layout(
+        polar = list(radialaxis = list(range = c(-0.4,0.4))),
+        showlegend = FALSE
+      )
+  })
+  
+  
+  #### Regression coefficients one season ####
+  regression_current_year_plot_debounced <- debounce(reactive(input$regression_current_year_plot), millis = 200)
+  regression_current_year_plot_level <- reactive({
+    val <- regression_current_year_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+  
+  one_season_full_coefficients <- list(panel_all_robust_theory, panel_pc, indices_all_plotting_data)
+  coefficient_one_year_data <- reactive({
+    one_season_full_coefficients[[regression_current_year_plot_level() + 1]]
+  })
+  
+  #plotting coefficients
+  output$regression_coefficients_current_plot <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+    if (regression_current_year_plot_level() %in% c(0,1)){
+      #version with all in one
+      coefficient_plot_function(regression_result = coefficient_one_year_data(), width = width)
+    } else if (regression_current_year_plot_level() == 2){
+      #indices version (multiple regressions combined in one)
+      coefficient_plot_function(regression_result = NULL, coefficient_plotting_data = coefficient_one_year_data(), width = width)
+    }
+  })
+  
+  
+  #### Regression coefficients next season ####
+  regression_next_year_plot_debounced <- debounce(reactive(input$regression_next_year_plot), millis = 200)
+  regression_next_year_plot_level <- reactive({
+    val <- regression_next_year_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+  
+  next_season_full_coefficients <- list(panel_all_lead, panel_lead_pc, indices_lead_plotting_data)
+  coefficient_next_year_data <- reactive({
+    next_season_full_coefficients[[regression_next_year_plot_level() + 1]]
+  })
+  
+  #plotting coefficients
+  output$regression_coefficients_next_plot <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+    if (regression_next_year_plot_level() %in% c(0,1)){
+      #version with all in one
+      coefficient_plot_function(regression_result = coefficient_next_year_data(), width = width, range_val = 15)
+    } else if (regression_next_year_plot_level() == 2){
+      #indices version (multiple regressions combined in one)
+      coefficient_plot_function(regression_result = NULL, coefficient_plotting_data = coefficient_next_year_data(), width = width, range_val = 15)
+    }
+  })
+  
+  
+  #### Regression coefficients by competition ####
+  regression_competition_plot_debounced <- debounce(reactive(input$regression_competition_plot), millis = 200)
+  regression_competition_plot_level <- reactive({
+    val <- regression_competition_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+  
+  comp_breakdown_season_full_coefficients <- list(panel_all_CL, panel_pc_CL, indices_comp_breakdown_plotting_data)
+  coefficient_comp_breakdown_data <- reactive({
+    comp_breakdown_season_full_coefficients[[regression_competition_plot_level() + 1]]
+  })
+  
+  #plotting coefficients
+  output$regression_coefficients_competition_plot <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+    if (regression_competition_plot_level() %in% c(0,1)){
+      #version with all in one
+      coefficient_plot_function(regression_result = coefficient_comp_breakdown_data(), width = width, range_val = 5)
+    } else if (regression_competition_plot_level() == 2){
+      #indices version (multiple regressions combined in one)
+      coefficient_plot_function(regression_result = NULL, coefficient_plotting_data = coefficient_comp_breakdown_data(), width = width, range_val = 5)
+    }
+  })
+  
+  
+  #### Clustering ####
+  clustering_plot_debounced <- debounce(reactive(input$clustering_plot), millis = 200)
+  clustering_plot_level <- reactive({
+    val <- clustering_plot_debounced()
+    if (is.null(val)) 0 else val
+  })
+
+  clustering_plot_selected_data <- reactive({
+    plotting_cluster_data %>% filter(cluster == clustering_plot_level() + 1)
+  })
+
+  #plotting radar chart
+  output$clustering_characteristics_plot  <- renderPlotly({
+    width <- ifelse(is.null(input$screen_width), 760, input$screen_width)
+
+    group_data <- clustering_plot_selected_data()
+    values <- as.numeric(group_data[1, 2:10])
+
+    custom_plotly(type = 'scatterpolar', fill = 'toself', hoverinfo = "none", xaxis_title = "", yaxis_title = "", width = width) %>%
+      add_trace(
+        r = c(values, values[1]),
+        theta = c(gsub("_", " ", names(group_data)[2:10]), gsub("_", " ", names(group_data)[2:10][1]))
+      ) %>%
+      layout(
+        polar = list(radialaxis = list(range = c(-0.7,1.5))),
+        showlegend = FALSE
+      )
+  })
+  
+}

text.R

@@ -0,0 +1,141 @@
+#Motivation
+title <- ("Does a more competitive domestic football league improve a team's performance in European competition?")
+motivation_one <- "Does competing in a more competitive domestic league provide an advantage or disadvantage in European competition? At a theoretical level, there are arguments supporting both perspectives. Teams from more competitive leagues are regularly exposed to high-quality opposition, which may better prepare players for the intensity and standard of European matches. Conversely, teams from less competitive leagues may be able to prioritise European competition by rotating players domestically, reducing fatigue and ensuring key players are available for important continental fixtures."
+motivation_two <- "The 2024/25 UEFA Champions League provides a useful illustration of these contrasting domestic scenarios. Paris Saint-Germain (PSG), one of the finalists, secured the Ligue 1 title on 5 April 2025 with six league matches remaining. As a result, they were able to focus all of their attention on the Champions League from the quarter finals, as they had already reached the Coupe de France final (which would not be played until the week before the Champions League final). In contrast, Internazionale remained locked in a Serie A title race until the final day of the season, ultimately finishing one point behind Napoli. They had also only played the first leg of their Coppa Italia semi-final (which they lost 4-1 on aggregate to city rivals Milan) before the quarter finals begun. PSG therefore represent a club succeeding in Europe while operating in a relatively less competitive domestic environment, whereas Internazionale demonstrate that strong European performances can also be achieved while facing significant domestic competition."
+motivation_three <- "This study investigates the factors that contribute to league competitiveness and examines how these characteristics relate to the performance of clubs from each country in European competitions. By analysing multiple dimensions of competitiveness rather than treating it as a single concept, the research seeks to identify whether particular competitive environments are associated with stronger or weaker European performance."
+
+date <- ("August 2026")
+name <- ("Ciarán Fitzsimons")
+
+#Exploratory analysis
+exploratory_analysis_one <- "Before exploring the relationship between league competitiveness and European performance, it is important to understand how the major European leagues differ from one another. This section examines the six most successful leagues from 1999/2000 onwards (Spain, England, Germany, Italy, France and Portugal), summarising their performance in European competitions and the characteristics of their domestic competitions. The selected variables are introduced and compared using league averages, providing a foundation for the subsequent analysis."
+
+#European performance over time
+europe_s_top_6_leagues_european_performance_since_1999_2000_one <- "Spanish teams have been the best performing in Europe across the period. The greatest era of dominance came between 2013 and 2018, where a Spanish side won the Champions league in all five seasons (Real Madrid (4) and Barcelona (1)) and four of the five Europa league titles (Sevilla (3) and Atletico Madrid (1))."
+europe_s_top_6_leagues_european_performance_since_1999_2000_two <- "English teams have performed consistently well across Europe throughout the period, with a few dominant seasons between 2004 and 2009, where an English side reached the Champions league final in each of the seasons. 2018/19 marked the season of English dominance, where the Champions League final was contested by Liverpool and Tottenham, and the Europa League by Chelsea and Arsenal."
+europe_s_top_6_leagues_european_performance_since_1999_2000_three <- "German teams have performed consistently across the period, without reaching the heights of the Spanish and English sides. This is largely down to fewer teams performing across the continent, with Bayern Munich and Borussia Dortmund largely carrying German performance, represented by a peak in 2012/13 where the two faced each other in the Champions league final."
+europe_s_top_6_leagues_european_performance_since_1999_2000_four <- "Italian teams’ European performance shows greater volatility over the period, with a range of seasons eclipsing Germany’s greatest seasons, however many seasons of low performance by the teams. 2002/03 represented Italian footballs best European performance season, with three of the four semi finalists being Italian and the final being contested by Juventus and AC Milan."
+europe_s_top_6_leagues_european_performance_since_1999_2000_five <- "French European performance has been consistently fairly poor over the period, with a few peaks over the period. Even with PSG’s rise in recent years and back-to-back Champions league titles, France’s overall performance has been poor due to the lack of other teams performing well."
+europe_s_top_6_leagues_european_performance_since_1999_2000_six <- "Portuguese teams have performed the worst of the six leagues over the period, with a few successful seasons with Porto winning the Champions league once and Europa league twice. The teams performing averagely in Europe from Portugal have consistently been the same sides (Porto, Benfica, Sporting CP and Braga)."
+
+#Variables used
+variables_used_in_analysis_of_leagues_competitiveness_one <- "<b>Points difference</b>"
+variables_used_in_analysis_of_leagues_competitiveness_two <- "- Between first and second place"
+variables_used_in_analysis_of_leagues_competitiveness_three <- "- Between first place and Champions League qualification (fourth)"
+variables_used_in_analysis_of_leagues_competitiveness_four <- "- Between first place and relegation (third bottom)"
+variables_used_in_analysis_of_leagues_competitiveness_five <- "- Between Champions League qualification and relegation (third bottom)"
+variables_used_in_analysis_of_leagues_competitiveness_six <- "These measures capture competitiveness at different levels of the league table. Smaller point gaps indicate a more closely contested league, while larger gaps suggest greater separation between clubs. To allow comparison across leagues of different sizes, all values are divided by the number of teams in the league."
