xgabora/Club-Football-Match-Data-2000-2025

⚽ Club Football Match Data (2000 - 2025)

This dataset offers a simple entrance to the world of football match data analysis. It offers football match data from 27 countries and 42 leagues worldwide, including some of the best leagues such as the English Premier League, German Bundesliga, and Spanish La Liga. The data spans from the 2000/01 season to the most recent results from the 2024/25 season. The dataset also includes Elo Ratings for the given time period with snapshots of ~500 of the best teams in Europe taken twice a month, on the 1st and 15th.

Match results and statistics provided in the table are taken from Football-Data.co.uk. Elo data are taken from ClubElo.

---

📅 DATASET OVERVIEW

📂 Files number: 2

🔗 Files type: .csv

⌨️ Total rows: ~470 000 as of 03/2025

💾 Total size: ~48MB as of 03/2025

The dataset is a great starting point for football match prediction, both pre-match and in-play, with huge potential lying in the amount of data and their accuracy. The dataset contains information about teams' strength and form prior to the match, as well as general market predictions via pre-match odds.

---

🔑 KEY FEATURES

1️⃣ SIZE - This is the biggest open and free dataset on the internet, keeping uniform information about tens of thousands of football matches, including match statistics, odds, and Elo and form information.

2️⃣ READABILITY - The whole dataset is tabular, and all of the data are clear to navigate and explain. Both tables in the dataset correspond to each other via remapped club names, and all of the formats within the table (such as odds) are uniform.

3️⃣ RECENCY - This is the most up-to-date open football dataset, containing data from matches as recent as March 2025. The plan is to update this dataset monthly or bi-monthly via a custom-made Python pipeline.

---

📋 COLUMNS AND DESCRIPTIONS

💪 TABLE 1 - ELO RATINGS.csv

This table is a collection of Elo ratings taken from ClubElo. Snapshots are taken twice a month, on the 1st and 15th day of the month, saving the whole Club Elo database. Some clubs' names are remapped to correspond with the Matches table (for example "Bayern" to "Bayern Munich").``

Column	Data Type	Description
📅 Date	date	Date of the snapshot.
🛡️ Club	string	Club name in English corresponding to Matches table.
🌍️ Country	enum	Club country three-letter code.
📈 Elo	float	Club's current Elo rating, rounded to two decimal spots.

💪 TABLE 2 - MATCHES.csv

This table is a collection of historical match results and statistics taken from Football-Data.co.uk. The data are ordered by date, starting from 28th July (the beginning of the 2000/01 season) up until 23rd December 2024 (the 2024/25 season). The table contains the most important information about both teams and about the match itself. The table is highly incomplete due to the differences in statistics provided by each league. However, it is possible to create smaller sub-tables or clean this one to remove null values. As of January 2025, the table contains about 226,000 unique match records.