+variables_used_in_analysis_of_leagues_competitiveness_seven <- "<b>Goal difference</b>"
+variables_used_in_analysis_of_leagues_competitiveness_eight <- "- League winner's goal difference"
+variables_used_in_analysis_of_leagues_competitiveness_nine <- "- Combined goal difference of teams finishing in Champions League qualification places"
+variables_used_in_analysis_of_leagues_competitiveness_ten <- "Goal difference provides an indication of how dominant the strongest teams are within a league. Larger goal differences suggest top clubs are outperforming their domestic opponents more consistently, which may indicate lower overall competitiveness. Values are divided by the number of teams in the league to provide relative measures across countries."
+variables_used_in_analysis_of_leagues_competitiveness_eleven <- "<b>Average number of players used</b>"
+variables_used_in_analysis_of_leagues_competitiveness_twelve <- "This measures the average number of players used in league matches by clubs competing in European competitions. A higher value may indicate greater squad rotation, potentially reflecting an ability to rest players domestically while maintaining performance levels."
+variables_used_in_analysis_of_leagues_competitiveness_thirteen <- "<b>Average goals scored by the top three scorers</b>"
+variables_used_in_analysis_of_leagues_competitiveness_fourteen <- "This metric captures the extent to which the league's leading goalscorers dominate attacking output. Higher values may suggest that elite attacking players face less resistance and are therefore able to record larger goal totals."
+variables_used_in_analysis_of_leagues_competitiveness_fifteen <- "<b>Gini coefficient</b>"
+variables_used_in_analysis_of_leagues_competitiveness_sixteen <- "The Gini coefficient measures the distribution of wealth across clubs within a league, ranging from 0 (perfect equality) to 1 (complete inequality). It is calculated using squad market values for each season. Higher values indicate wealth is concentrated among a smaller number of clubs, which may provide those clubs with a competitive advantage and reduce overall league competitiveness."
+
+#League characteristics
+characteristics_of_leagues_since_1999_2000_one <- "Spain combines strong top-club performance with a heavy reliance on elite attacking players, as shown by the very high contribution from the top three scorers. Although competitive gaps across the league are relatively modest, financial inequality is among the highest of the six leagues, suggesting that a small number of wealthy clubs hold a significant economic advantage. As a result, Spanish football appears to be characterised by influential elite clubs operating within a league that remains reasonably competitive on the pitch. The characteristics mirror what we see in the real world, with financial and competitive dominance of Real Madrid and Barcelona (and to a lesser extent, Atletico Madrid), as well as superstar signings throughout the period."
+characteristics_of_leagues_since_1999_2000_two <- "England combines relatively strong competitive balance at the top of the table with a substantial gap between the Champions League places and the relegation zone. The league's top clubs perform strongly, as shown by the high aggregate goal difference among the top four teams, but financial resources are distributed more evenly than in any of the other leagues, reflected by the lowest Gini coefficient. Overall, England appears to balance strong elite clubs with relatively broad financial strength across the league. This reflects the dominance being spread between the traditional ‘big six’ clubs, with the smaller clubs having more financial resources than counterparts in other leagues due to extensive worldwide TV deals."
+characteristics_of_leagues_since_1999_2000_three <- "Germany combines strong competitive balance near the top of the table with some of the highest-performing champions. Winners typically achieve large goal differences, and the league's strongest clubs remain highly effective despite using fewer players on average than other leagues. Financial inequality is relatively low, indicating that strong on-field performances are achieved within a comparatively balanced economic structure. This is reflective of the strong “50 + 1” rule around ownership, as well as the consistent dominance of Bayern Munich, with Borussia Dortmund running them close on occasion."
+characteristics_of_leagues_since_1999_2000_four <- "Italy presents a relatively balanced profile across most indicators. There is a moderate separation between the leading clubs and the rest of the league, while squad usage is relatively high and a large share of goals comes from the top three scorers. Financial inequality is similar to that of Germany and France, suggesting that economic concentration is present but not to the extent seen in Spain or Portugal. This is reflective of the title winner changing regularly, but largely between the same four clubs, as well as weaker financial power than other countries."
+characteristics_of_leagues_since_1999_2000_five <- "France stands out for having one of the closest title races, with a relatively small gap between first and second place. However, league-wide dominance is limited, as shown by the low goal differences recorded by its strongest clubs. French clubs tend to use relatively large squads, while financial inequality is moderate, suggesting a league that is neither highly concentrated nor heavily dominated by a small elite. This is reflective of a tight and not financially powerful league (particularly with the lack of TV deal), despite PSG’s recent dominance, but usually with a close title rival."
+characteristics_of_leagues_since_1999_2000_six <- "Portugal exhibits the clearest signs of competitive imbalance. The largest gaps between Champions League qualification and relegation, coupled with the strongest winners' and top-four goal differences, point to a league dominated by a small group of elite clubs. This is reinforced by the highest Gini coefficient, indicating that wealth is heavily concentrated among a few clubs. Portuguese clubs also use the largest squads on average, creating a picture of a league with a pronounced divide between the top teams and the rest. This is clearly seen by the dominance of the same four clubs consistently finishing in the top 4 positions (Porto, Benfica, Sporting CP and Braga)."
+characteristics_of_leagues_extra_one <- "Overall, the pattern of European performance suggests that success is not associated with a single domestic league structure. Spain, the strongest-performing league over the period, combines strong elite clubs, high contributions from star players, and relatively high financial inequality, indicating that concentrated quality at the top can translate into European success. England and Germany also perform strongly, but do so with noticeably different characteristics. England combines strong top-club performance with the most even distribution of wealth, while Germany pairs powerful champions with relatively low financial inequality and consistent squad usage. This suggests there are multiple pathways to European success, ranging from concentrated excellence among elite clubs to broader league-wide strength."
+characteristics_of_leagues_extra_two <- "In contrast, Italy sits in the middle of the European performance rankings and is characterised by moderate values across most measures, reflecting a league that is neither highly concentrated nor exceptionally competitive on any single dimension. France and Portugal, despite representing opposite ends of several domestic characteristics, have generally achieved weaker European outcomes. France combines close title races, lower goal differences and relatively balanced finances, while Portugal is characterised by extreme dominance from a small group of wealthy clubs. The fact that both leagues underperform relative to Spain, England and Germany suggests that neither competitive balance nor competitive imbalance alone guarantees European success. Instead, the strongest-performing leagues appear to be those that combine strong elite clubs with sufficient overall league quality to sustain high levels of competition and development over time."
+
+#Performance characteristics
+characteristics_of_leagues_grouped_by_performance <- "To explore whether domestic league characteristics differ according to European success, seasons were divided into three equally sized groups based on their level of European performance: low, medium and high performance. The average value of each competitiveness characteristic was then calculated within each group to identify any broad patterns."
+characteristics_of_leagues_grouped_by_performance_one <- "The lowest-performing group is characterised by negative scores across most competitiveness measures, indicating smaller gaps between teams and a more competitively balanced domestic environment. The title race, qualification places, and overall league structure tend to be more closely contested, while top clubs record lower goal differences than the other groups. While this does not prove a causal relationship, it suggests that the most competitive leagues, at least on these measures, tend to be associated with weaker European performance."
+characteristics_of_leagues_grouped_by_performance_two <- "The highest-performing group generally sits at the opposite end of the spectrum, with positive scores for most competitiveness variables. Larger gaps throughout the league and stronger goal differences for champions and top clubs indicate a greater degree of domestic dominance. This pattern is consistent with the idea that leagues containing more dominant elite clubs may be better positioned to compete in Europe, although several other characteristics vary across groups, making it difficult to isolate any single driver of success."
+characteristics_of_leagues_grouped_by_performance_three <- "The middle-performing group presents a mixed picture. Some measures point towards greater competitive balance, while others suggest stronger elite clubs and higher levels of dominance. Compared with the highest-performing group, it tends to have smaller gaps and lower goal differences, but it is generally less competitive than the lowest-performing group. Overall, the progression from Group 1 to Group 2 hints at a relationship where less competitive domestic leagues achieve stronger European results, but the variation across individual characteristics suggests the relationship is neither straightforward nor uniform across all dimensions of competitiveness."
+
+#Modelling
+modelling_one <- "This section moves beyond the descriptive analysis to investigate the relationship between domestic league characteristics and European performance. While the exploratory analysis identified a number of potential patterns, modelling techniques are used to assess the strength and significance of these relationships more formally."
+modelling_two <- "Both supervised and unsupervised machine learning approaches are applied. Panel regression is used as a supervised method to examine how changes in league characteristics are associated with European performance and to identify which factors have the strongest statistical relationship with success. Clustering is then used as an unsupervised method to group seasons with similar characteristics, providing an alternative perspective on how different league profiles relate to European performance. Together, these approaches help determine whether consistent patterns exist between domestic competitiveness and success in European competitions."
+
+#Variable creation
+competitiveness_indices <- "Competitiveness indices"
+competitiveness_indices_one <- "The competitiveness indices were developed using football theory to capture different dimensions of domestic league competitiveness. Rather than being derived statistically, the indices combine related variables based on their expected influence on the competitive structure of a league."
+competitiveness_indices_two <- "- <b>Competitive title race</b> combines the points differences between first and second place and between first and the Champions League qualification places. These measures represent how closely contested the race for the title and top positions is, with smaller gaps indicating a more competitive battle at the top of the league."
+competitiveness_indices_three <- "- <b>Elite dominance</b> combines the points difference between Champions League qualification and relegation, the total goal difference of the top four clubs, the Gini coefficient, and the contribution of the top three goalscorers. This index captures the extent to which a small number of clubs dominate the league competitively, financially and offensively. Higher values indicate a greater concentration of success and resources among elite clubs."
+competitiveness_indices_four <- "- <b>Competitive balance</b> combines all points-difference measures, top-four goal difference, and the Gini coefficient, with the signs reversed where appropriate so that smaller gaps and lower concentrations increase the index. It is intended to measure how evenly matched clubs are across the league as a whole, rather than focusing solely on competition at the top."
+competitiveness_indices_five <- "- <b>Player rotation</b> is represented by the average number of players used by clubs during the season. This reflects squad utilisation and rotation practices, which may indicate differences in squad depth, fixture congestion, or approaches to balancing domestic and European commitments."
+competitiveness_indices_six <- "- <b>Multi-level competitiveness</b> combines the three points-difference measures across the league table. Unlike the other indices, it focuses solely on competitive outcomes and league standings, providing a direct measure of how closely contested the league is at multiple levels, including the title race, qualification positions and relegation battle."
+competitiveness_indices_seven <- "Together, these indices provide a range of theoretically motivated measures that capture different aspects of league competitiveness, allowing the analysis to distinguish between competition at the top of the table, overall competitive balance, elite club dominance and squad management strategies."
+principal_component_analysis_pca <- "Principal Component Analysis (PCA)"
+principal_component_analysis_pca_one <- "Many of the competitiveness variables are closely related and may capture similar underlying concepts. PCA was therefore used to reduce the number of variables by combining them into a smaller set of uncorrelated components that capture the main sources of variation in league competitiveness. This helps simplify the analysis and reduces potential multicollinearity issues when modelling European performance."
+the_five_principal_components_can_be_broadly_interpreted_as <- "The five principal components can be broadly interpreted as:"
+the_five_principal_components_can_be_broadly_interpreted_as_one <- "- <b>PC1: Overall dominance and inequality:</b> Primarily captures large gaps throughout the league and strong performances from top clubs. The relatively low loading for the gap between first and second place suggests that title-race competitiveness is less important than broader dominance throughout the league. Higher values indicate less competitive leagues with greater separation between elite clubs and the rest."
+the_five_principal_components_can_be_broadly_interpreted_as_two <- "- <b>PC2: Dominant champion, competitive remainder:</b> Reflects leagues where the champion is clearly ahead of its rivals, but the teams below are more evenly matched. With positive loadings for gaps between first and other positions, whilst negative loadings on the points gap between champions league places and relegation and goal difference of the top 4 clubs."
+the_five_principal_components_can_be_broadly_interpreted_as_three <- "- <b>PC3: Squad structure and player reliance:</b> Driven mainly by player usage, goalscoring concentration and wealth distribution. This component represents structural characteristics of leagues rather than competitive balance itself."
+the_five_principal_components_can_be_broadly_interpreted_as_four <- "- <b>PC4: Financial balance with mixed competitiveness:</b> Relates strongly to wealth distribution and competitive separation across different parts of the league table. It suggests leagues with relatively balanced finances but varying levels of competition, which is relatively close among top teams despite larger gaps elsewhere."
+the_five_principal_components_can_be_broadly_interpreted_as_five <- "- <b>PC5: Rotation and star-player influence:</b> Captures greater player rotation and reliance on key goalscorers, alongside a dominant league winner but relatively tighter competition across the remainder of the league. It may reflect leagues where top clubs possess sufficient depth to rotate players while still maintaining a competitive advantage."
+the_five_principal_components_can_be_broadly_interpreted_as_six <- "Together, these components provide a condensed representation of the different dimensions of league competitiveness and are used alongside the theory-driven competitiveness indices to investigate their relationship with European performance."
+
+#Regression
+regression_analysis_one <- "To examine the relationship between domestic league characteristics and European performance, panel regression was used. Panel regression is particularly suited to this dataset as it combines observations across both countries and seasons, allowing variation over time and between leagues to be analysed simultaneously. Two-way fixed effects models were applied, incorporating both country fixed effects and season fixed effects. This approach controls for unobserved characteristics that are constant within each league (such as football culture or league structure) and factors that affect all leagues in a given season (such as changes to European competition formats through the introduction of the League phase and the removal of the Away goals rule). As a result, the analysis focuses on how changes within a particular league over time relate to changes in its European performance."
+regression_analysis_two <- "Three regression specifications were estimated. The first used the original competitiveness variables, excluding variables with high multicollinearity where necessary to avoid distorted coefficient estimates. The second used the principal components generated through PCA, allowing the relationship between broader dimensions of competitiveness and European performance to be assessed. The third used the theory-driven competitiveness indices, examining whether the underlying concepts of competitiveness were more informative than the individual variables themselves."
+hidden_click_to_view_more_detailed_description_of_methodology_one <- "To account for potential violations of standard regression assumptions, clustered standard errors were used. These adjust for both heteroskedasticity (non-constant variance of errors) and autocorrelation (correlation of errors across seasons within the same country), providing more reliable estimates of statistical significance. Variable significance was therefore assessed using these robust standard errors rather than conventional regression outputs."
+hidden_click_to_view_more_detailed_description_of_methodology_two <- "Multiple regression models were estimated throughout the analysis in order to identify the independent relationship between each competitiveness measure and European performance while controlling for the effects of other variables. This provides greater confidence that observed relationships are not simply driven by omitted factors captured elsewhere in the model. For the competitiveness indices, single regressions were run as each represents a combination of variables so controls for other factors."
+hidden_click_to_view_more_detailed_description_of_methodology_three <- "Initially, full models containing all selected variables, principal components, or competitiveness indices were estimated. As a robustness check, additional regressions were then run using only statistically significant variables and those considered theoretically important. Comparing these specifications allowed both model performance and coefficient stability to be assessed. Where appropriate, results presented in this report focus on the best-performing specification, while comparisons between models are used to evaluate the robustness of the findings."
+
+#Regression on same
+on_league_characteristics_and_same_season_s_european_performance_one <- "Across all three modelling approaches, a consistent theme emerges: domestic league competitiveness appears to be negatively associated with European performance. While the explanatory power of the models is relatively modest, the strongest and most consistent relationships suggest that leagues characterised by greater dominance from their leading clubs tend to perform better in European competitions. Larger performance gaps between top clubs and the rest of the league, higher goal differences among elite teams, and greater concentrations of financial resources are generally associated with stronger European outcomes. Conversely, leagues with tighter competition throughout the table, smaller points gaps, and more evenly distributed resources tend to achieve weaker European performance. Although individual variables and indices produce some mixed findings, the overall evidence points towards elite domestic dominance being more beneficial for European success than broad competitive balance."
+on_league_characteristics_and_same_season_s_european_performance_two <- "The results suggest that European performance is strongest in leagues where the leading clubs are particularly dominant. Larger goal differences among the top four clubs and greater financial inequality are both positively associated with European success, indicating that strong elite clubs and concentrated resources may provide an advantage in continental competitions. Similarly, a larger gap between first and second place suggests that a less competitive title race is associated with improved European performance. However, the negative coefficient for the gap between Champions League qualification and relegation and the negative relationship with top-scorer goals indicate that some forms of competitiveness in the wider league may still be beneficial. Overall, the evidence points towards the strength and dominance of elite clubs being more important than maintaining competitive balance across the league as a whole."
+on_league_characteristics_and_same_season_s_european_performance_three <- "PC1, the strongest predictor, primarily captures overall league dominance and inequality, particularly through large points gaps and strong performances from top clubs. Its positive relationship with European performance reinforces the conclusion that less competitive domestic leagues tend to produce better European results. PC4, however, introduces some nuance to this finding. This component reflects greater financial equality and lower concentration among goalscorers, while also containing mixed signals regarding competitiveness across different parts of the league table. The negative coefficient for PC4 suggests that some aspects of equality and tighter competition may be associated with reduced European performance, while some tighter points gaps throughout the league may improve performance. Taken together, the PCA results broadly support the findings from the individual-variable models, although the interpretation is less direct because each component combines multiple underlying characteristics."
+on_league_characteristics_and_same_season_s_european_performance_four <- "The Competitive Balance Index is negatively related to European performance, suggesting that leagues with smaller points gaps, lower top-four goal differences and more evenly distributed wealth generally perform worse in European competition. This finding is supported by the Multi-Level Competitiveness Index, which focuses solely on points differences and also shows a significant negative relationship with performance. Similarly, a more competitive title race is associated with weaker European outcomes. In contrast, the Elite Dominance Index has a positive relationship with European performance, indicating that leagues characterised by dominant top clubs, greater financial concentration and stronger performances from elite teams tend to achieve greater success in Europe. Collectively, these results provide the strongest support for the idea that domestic dominance among elite clubs, rather than league-wide competitive balance, is associated with improved European performance."
+
+#Regression on next
+on_league_characteristics_and_next_season_s_european_performance_one <- "The lagged analysis examines whether domestic league characteristics in one season are associated with European performance in the following season. Compared with the contemporaneous models, the relationships are generally stronger, with higher R² values across all three modelling approaches. The results provide more consistent evidence that domestic dominance and competitive imbalance may create conditions that support future European success. Larger performance gaps between clubs, greater concentration of goals among leading players, and less equal distributions of wealth are repeatedly associated with stronger European outcomes in the following season. While some findings remain mixed, the lagged models provide stronger support for the argument that dominant domestic environments are linked to improved European performance than the models using European performance from the same season. An explanation for this would be that dominant sides have time to develop in lower pressured environments over multiple seasons, whilst also being able to hold onto their star players and sign others due to strong performance in their domestic leagues."
+on_league_characteristics_and_next_season_s_european_performance_two <- "These results suggest that leagues characterised by larger performance gaps throughout the table tend to achieve stronger European outcomes in the following season. A less competitive title race, greater separation between clubs, larger contributions from elite goalscorers and greater wealth inequality are all associated with improved future European performance. Compared with the current-season model, the findings are more consistent, as all significant variables point towards stronger European performance being associated with greater levels of domestic dominance and concentration."
+on_league_characteristics_and_next_season_s_european_performance_three <- "As with the contemporaneous analysis, PC1 remains the strongest predictor, reinforcing the importance of overall league dominance and separation between elite clubs and the rest of the league. PC3 suggests that greater reliance on leading goalscorers is associated with improved performance, although the component also contains mixed signals relating to player usage and wealth distribution. PC4 continues to display a negative relationship with European performance, indicating that characteristics associated with greater equality may reduce future success. Overall, the PCA results largely support the findings from the current-season analysis, but the relationships are stronger and more consistent when predicting European performance one season ahead."