Column	Data Type	Description
🏆 Division	enum	League that the match was played in - country code + division number (I1 for Italian First Division). For countries where we only have one league, we use 3-letter country code (ARG for Argentina).
📆 MatchDate	date	Match date in the classic YYYY-MM-DD format.
🕘 MatchTime	time	Match time in the HH:MM:SS format. CET-1 timezone.
🏠 HomeTeam	string	Home team's club name in English, abbreviated if needed.
🚗 AwayTeam	string	Home team's club name in English, abbreviated if needed.
📊 HomeElo	float	Home team's most recent Elo rating.
📊 AwayElo	float	Away team's most recent Elo rating.
📉 Form3Home	int	Number of points gathered by home team in the last 3 matches (Win = 3 points, Draw = 1 point, Loss = 0 points, so this value is between 0 and 9).
📈 Form5Home	int	Number of points gathered by home team in the last 5 matches (Win = 3 points, Draw = 1 point, Loss = 0 points, so this value is between 0 and 15).
📉 Form3Away	int	Number of points gathered by away team in the last 3 matches (Win = 3 points, Draw = 1 point, Loss = 0 points, so this value is between 0 and 9).
📈 Form5Away	int	Number of points gathered by away team in the last 5 matches (Win = 3 points, Draw = 1 point, Loss = 0 points, so this value is between 0 and 15).
⚽ FTHome	int	Full-time goals scored by home team.
⚽ FTAway	int	Full-time goals scored by away team.
🏁 FTResult	enum	Full-time result (H for Home win, D for Draw and A for Away win).
⚽ HTHome	int	Half-time goals scored by home team.
⚽ HTAway	int	Half-time goals scored by away team.
⏱️ HTResult	enum	Half-time result (H for Home win, D for Draw and A for Away win).
🏹 HomeShots	int	Total shots (goal, saved, blocked, off-target) by home team.
🏹 AwayShots	int	Total shots (goal, saved, blocked, off-target) by away team.
🎯 HomeTarget	int	Total shots on target (goal, saved) by home team.
🎯 AwayTarget	int	Total shots on target (goal, saved) by away team.
🤕 HomeFouls	int	Total fouls by home team.
🤕 AwayFouls	int	Total fouls by away team.
🚩 HomeCorners	int	Total corners taken by home team.
🚩 AwayCorners	int	Total corners taken by away team.
🟨 HomeYellow	int	Total yellow cards awarded to home team players (excl. staff).
🟨 AwayYellow	int	Total yellow cards awarded to away team players (excl. staff).
🟥 HomeRed	int	Total red cards awarded to home team players (excl. staff).
🟥 AwayRed	int	Total red cards awarded to away team players (excl. staff).
1️⃣ OddHome	float	Bet365's Home Team Win Odd.
0️⃣ OddDraw	float	Bet365's Draw Odd.
2️⃣ OddAway	float	Bet365's Away Team Win Odd.
1️⃣ MaxHome	float	Maximum Home Team Win Odd from ~17 European bookmakers.
0️⃣ MaxDraw	float	Maximum Draw Odd from ~17 European bookmakers.
2️⃣ MaxAway	float	Maximum Away Team Win Odd from ~17 European bookmakers.
⬆️ Over25	float	Bet365's Over 2.5 Total Goals Scored Odd.
⬇️ Under25	float	Bet365's Under 2.5 Total Goals Scored Odd.
⬆️ MaxOver25	float	Maximum Over 2.5 Total Goals Scored Odd from ~17 European bookmakers.
⬇️ MaxUnder25	float	Maximum Under 2.5 Total Goals Scored Odd from ~17 European bookmakers.
🟰 HandiSize	float	Asian handicap size for home team (negative number indicating stronger home team) .
➕️ HandiHome	float	Bet365's Home Team Win Odd with the given handicap size for Home team.
➖️ HandiAway	float	Bet365's Away Team Win Odd with the given handicap size for Home team.

---

🔭 FEATURE ENGINEERING

🔗 LAGGED FEATURES

Given that the dataset includes a time data, it offers a possibility to create lagged features. One of those features (Form) is already included in the dataset, but others might include things like goal potency, points gathered during the season, table position, streaks or Elo shifts. These are planned to be added into the dataset in the future as they might contain valuable information for match prediction:

Column	Data Type	Description
⚽️ GF3Home	int	Goals scored by the Home team in last 3 matches.
⚽️ GF3Away	int	Goals scored by the Away team in last 3 matches.
🥅 GA3Home	int	Goals conceded by the Home team in last 3 matches.
🥅 GA3Away	int	Goals conceded by the Away team in last 3 matches.
⚽️ GF5Home	int	Goals scored by the Home team in last 5 matches.
⚽️ GF5Away	int	Goals scored by the Away team in last 5 matches.
🥅 GA5Home	int	Goals conceded by the Home team in last 5 matches.
🥅 GA5Away	int	Goals conceded by the Away team in last 5 matches.
🥅 GA5Away	int	Goals conceded by the Away team in last 5 matches.
🏆 PointsHome	int	Points gathered in the respective season in the league by Home team.
🏆 PointsAway	int	Points gathered in the respective season in the league by Away team.
🏆 PositionHome	int	Home team's current position in the league.
🏆 PositionAway	int	Away team's current position in the league.
🔥 WStreakHome	int	Home team's number of won games in a row.
🔥 WStreakAway	int	Away team's number of won games in a row.
🔥 LStreakHome	int	Home team's number of lost games in a row.
🔥 LStreakAway	int	Away team's number of lost games in a row.
🔥 SStreakHome	int	Home team's number of games in a row they scored in.
🔥 SStreakAway	int	Away team's number of games in a row they scored in.
🏃 EloChange1Home	float	The difference between today's Home team Elo and Home team Elo a month before.
🏃 EloChange1Away	float	The difference between today's Away team Elo and Away team Elo a month before.
🏃 EloChange2Home	float	The difference between today's Home team Elo and Home team Elo two months before.
🏃 EloChange2Away	float	The difference between today's Away team Elo and Away team Elo two months before.
🏠 HomeRecord	int	Number of points gathered by home team in the last 5 home matches (Win = 3 points, Draw = 1 point, Loss = 0 points, so this value is between 0 and 15).
🚗 AwayRecord	int	Number of points gathered by away team in the last 5 away matches (Win = 3 points, Draw = 1 point, Loss = 0 points, so this value is between 0 and 15).
💆‍♂️ RestDaysHome	int	Number of days between Home team's last match and today (note this might not give correct and accurate results because of other competitions played in between domestic league matches, such as domestic cups and continental cups, as well as international break games).
💆‍♂️ RestDaysAway	int	Number of days between Away team's last match and today (note this might not give correct and accurate results because of other competitions played in between domestic league matches, such as domestic cups and continental cups, as well as international break games).