+on_league_characteristics_and_next_season_s_european_performance_four <- "The Elite Dominance Index emerges as the strongest predictor of future European performance, suggesting that leagues characterised by dominant top clubs, larger gaps between elite and weaker teams, higher top-club goal differences and concentrated financial resources tend to achieve greater European success in the following season. Conversely, the Competitive Balance Index remains negatively related to performance, indicating that more evenly matched leagues tend to perform less well. These findings closely mirror those obtained for current-season European performance, but the relationships are stronger in the lagged models, providing further evidence that domestic dominance may contribute to sustained European success over time."
+
+#Regression by comp
+on_league_characteristics_and_same_season_s_european_performance_by_co_one <- "When European performance is split by competition, the relationship between domestic league characteristics appears to be driven almost entirely by clubs performance in the Champions League. No statistically significant relationships were identified for Europa League performance, suggesting that domestic competitiveness has little measurable impact on success in that competition. In contrast, the Champions League results closely mirror the findings from the overall European performance models, with slightly stronger relationships found. Across all three modelling approaches, the strongest evidence points towards leagues with dominant elite clubs, greater financial inequality and stronger top-team performances achieving better Champions League results. This suggests that the characteristics associated with success in European competition overall are largely being driven by performance at the highest level of continental football."
+on_league_characteristics_and_same_season_s_european_performance_by_co_two <- "The results reinforce several conclusions found in the main analysis. Greater financial inequality and stronger performances from the league's leading clubs, measured through top-four goal difference, are both associated with improved Champions League performance. As with the overall model, the negative relationship between the Champions League-to-relegation gap suggests that some forms of competitiveness below the elite clubs may coexist with European success. Compared with the overall European performance model, the Champions League results place even greater emphasis on the importance of strong elite clubs and concentrated resources, while the role of title-race competitiveness becomes less important."
+on_league_characteristics_and_same_season_s_european_performance_by_co_three <- "The findings are highly consistent with those obtained using overall European performance. PC1 again emerges as the strongest predictor, highlighting the importance of dominance and separation between leading clubs and the rest of the league. PC4 remains negatively associated with performance, suggesting that characteristics linked to greater equality are generally associated with weaker Champions League outcomes. PC3 introduces some uncertainty as it contains a mixture of competitiveness and structural league characteristics, although the overall pattern continues to point towards less competitive domestic environments being linked to stronger Champions League performance."
+on_league_characteristics_and_same_season_s_european_performance_by_co_four <- "These results suggest that leagues characterised by dominant top clubs, strong top-four performances and concentrated financial resources tend to achieve better Champions League outcomes. Conversely, leagues exhibiting greater competitive balance generally achieve weaker results. While the explanatory power of the models is relatively modest, the consistency of these findings with the main regression analysis suggests that the negative relationship between domestic competitiveness and European success is largely driven by performance in the Champions League rather than the Europa League."
+
+#Clustering
+clustering_to_group_similar_seasons_together_one <- "K-means clustering was used as an unsupervised machine learning technique to group together seasons with similar domestic competitiveness characteristics. Unlike panel regression, it does not test the statistical significance of individual variables. Instead, it identifies whether certain combinations of league characteristics naturally appear together across seasons."
+clustering_to_group_similar_seasons_together_two <- "After the clusters were created, the average European performance and next-season European performance were calculated for each group. The average league characteristics within each cluster were then compared to assess whether certain competitiveness profiles were associated with stronger or weaker European outcomes. This was applied using three approaches: all individual variables, the principal components, and the theory-driven competitiveness indices. The competitiveness indices produced the clearest cluster separation, although all approaches showed broadly similar trends."
+clustering_to_group_similar_seasons_together_three <- "Overall, the clustering results support the pattern found in the regression analysis. When the number of clusters was set to two, the distinction was mainly between competitive leagues, with smaller or negative values across most variables, and less competitive leagues, with larger positive values showing greater separation between clubs. The less competitive cluster consistently achieved higher European performance, both in the current season and the following season. However, the differences in average European performance between some clusters are not especially large, so these results should be interpreted as descriptive patterns rather than statistically significant evidence. The clearest exception is that the least competitive cluster is consistently associated with higher European performance, particularly in the following season."
+radar_plots_for_clusters_one <- "This group performs best overall, particularly in the following European season. It is characterised by large gaps between first and second place and between first and the Champions League places, suggesting that the leading club is often clearly ahead domestically. However, the gap between Champions League qualification and relegation is relatively small, and the Gini coefficient is low. This suggests a league where the title race may be less competitive, but there is still some competitiveness across the rest of the league. This profile appears to be favourable for European performance, potentially because elite clubs are strong enough to compete in Europe while still operating in a reasonably competitive domestic environment. The clearest example is England in 2004/05, when Chelsea were runaway league winners and set a record for the fewest goals conceded in a Premier League season. Despite Chelsea's dominance, wealth distribution and the points gap between the top four and the rest of the league remained relatively balanced. English clubs also performed strongly in Europe, with Liverpool winning the Champions League and Arsenal reaching the final the following season."
+radar_plots_for_clusters_two <- "This group also performs well in the current European season, although its next-season performance is weaker than Cluster 1. This group shows signs of stronger elite dominance, with high top-four goal difference, high goals from top scorers, a larger gap between Champions League qualification and relegation, and the highest Gini coefficient. However, the gap between the top two sides is relatively small, reflecting a tight title battle. This suggests a league where a small number of wealthy and high-performing clubs are clearly separated from the rest. The strong current European performance supports the idea that dominant elite clubs can drive continental success, although the weaker performance in the next season suggests this advantage may be less consistent over time. Spain in 2015/16 typifies this cluster. Despite a title race in which the top three teams finished within three points of one another, Barcelona and Real Madrid recorded exceptional goal differences (83 and 76 respectively), while Luis Suárez, Cristiano Ronaldo and Lionel Messi scored a combined 101 league goals. European performance was equally impressive, with Real Madrid and Atlético Madrid contesting the Champions League final, while Sevilla won the Europa League and Villarreal reached the semi-finals."
+radar_plots_for_clusters_three <- "This group performs worst in both current and next-season European performance. It has negative values across most characteristics, including smaller points gaps, lower winners' and top-four goal differences, fewer players used, fewer goals from top scorers, and lower wealth inequality. This suggests the most competitively balanced league profile, where clubs are more evenly matched throughout the table. While this may indicate a healthier domestic competitive structure, it appears to be associated with weaker European performance, supporting the wider finding that highly competitive leagues may not necessarily produce the strongest European outcomes. A clear example is the Premier League in 2015/16, when Leicester City won the league for the first time despite beginning the season as outsiders, with the lowest points tally of a winner since 2010/11. The season was characterised by relatively small gaps between clubs and a lack of sustained dominance from the traditional elite teams. English clubs subsequently performed relatively poorly in European competitions both during that season and in the following campaign, making it a strong illustration of this cluster's characteristics."
+radar_plots_for_clusters_overall <- "Taken together, the clustering suggests that the strongest European performance is generally found in leagues that are not fully competitive throughout, but also not necessarily dominated in every area. The best-performing profiles tend to combine some form of elite strength with either a closer gap between the Champions League places and the rest of the league, or a more competitive title race. In contrast, the most balanced leagues, where points gaps, goal differences and wealth inequality are all relatively low, tend to perform worst in Europe. This supports the broader conclusion that European success is more closely associated with the strength and dominance of leading clubs than with domestic competitive balance across the league as a whole."
+
+#Conclusion
+overall_conclusion_one <- "The results suggest that domestic competitiveness and European performance are related, although the relationship is neither simple nor driven by a single characteristic. Across the exploratory analysis, the strongest-performing leagues generally exhibited larger gaps between clubs, stronger performances among the top teams, or greater concentrations of financial resources than lower-performing leagues. At the same time, the leading European leagues displayed a variety of domestic structures, indicating that there is no single model for success."
+overall_conclusion_two <- "The regression analysis found a broadly consistent relationship between European performance and measures associated with lower competitiveness, including larger points gaps, greater top-club goal differences and higher levels of wealth concentration. Conversely, variables and indices representing competitive balance, such as smaller gaps throughout the league table and more equal resource distribution, were often associated with lower European performance. These patterns were evident across the original variables, principal components and theory-driven competitiveness indices, with the latter providing the most consistent results."
+overall_conclusion_three <- "An even stronger relationship was found when analysing the relationship between league characteristics and performance in the following European season. This suggests that domestic dominance may not only coincide with European success, but may also help create conditions for future performance. Leagues where top clubs are already dominant domestically may allow those clubs to build confidence, manage squads more effectively, attract and retain stronger players, and sustain performance across multiple competitions."
+overall_conclusion_four <- "Analysis by competition further indicated that this relationship is primarily driven by the Champions League. While significant relationships were identified for Champions League performance, no significant relationships were found for Europa League performance. This suggests that domestic league structure has a stronger connection with performance at the highest level of European club football than in secondary competitions."
+overall_conclusion_five <- "The clustering analysis reinforced these findings by grouping seasons with similar competitiveness characteristics. The most competitively balanced cluster consistently produced the weakest European performance in both the current and subsequent season. In contrast, clusters characterised by greater separation between clubs generally achieved stronger European outcomes. However, the best-performing clusters were not necessarily those displaying the greatest lack of competitiveness in every dimension. Rather, they tended to combine some degree of domestic dominance with competitiveness in specific areas of the league, such as tighter competition among elite clubs or a smaller gap between European qualification places and the rest of the table."
+overall_conclusion_six <- "Taken together, the findings suggest that the domestic characteristics most closely associated with European success are generally those reflecting lower levels of competitiveness across the league as a whole. Leagues characterised by larger performance gaps, stronger top-club dominance and greater financial concentration tended to achieve higher levels of European performance than leagues where clubs were more evenly matched. However, the relatively modest explanatory power of the models indicates that competitiveness alone explains only part of European success, and that many other sporting, financial and organisational factors are also likely to play an important role."
+overall_conclusion_seven <- "Overall, the analysis provides evidence of a consistent association between domestic league structure and European performance. Across multiple analytical approaches, the results suggest that seasons characterised by lower overall competitiveness tend to be associated with stronger European outcomes, particularly in the Champions League. While this relationship should not be interpreted as causal, it represents a recurring pattern throughout the period analysed and offers insight into how different competitive environments may relate to success in European football."
+
+#Limitations
+limitations_one <- "Some limitations should be considered when interpreting the findings of this analysis."
+limitations_two <- "Firstly, the measure of European performance is based on a constructed scoring system and therefore reflects one possible interpretation of success in European competitions. While it aims to capture performance consistently across clubs, countries and seasons, different weighting schemes or performance metrics may produce slightly different results."
+limitations_three <- "Similarly, league competitiveness is not directly observable and must be defined through measurable characteristics. The variables, indices and principal components used in this report represent theoretically informed proxies for competitiveness, but alternative definitions could place greater emphasis on different aspects of competition, such as attendance, uncertainty of outcomes, or competitive balance measured through other statistical methods."
+limitations_four <- "The analysis is also constrained by a relatively small sample size, covering only six leagues and seasons from 1999/00 onwards. While this period captures the modern era of European football and countries with meaningful consistent European performance, it excludes earlier seasons and limits the number of observations available for statistical analysis."
+limitations_five <- "Another limitation is that results may be influenced by exceptional teams or exceptional seasons. A single dominant club can substantially improve a country's European performance, potentially making a league appear more successful than its overall competitiveness would suggest. Similarly, a team that secures domestic success early may rotate players later in the season, reducing its final points total and making the league appear more competitive than it was in practice."
+limitations_six <- "Finally, the analysis excludes the UEFA Europa Conference League. The competition was introduced part way through the study period, making consistent comparisons across all seasons difficult. In addition, the countries included in the dataset typically had only one representative in the competition each season, limiting its contribution to overall European performance and reducing the value of including it within the analysis."
+limitations_seven <- "These limitations mean that the results should be interpreted as evidence of associations between domestic league characteristics and European performance rather than definitive causal relationships."
+
+#Process
+data_sources_and_how_data_was_collected_and_analysis_was_done_one <- "The majority of the data was collected from Transfermarkt, a football database containing information on players, clubs and competitions. Where data was unavailable on Transfermarkt, supplementary information was obtained from Wikipedia. Data was collected using Python web-scraping techniques and subsequently cleaned and transformed into datasets for each variable. Additional calculations were performed to derive the European Performance metric and the Gini Coefficient."
+data_sources_and_how_data_was_collected_and_analysis_was_done_two <- "For each season between 1999/00 and 2025/26, each country was assigned a European Performance score. This measure is calculated by summing points awarded for the progress of clubs from that country in the UEFA Champions League and UEFA Europa League:"
+data_sources_and_how_data_was_collected_and_analysis_was_done_three <- "- 20 points: Champions League winners"
+data_sources_and_how_data_was_collected_and_analysis_was_done_four <- "- 16 points: Champions League runners-up"
+data_sources_and_how_data_was_collected_and_analysis_was_done_five <- "- 12 points: Champions League semi-finalists"
+data_sources_and_how_data_was_collected_and_analysis_was_done_six <- "- 10 points: Europa League winners"
+data_sources_and_how_data_was_collected_and_analysis_was_done_seven <- "- 8 points: Champions League quarter-finalists"
+data_sources_and_how_data_was_collected_and_analysis_was_done_eight <- "- 8 points: Europa League runners-up"
+data_sources_and_how_data_was_collected_and_analysis_was_done_nine <- "- 6 points: Europa League semi-finalists"
+data_sources_and_how_data_was_collected_and_analysis_was_done_ten <- "- 4 points: Europa League quarter-finalists"
+data_sources_and_how_data_was_collected_and_analysis_was_done_eleven <- "Subsequent analysis was conducted in R, while the interactive visualisations were developed using R Shiny, CSS and JavaScript."

ui.R

@@ -0,0 +1,693 @@
+ui <- fluidPage(
+
+  title="Football league competitiveness and countries European performance",lang="en",
+
+  theme="css.css",
+  useShinyjs(),
+  tags$head(
+    tags$script(HTML("
+      Shiny.addCustomMessageHandler('bindModalHidden', function(message) {
+        $('#shiny-modal').off('hidden.bs.modal.modalListener');
+        $('#shiny-modal').on('hidden.bs.modal.modalListener', function () {
+          Shiny.setInputValue('modal_closed_nonce', Math.random(), {priority: 'event'});
+        });
+      });
+    "),
+    "function sendWidth() {
+      Shiny.setInputValue('screen_width', window.innerWidth);
+    }
+    $(document).on('shiny:connected', sendWidth);
+    $(window).on('resize', sendWidth);"),
+    tags$style(
+      # the create_scrolly_side will return a list, the first object is the CSS classes that create the stick position scrolling behaviour.
+      # the second object is the javascript that links the scrolling to change CSS classes and update shiny inputs (loaded at end of page)
+      
+      HTML(as.character(create_scrolly_side("time_plot")[1])),
+      HTML(as.character(create_scrolly_side("league_averages_plot")[1])),
+      HTML(as.character(create_scrolly_side("basic_profiling_plot")[1])),
+      HTML(as.character(create_scrolly_side("regression_current_year_plot")[1])),
+      HTML(as.character(create_scrolly_side("regression_next_year_plot")[1])),
+      HTML(as.character(create_scrolly_side("regression_competition_plot")[1])),
+      HTML(as.character(create_scrolly_side("clustering_plot")[1]))
+    ),
+    
+    HTML(
+      "<meta name=\"viewport\" content=\"width=device-width,initial-scale=1\">
+<meta property=\"og:title\" content=\"Football league competitiveness and countries European performance\">
+<meta property=\"og:description\" content=\"Football league competitiveness and countries European performance\">"
+      )
+  ),
+  
+  #### header ####
+  div(class = "header",
+      tags$div(
+        style = "display: flex; flex-direction: column;",
+        tags$a(`aria-label` = "Jump back to contents", href = "#contents", class = "jump-link", "Jump back to contents")
+      )
+  ),
+  
+  #### title and contents ####
+  
+  div(id = "contents", class = "toc",
+      tags$h1(title),
+      tags$a(href = "https://youtu.be/_mnsaDJzpso?si=3EnfpdRqXp4afBiI", tags$strong("Click here for a link to a static version of the report")),
+      p(style = 'font-weight: bold; text-align: right; font-size: 1.2em;', "Scroll to begin  ▼"),
+      tags$h3("Table of Contents"),
+      tags$ul(
+        tags$li(tags$a(href = "#motivation", tags$strong("Motivation"))),
+        tags$li(tags$a(href = "#exploratory_analysis", tags$strong("Exploratory analysis"))),
+        tags$ul(
+          tags$li(tags$a(href = "#time_plot", "Europe's top 6 leagues European performance since 1999-2000")),
+          tags$li(tags$a(href = "#variable_selection", "Variables used in analysis of leagues competitiveness")),
+          tags$li(tags$a(href = "#league_averages_plot", "Characteristics of leagues since 1999-2000")),
+          tags$li(tags$a(href = "#basic_profiling", "Characteristics of leagues grouped by performance"))
+          ),
+        tags$li(tags$a(href = "#modelling", tags$strong("Modelling"))),
+        tags$ul(
+                tags$li(tags$a(href = "#variable_creation", tags$strong("Variable creation"))),
+                tags$li(tags$a(href = "#regression", tags$strong("Regression analysis"))),
+                tags$ul(
+                  tags$li(tags$a(href = "#regression_current_year", "On league characteristics and same season's European performance")),
+                  tags$li(tags$a(href = "#regression_next_year", "On league characteristics and next season's European performance")),
+                  tags$li(tags$a(href = "#regression_competition", "On league characteristics and same season's European performance by competition"))
+                ),
+                tags$li(tags$a(href = "#clustering", tags$strong("Clustering to group similar seasons together")))
+        ),
+        tags$li(tags$a(href = "#conclusion", tags$strong("Conclusion")))
+        )
+  ),
+  
+  br(),
+  br(),
+  
+  #### motivation ####
+  
+  tags$header(
+    id = "motivation", 
+    class = "colourheader",
+    tags$div(
+      class = "paddeddiv",
+      style = "position: relative;",
+      tags$h2(paste0("Motivation"), style = "font-weight: bold; text-align: center;"),
+      p(style = "margin-top: 5px; white-space: normal;", HTML(motivation_one)),
+      p(style = "margin-top: 5px;", paste0(motivation_two)),
+      p(style = "margin-top: 5px;", paste0(motivation_three)),
+      p(style = "margin-top: 20px;", paste0(date))
+    )
+  ),
+  
+  br(),
+  br(),
+  
+  div(
+    class = "bigfixedimage",
+    style =  paste0("background-image: url('psg.jpg');min-height: 100vh;")
+  ),
+  
+  br(),
+  br(),
+  
+  #### Exploratory analysis ####
+  tags$header(id = "exploratory_analysis", class="colourheader",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Exploratory analysis"), style = "font-weight: bold; text-align: center;"),
+                       p(style = "margin-top: 5px;", paste0(exploratory_analysis_one))
+              )
+  ),
+  
+  br(),
+  br(),
+ 
+  #### European performance over time ####
+  
+  tags$div(ID = "time_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "European performance over time, from 1999-2000 to 2025-26"),
+                           plotlyOutput("european_performance_over_time_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="time_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Spanish teams European performance "),
+                       br(),
+                       europe_s_top_6_leagues_european_performance_since_1999_2000_one)),
+                 
+                 div(class="time_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" English teams European performance "),
+                       br(),
+                       europe_s_top_6_leagues_european_performance_since_1999_2000_two)),
+                 
+                 div(class="time_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" German teams European performance "),
+                       br(),
+                       europe_s_top_6_leagues_european_performance_since_1999_2000_three)),
+                 
+                 div(class="time_plot text-article", `data-width`="1",`data-index`="3",
+                     p(strong(" Italian teams European performance "),
+                       br(),
+                       europe_s_top_6_leagues_european_performance_since_1999_2000_four)),
+                 
+                 div(class="time_plot text-article", `data-width`="1",`data-index`="4",
+                     p(strong(" French teams European performance "),
+                       br(),
+                       europe_s_top_6_leagues_european_performance_since_1999_2000_five)),
+                 
+                 div(class="time_plot text-article", `data-width`="1",`data-index`="5",
+                     p(strong(" Portuguese teams European performance "),
+                       br(),
+                       europe_s_top_6_leagues_european_performance_since_1999_2000_six))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  #### Variable selection ####
+  tags$header(id = "variable_selection", class="colourheader3",
+              tags$div(
+                class = "paddeddiv",
+                style = "position: relative;",
+                tags$h3(
+                  "Variables used in analysis of leagues competitiveness",
+                  style = "font-weight: bold; text-align: center;"
+                ),
+                lapply(
+                  c(
+                    variables_used_in_analysis_of_leagues_competitiveness_one,
+                    variables_used_in_analysis_of_leagues_competitiveness_two,
+                    variables_used_in_analysis_of_leagues_competitiveness_three,
+                    variables_used_in_analysis_of_leagues_competitiveness_four,
+                    variables_used_in_analysis_of_leagues_competitiveness_five,
+                    variables_used_in_analysis_of_leagues_competitiveness_six,
+                    variables_used_in_analysis_of_leagues_competitiveness_seven,
+                    variables_used_in_analysis_of_leagues_competitiveness_eight,
+                    variables_used_in_analysis_of_leagues_competitiveness_nine,
+                    variables_used_in_analysis_of_leagues_competitiveness_ten,
+                    variables_used_in_analysis_of_leagues_competitiveness_eleven,
+                    variables_used_in_analysis_of_leagues_competitiveness_twelve,
+                    variables_used_in_analysis_of_leagues_competitiveness_thirteen,
+                    variables_used_in_analysis_of_leagues_competitiveness_fourteen,
+                    variables_used_in_analysis_of_leagues_competitiveness_fifteen,
+                    variables_used_in_analysis_of_leagues_competitiveness_sixteen
+                  ),
+                  function(x) p(style = "margin-top: 5px;", HTML(x))
+                )
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### League average characteristics ####
+  
+  tags$div(ID = "league_averages_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "Average characteristics of leagues, from 1999-2000 to 2025-26"),
+                           plotlyOutput("league_characteristics_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="league_averages_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Spanish teams European performance "),
+                       br(),
+                       characteristics_of_leagues_since_1999_2000_one)),
+                 
+                 div(class="league_averages_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" English teams European performance "),
+                       br(),
+                       characteristics_of_leagues_since_1999_2000_two)),
+                 
+                 div(class="league_averages_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" German teams European performance "),
+                       br(),
+                       characteristics_of_leagues_since_1999_2000_three)),
+                 
+                 div(class="league_averages_plot text-article", `data-width`="1",`data-index`="3",
+                     p(strong(" Italian teams European performance "),
+                       br(),
+                       characteristics_of_leagues_since_1999_2000_four)),
+                 
+                 div(class="league_averages_plot text-article", `data-width`="1",`data-index`="4",
+                     p(strong(" French teams European performance "),
+                       br(),
+                       characteristics_of_leagues_since_1999_2000_five)),
+                 
+                 div(class="league_averages_plot text-article", `data-width`="1",`data-index`="5",
+                     p(strong(" Portuguese teams European performance "),
+                       br(),
+                       characteristics_of_leagues_since_1999_2000_six))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  tags$header(class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       p(style="margin-top: 5px;",paste0(characteristics_of_leagues_extra_one)),
+                       p(style="margin-top: 5px;",paste0(characteristics_of_leagues_extra_two))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### Basic profiles characteristics ####
+  tags$header(id = "basic_profiling", class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       p(style="margin-top: 5px;",paste0(characteristics_of_leagues_grouped_by_performance))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  tags$div(ID = "basic_profiling_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "Average characteristics of leagues, grouped by performance"),
+                           plotlyOutput("basic_profiles_characteristic_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="basic_profiling_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Characteristics of worst performing European seasons "),
+                       br(),
+                       characteristics_of_leagues_grouped_by_performance_one)),
+                 
+                 div(class="basic_profiling_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" Characteristics of best performing European seasons "),
+                       br(),
+                       characteristics_of_leagues_grouped_by_performance_two)),
+                 
+                 div(class="basic_profiling_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" Characteristics of middle performing European seasons "),
+                       br(),
+                       characteristics_of_leagues_grouped_by_performance_three))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  div(
+    class = "bigfixedimage",
+    style =  paste0("background-image: url('aston-villa.jpg');min-height: 100vh;")
+  ),
+  
+  br(),
+  br(),
+  
+  #### Modelling ####
+  tags$header(id = "modelling", class="colourheader",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Modelling"), style = "font-weight: bold; text-align: center;"),
+                       p(style="margin-top: 5px;",paste0(modelling_one)),
+                       p(style="margin-top: 5px;",paste0(modelling_two))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### Variable creation ####
+  tags$header(id = "variable_creation", class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Variable creation"), style = "font-weight: bold; text-align: center;"),
+                       tags$h3(paste0(competitiveness_indices), style = "font-weight: bold;"),
+                       lapply(
+                         c(
+                           competitiveness_indices_one,
+                           competitiveness_indices_two,
+                           competitiveness_indices_three,
+                           competitiveness_indices_four,
+                           competitiveness_indices_five,
+                           competitiveness_indices_six,
+                           competitiveness_indices_seven
+                         ),
+                         function(x) p(style = "margin-top: 5px;", HTML(x))
+                       ),
+                       br(),
+                       tags$h3(paste0(principal_component_analysis_pca), style = "font-weight: bold;"),
+                       lapply(
+                         c(
+                           principal_component_analysis_pca_one,
+                           the_five_principal_components_can_be_broadly_interpreted_as,
+                           the_five_principal_components_can_be_broadly_interpreted_as_one,
+                           the_five_principal_components_can_be_broadly_interpreted_as_two,
+                           the_five_principal_components_can_be_broadly_interpreted_as_three,
+                           the_five_principal_components_can_be_broadly_interpreted_as_four,
+                           the_five_principal_components_can_be_broadly_interpreted_as_five,
+                           the_five_principal_components_can_be_broadly_interpreted_as_six
+                         ),
+                         function(x) p(style = "margin-top: 5px;", HTML(x))
+                       )
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### Regression ####
+  tags$header(id = "regression", class="colourheader2",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Regression analysis"), style = "font-weight: bold; text-align: center;"),
+                       p(style="margin-top: 5px;",paste0(regression_analysis_one)),
+                       p(style="margin-top: 5px;",paste0(regression_analysis_two)),
+                       tags$button(
+                         class = "btn btn-light mt-3 detail-toggle-btn-black",
+                         type = "button",
+                         "data-toggle-target" = "#regression-moreContent",
+                         "data-show-text" = "Click to view more detailed description of methodology ▼",
+                         "data-hide-text" = "Show less ▲",
+                         "Click to view more detailed description of methodology ▼"
+                       ),
+                       tags$div(
+                         id = "regression-moreContent",
+                         class = "collapse mt-3",
+                         p(style = "margin-top: 5px;", paste0(hidden_click_to_view_more_detailed_description_of_methodology_one)),
+                         p(style = "margin-top: 5px;", paste0(hidden_click_to_view_more_detailed_description_of_methodology_two)),
+                         p(style = "margin-top: 5px;", paste0(hidden_click_to_view_more_detailed_description_of_methodology_three))
+                       ),
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### Regression in one season ####
+  tags$header(id = "regression_current_year", class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Regression analysis on league characteristics and same season's European performance"), style = "font-weight: bold; text-align: center;"),
+                       p(style = "margin-top: 5px;", paste0(on_league_characteristics_and_same_season_s_european_performance_one))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  tags$div(ID = "regression_current_year_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "Coefficient estimates for regression"),
+                           plotlyOutput("regression_coefficients_current_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="regression_current_year_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Using all variables (R2 = 0.069) "),
+                       br(),
+                       on_league_characteristics_and_same_season_s_european_performance_two)),
+                 
+                 div(class="regression_current_year_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" Using principle components (R2 = 0.049) "),
+                       br(),
+                       on_league_characteristics_and_same_season_s_european_performance_three)),
+                 
+                 div(class="regression_current_year_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" Using competitive indices (R2 = 0.040) "),
+                       br(),
+                       on_league_characteristics_and_same_season_s_european_performance_four))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  #### Regression in next season ####
+  tags$header(id = "regression_next_year", class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Regression analysis on league characteristics and next season's European performance"), style = "font-weight: bold; text-align: center;"),
+                       p(style = "margin-top: 5px;", paste0(on_league_characteristics_and_next_season_s_european_performance_one))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  tags$div(ID = "regression_next_year_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "Coefficient estimates for regression"),
+                           plotlyOutput("regression_coefficients_next_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="regression_next_year_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Using all variables (R2 = 0.128) "),
+                       br(),
+                       on_league_characteristics_and_next_season_s_european_performance_two)),
+                 
+                 div(class="regression_next_year_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" Using principle components (R2 = 0.112) "),
+                       br(),
+                       on_league_characteristics_and_next_season_s_european_performance_three)),
+                 
+                 div(class="regression_next_year_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" Using competitive indices (R2 = 0.090) "),
+                       br(),
+                       on_league_characteristics_and_next_season_s_european_performance_four))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  #### Regression by competition ####
+  tags$header(id = "regression_competition", class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Regression analysis on league characteristics and same season's European performance by competition"), style = "font-weight: bold; text-align: center;"),
+                       p(style = "margin-top: 5px;", paste0(on_league_characteristics_and_same_season_s_european_performance_by_co_one))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  tags$div(ID = "regression_competition_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "Coefficient estimates for regression"),
+                           plotlyOutput("regression_coefficients_competition_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="regression_competition_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Using all variables (R2 = 0.099) "),
+                       br(),
+                       on_league_characteristics_and_same_season_s_european_performance_by_co_two)),
+                 
+                 div(class="regression_competition_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" Using principle components (R2 = 0.049) "),
+                       br(),
+                       on_league_characteristics_and_same_season_s_european_performance_by_co_three)),
+                 
+                 div(class="regression_competition_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" Using competitive indices (R2 = 0.042) "),
+                       br(),
+                       on_league_characteristics_and_same_season_s_european_performance_by_co_four))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  #### Clustering ####
+  tags$header(id = "clustering", class="colourheader2",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Clustering to group similar seasons together"), style = "font-weight: bold; text-align: center;"),
+                       p(style = "margin-top: 5px;", paste0(clustering_to_group_similar_seasons_together_one)),
+                       p(style = "margin-top: 5px;", paste0(clustering_to_group_similar_seasons_together_two)),
+                       p(style = "margin-top: 5px;", paste0(clustering_to_group_similar_seasons_together_three))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  tags$div(ID = "clustering_plot",
+           div(class="scrolly",
+               tags$figure(class="sticky",style=("order: 0; display: flex; flex-direction: column;"),
+                           tags$h4(style="text-align: center;", "Average characteristics of clusters"),
+                           plotlyOutput("clustering_characteristics_plot", height = "100%")
+               ),
+               
+               tags$article(
+                 div(class="clustering_plot text-article", `data-width`="1",`data-index`="0",
+                     p(strong(" Elite competition with wider league separation "),
+                       br(),
+                       radar_plots_for_clusters_one)),
+                 
+                 div(class="clustering_plot text-article", `data-width`="1",`data-index`="1",
+                     p(strong(" Elite dominance across the league "),
+                       br(),
+                       radar_plots_for_clusters_two)),
+                 
+                 div(class="clustering_plot text-article", `data-width`="1",`data-index`="2",
+                     p(strong(" Most competitive league profile "),
+                       br(),
+                       radar_plots_for_clusters_three))
+               ))
+  ),
+  
+  br(),
+  br(),
+  
+  tags$header(class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       p(style = "margin-top: 5px;", paste0(radar_plots_for_clusters_overall))
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  div(
+    class = "bigfixedimage",
+    style =  paste0("background-image: url('cpfc.jpg');min-height: 100vh;")
+  ),
+  
+  #### Conclusion ####
+  tags$header(id = "conclusion", class="colourheader",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Conclusion"), style = "font-weight: bold; text-align: center;"),
+                       lapply(
+                         c(
+                           overall_conclusion_one,
+                           overall_conclusion_two,
+                           overall_conclusion_three,
+                           overall_conclusion_four,
+                           overall_conclusion_five,
+                           overall_conclusion_six,
+                           overall_conclusion_seven
+                         ),
+                         function(x) {
+                           p(
+                             style = "margin-top: 5px;",
+                             HTML(x)
+                           )
+                         }
+                       )
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### Limitations ####
+  tags$header(class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Limitations"), style = "font-weight: bold; text-align: center;"),
+                       lapply(
+                         c(
+                           limitations_one,
+                           limitations_two,
+                           limitations_three,
+                           limitations_four,
+                           limitations_five,
+                           limitations_six,
+                           limitations_seven
+                         ),
+                         function(x) {
+                           p(
+                             style = "margin-top: 5px;",
+                             HTML(x)
+                           )
+                         }
+                       )
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### Footnote ####
+  tags$header(class="colourheader3",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h2(paste0("Data collection, sources and analysis process"), style = "font-weight: bold; text-align: center;"),
+                       lapply(
+                         c(
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_one,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_eleven,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_two,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_three,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_four,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_five,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_six,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_seven,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_eight,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_nine,
+                           data_sources_and_how_data_was_collected_and_analysis_was_done_ten
+                         ),
+                         function(x) {
+                           p(
+                             style = "margin-top: 5px;",
+                             HTML(x)
+                           )
+                         }
+                       )
+              )
+  ),
+  
+  br(),
+  br(),
+  
+  #### ending section ####
+  
+  tags$header(class="colourheader",
+              style="background:#010056 ; color:white;",
+              tags$div(class="paddeddiv",style="position: relative;",
+                       tags$h1(paste0(title)),
+                       tags$h3(name),
+                       tags$h3(date))
+              ),
+  
+  
+  #### expandable/collapsable text function ####
+  
+  tags$script(HTML("
+    document.addEventListener('DOMContentLoaded', function() {
+      document.querySelectorAll('[data-toggle-target]').forEach(function(button) {
+        const targetSelector = button.getAttribute('data-toggle-target');
+        const target = document.querySelector(targetSelector);
+        if (!target) return;
+  
+        const showText = (button.getAttribute('data-show-text') || button.textContent || '').trim();
+        const hideText = (button.getAttribute('data-hide-text') || 'Show less ▲').trim();
+  
+        const isVisible = function(el) {
+          return window.getComputedStyle(el).display !== 'none' && !el.hidden;
+        };
+        button.textContent = isVisible(target) ? hideText : showText;
+  
+        button.addEventListener('click', function() {
+          if (isVisible(target)) {
+            target.classList.remove('show');
+            target.style.display = 'none';
+            button.textContent = showText;
+          } else {
+            target.classList.add('show');
+            target.style.display = 'block';
+            button.textContent = hideText;
+          }
+        });
+      });
+    });
+  ")),
+  
+  #### scrolly javascript ####
+  
+  HTML("<script src='https://unpkg.com/d3@5.4.0/dist/d3.min.js'></script>"),
+  HTML("<script src='https://unpkg.com/enter-view@1.0.0/enter-view.min.js'></script>"),
+  HTML("<script src='https://unpkg.com/stickyfilljs@2.0.5/dist/stickyfill.js'></script>"),
+  
+  HTML(as.character(create_scrolly_side("time_plot")[2])),
+  HTML(as.character(create_scrolly_side("league_averages_plot")[2])),
+  HTML(as.character(create_scrolly_side("basic_profiling_plot")[2])),
+  HTML(as.character(create_scrolly_side("regression_current_year_plot")[2])),
+  HTML(as.character(create_scrolly_side("regression_next_year_plot")[2])),
+  HTML(as.character(create_scrolly_side("regression_competition_plot")[2])),
+  HTML(as.character(create_scrolly_side("clustering_plot")[2]))
+)

www/css.css

@@ -0,0 +1,553 @@
+html {scroll-behavior: smooth;}
+*:focus-visible{
+outline: 2px solid crimson !important; 
+scroll-margin-top: 85vh !important;
+}
+
+  .hidden {
+visibility: hidden;
+
+
+}
+
+  .hidden2 {
+visibility: hidden;
+
+
+}
+
+.displayed{
+  width: 100%;
+  justify-content: center;
+  height:100vh;
+
+
+  
+}
+
+
+
+                .switch {
+  position: relative;
+  display: inline-block;
+  width: 60px;
+  height: 34px;
+}
+
+.switch input { 
+  opacity: 0;
+  width: 0;
+  height: 0;
+}
+
+.slider {
+  position: absolute;
+  cursor: pointer;
+  top: 0;
+  left: 0;
+  right: 0;
+  bottom: 0;
+  background-color: #ccc;
+  -webkit-transition: .4s;
+  transition: .4s;
+}
+
+.slider:before {
+  position: absolute;
+  content: "";
+  height: 26px;
+  width: 26px;
+  left: 4px;
+  bottom: 4px;
+  background-color: white;
+  -webkit-transition: .4s;
+  transition: .4s;
+}
+
+input:checked + .slider {
+  background-color: #010056;
+}
+
+input:focus + .slider {
+  box-shadow: 0 0 1px #010056;
+}
+
+input:checked + .slider:before {
+  -webkit-transform: translateX(26px);
+  -ms-transform: translateX(26px);
+  transform: translateX(26px);
+}
+
+/* Rounded sliders */
+.slider.round {
+  border-radius: 34px;
+}
+
+.slider.round:before {
+  border-radius: 50%;
+}
+
+/* Table of contents */
+.colourheader, .toc {
+  background: #010056;
+  color: white;
+  padding: 6vh;
+  border-radius: 10px;
+  font-family: Arial, Helvetica, sans-serif;
+}
+
+.toc ul {
+  list-style-type: disc;
+  padding-left: 25px;
+  margin-top: 5px;
+}
+
+.toc li {
+  margin-bottom: 6px;
+  font-weight: normal;
+}
+
+.toc a {
+  color: white;
+  text-decoration: none;
+  font-weight: normal;
+}
+
+.toc a:hover {
+  text-decoration: underline;
+}
+
+.toc h1 {
+  margin-top: 25px;
+  margin-bottom: 25px;
+  font-size: 2.2em;
+  font-weight: bold;
+  text-align: center;
+  padding-bottom: 10px;
+  letter-spacing: 1px;
+}
+
+.toc h3 {
+  display: inline;
+  margin: 0;
+  font-size: 1.2em;
+  font-weight: bold;
+  color: white;
+}
+
+/* Expandable text button */
+.detail-toggle-btn {
+  display: inline-block !important;
+
+  /* allow wrapping */
+  max-width: 100% !important;
+  white-space: normal !important;
+  overflow-wrap: break-word;   /* ensures long text wraps */
+
+  /* reset sizing constraints that prevent wrapping */
+  width: auto;
+  flex: 0 1 auto !important;
+
+  box-sizing: border-box;
+
+  /* styling */
+  color: white;
+  border: 1px solid white;
+  background-color: transparent;
+  font-weight: 500;
+  font-family: Arial, Helvetica, sans-serif;
+  font-size: 1em;
+  text-align: left;
+  padding: 0.25rem 0.5rem;
+  cursor: pointer;
+}
+
+.detail-toggle-btn:hover,
+.detail-toggle-btn:focus,
+.detail-toggle-btn:active {
+  color: white !important;
+  border-color: white !important;
+  background-color: transparent !important;
+  outline: none;
+  box-shadow: none;
+}
+
+
+.detail-toggle-btn-black {
+  color: black !important;
+  border-color: black !important;
+  background-color: transparent !important;
+}
+.detail-toggle-btn-black:hover,
+.detail-toggle-btn-black:focus,
+.detail-toggle-btn-black:active {
+  color: black !important;
+  border-color: black !important;
+  background-color: transparent !important;
+}
+
+
+
+
+                  /* GLOBALS */
+
+                    html, body {
+                        position: relative;
+                        width: 100%;
+                        height: 100%;
+                      }
+
+
+                @media screen and (min-width: 768px){
+
+
+                    body {
+                      color: #222;
+                      background-color: #fff;
+                      margin: 0;
+                      padding: 0;
+                      -webkit-box-sizing: border-box;
+                      box-sizing: border-box;
+                      font-family: Arial, Helvetica, sans-serif;
+                      font-size: 18px;
+                      line-height: 1.5;
+                      word-wrap: break-word;
+                    }
+                    
+                }
+                
+                  @media screen and (max-width: 767px){
+
+
+                    body {
+                      color: #222;
+                      background-color: #fff;
+                      margin: 0;
+                      padding: 0;
+                      -webkit-box-sizing: border-box;
+                      box-sizing: border-box;
+                      font-family: Arial, Helvetica, sans-serif;
+                      font-size: 15px;
+                      line-height: 1.5;
+                      word-wrap: break-word;
+                    }
+                    
+                }
+                  
+                  header, section, nav, footer, figure, caption { /* This places text in center of various elements */
+                  display: -webkit-box;
+                  display: -ms-flexbox;
+                  display: flex;
+                  -webkit-box-pack: center;
+                  -ms-flex-pack: center;
+                  justify-content: center;
+                  background-position: center;
+                  background-repeat: no-repeat;
+                  background-size: cover;
+                  margin: 0;
+                  padding: 0;
+                }
+
+                footer {
+                  margin: 60px 0 0 0;
+                }
+                
+                /* These are the default heading styles */
+                @media screen and (max-width: 767px){
+                
+                  
+                .text-big {
+                  font-size: 15px;
+                  margin: 10px 0;
+                }
+                
+                .text-article{
+                  font-size: 15px;
+                  margin: 10px 0;
+                }
+                  
+                }
+                
+                @media screen and (min-width: 768px){
+                
+                  
+                                 .text-big {
+                  font-size: 15px;
+                  margin: 20px 0;
+                }
+                
+                    .text-article{
+                  font-size: 15px;
+                  margin: 10px 0;
+                }
+                
+                
+                }
+                
+                
+                
+                /* These are the default paragraph, image and blockquote styles */
+
+
+                p {
+                  margin: 0 0 0 0;
+                }
+
+                img {
+                  max-width: 100%;
+                  height: auto;
+                  vertical-align: middle;
+                }
+
+                blockquote {
+                  margin: 30px 0 6px 0;
+                  font-size: 30px;
+                  color: #777;
+                }
+
+                small {
+                  font-size: 14px;
+                }
+                
+                
+                /* CLASSES */
+                .header {
+                  background-color: #f4f4f4;
+                  position: fixed;
+                  top: 0;
+                  left: 0;
+                  right: 0;
+                  height: 2vh;
+                  display: flex;
+                  align-items: center;
+                  justify-content: space-between; /* space between jump-link and logo */
+                  z-index: 10;
+                  padding: 0 15px;
+                }
+                
+                /* Jump link on the left */
+                .jump-link {
+                  font-size: 16px;
+                  white-space: nowrap;
+                }
+                
+                /* Responsive adjustments */
+                @media (max-width: 767px) {
+                  .header {
+                    padding: 0 10px;
+                  }
+                
+                  .jump-link {
+                    font-size: 10px;
+                    margin-right: 10px;
+                  }
+                
+                  .logo-img {
+                    height: 5vh;
+                  }
+                }
+
+
+                  
+                /* This is for fixed background images sometimes visible between content */
+                
+
+                @media screen and (min-width: 768px){
+
+                .bigfixedimage{
+                height: 93vh;
+                width:100%;
+                background-attachment: fixed;
+                background-position: center;
+                background-repeat: no-repeat;
+                background-size:cover;cover;
+
+                border-radius: 25px;}
+                
+                .bigfixedimage2{
+                height: 93vh;
+                width:100%;
+                background-attachment: relative;
+                background-position: center;
+                background-repeat: no-repeat;
+                                background-size:cover;cover;
+
+                border-radius: 25px;}
+                
+                }
+                
+                @media screen and (max-width: 767px){
+
+                .bigfixedimage{
+                height: 30vh;
+                background-position: center;
+                background-repeat: no-repeat;
+                background-size:cover;auto;
+                border-radius: 25px;}
+                
+                .bigfixedimage2{
+               height: 30vh;
+                               width:100%;
+
+                background-position: center;
+                background-repeat: no-repeat;
+                background-size:cover;auto;
+                border-radius: 25px;}
+                
+                }
+                
+                
+                
+                @media screen and (min-width: 768px){
+
+                        
+                /* This is a colour panel */
+                .colourheader{
+                color: #fff;
+                background-color: #010056;
+                background-image: none;
+                background-attachment: fixed;
+                /*min-height: 93vh;*/
+                border-radius: 25px;
+                }
+                
+                .colourheader2{
+                color: black !important;
+                background-color: #F36617 !important;
+                background-image: none;
+                background-attachment: fixed;
+                /*min-height: 93vh;*/
+                border-radius: 25px;
+                }
+                
+                .colourheader3{
+                color: black;
+                background-color: #F4F4F4;
+                background-image: none;
+                background-attachment: fixed;
+                /*min-height: 93vh;*/
+                border-radius: 25px;
+                }
+                  
+                /* This pads element so it is a smaller block within contaiing element, eg to make text start closer to center */
+
+                .paddeddiv{
+                position: relative;
+                box-sizing: border-box;
+                padding: 40px 0;
+                /*min-height: 85vh;*/
+                height: 100%;
+                display: flex;
+                flex-direction: column;
+                justify-content: center;
+                width: 100%;
+                max-width: 980px;
+                margin: 0 24px;
+                }
+                  
+                }
+                
+                @media screen and (max-width: 767px){
+
+                        
+                /* This is a colour panel */
+                .colourheader{
+                color: #fff;
+                background-color: #010056;
+                background-image: none;
+                background-attachment: fixed;
+                /*min-height: 30vh!important;*/
+                border-radius: 25px;
+                }
+                
+                .colourheader2{
+                color: black !important;
+                background-color: #F36617 !important;
+                background-image: none;
+                background-attachment: fixed;
+                /*min-height: 30vh!important;*/
+                border-radius: 25px;
+                }
+                
+                .colourheader3{
+                color: black;
+                background-color: #F4F4F4;
+                background-image: none;
+                background-attachment: fixed;
+                /*min-height: 93vh;*/
+                border-radius: 25px;
+                }
+                
+                
+                .paddeddiv{
+                position: relative;
+                box-sizing: border-box;
+                padding: 10px 0;
+                min-height: 25vh;
+                height: 100%;
+                display: flex;
+                flex-direction: column;
+                justify-content: center;
+                width: 100%;
+                max-width: 980px;
+                margin: 0 24px;
+                }  
+                  
+                  
+                }
+                
+                
+
+               .col-medium {
+                width: 100%;
+                max-width: 680px;
+                margin: 0 24px;
+                }
+ 
+
+                .text-indent {
+                  margin-left: 30px;
+                }
+ 
+.selectize-input { font-size: 16px; line-height: 16px;}
+.selectize-dropdown { font-size: 16px; line-height: 16px; } 
+
+.sankey-link:hover { fill-opacity: 0.75 !important; }
+
+.plotly .hoverlayer .hovertext { fill-opacity: 1 !important; stroke-opacity: 1 !important; }
+
+/* Allow scrolling on mobile over graphs */
+.scrolly {
+    touch-action: pan-y !important;
+    -ms-touch-action: pan-y !important;
+}
+
+.scrolly figure.sticky {
+    touch-action: pan-y !important;
+    pointer-events: auto !important;
+}
+
+.plotly-scrolling-over {
+  touch-action: pan-y !important;
+  overflow-y: auto;
+}
+
+.scrolly .maindrag,
+.scrolly .angulardrag,
+.scrolly .radialdrag,
+.scrolly .radialdrag-inner {
+    pointer-events: none !important;
+}
+
+/* Make cut 1 over time larger on mobile screen*/
+@media screen and (max-width: 767px){
+  .scrolly figure.sticky {
+    height: 60vh !important;
+    top: 10vh !important;
+  }
+
+  .scrolly figure.sticky .plotly {
+    height: 100% !important;
+  }
+}
+