⛓️‍💥 DERIVED FEATURES

Derived features are computed quickly on-the-go and that's why they are not included in the base dataset, however they may still contain some very useful information that can affect teams' performances such as their relative chances, attacking potence or defensive strength. These are just some of the suggestions for the derived features you can get from the dataset that could come in handy when using the dataset:

Column	Data Type	Description
🕘 MatchDateTime	datetime	Combining match day and match time as one data type.
⚽️ TotalGoals	int	Goals scored by Home team + Goals scored by Away team.
1️⃣0️⃣ 1XOdd	float	Combined odd for Double chance bet - Home Team Win or a Draw (making Max1XOdd is also possible).
0️⃣2️⃣ X2Odd	float	Combined odd for Double chance bet - Away Team Win or a Draw (making MaxX2Odd is also possible).
1️⃣2️⃣ 12Odd	float	Combined odd for Double chance bet - Home Team Win or an Away Team Win (making Max12Odd is also possible).
📊 EloDifference	float	Difference between Home Team Elo rating and Away Team Elo rating.
📊 EloTotal	float	Home Team Elo rating + Away Team Elo rating.
📊 EloAdvantage	float	EloDifference divided by EloTotal - normalizes the Elo difference to become a percentage.
📉 Form3Difference	int	Home Team Form3 - Away Team Form3.
📉 Form5Difference	int	Home Team Form5 - Away Team Form5.
↗️ FormMomentumHome	int	Difference between Home Team's recent and older form, derived from the formula Form3Home - (Form5Home - Form3Home). Values lie between -15 for the worst momentum and 18 for the best momentum.
↗️ FormMomentumAway	int	Difference between Away Team's recent and older form, derived from the formula Form3Away - (Form5Away - Form3Away). Values lie between -15 for the worst momentum and 18 for the best momentum.
1️⃣ ImpliedProbHome	float	Probability value for Home Win derived from OddHome using formula 1/OddHome.
0️⃣ ImpliedProbDraw	float	Probability value for a Draw derived from OddDraw using formula 1/OddDraw.
2️⃣ ImpliedProbAway	float	Probability value for Away Win derived from OddAway using formula 1/OddAway.
➗️ ImpliedProbTotal	float	ImpliedProbHome + ImpliedProbDraw + ImpliedProbAway.
💰️ BookmakerMargin	float	ImpliedProbTotal - 1. Understoot as a "market uncertainty", can help distinguishing between clear-favorite matches and matches that can go either way.
🏹 ShotsDifference	int	HomeShots - AwayShots.
🏹 ShotsTotal	int	HomeShots + AwayShots.
🎯 ShotAccuracyHome	float	HomeTarget / HomeShots. Gives us % accuracy, this tends to go up as match progresses. Better team does not necessarily need to have better shot accuracy.
🎯 ShotAccuracyAway	float	AwayTarget / AwayShots. Gives us % accuracy, this tends to go up as match progresses. Better team does not necessarily need to have better shot accuracy.
🎯 ShotAccuracyDiff	float	ShotAccuracyHome - ShotAccuracyAway.
🗡️ ScoringEfficiencyHome	float	Derived from Home Team's goals and Home Team's shots on target. Home Elo, Odds and Form can also be taken into account as well as historic scoring ratios.
🗡️ ScoringEfficiencyAway	float	Derived from Away Team's goals and Away Team's shots on target. Away Elo, Odds and Form can also be taken into account as well as historic scoring ratios.
🛡️ DefensiveRatingHome	float	Derived from Home Team's fouls and Away Team's shots and shots on target. Home Elo, Odds and Form can also be taken into account as well as historic scoring ratios.
🛡️ DefensiveRatingAway	float	Derived from Away Team's fouls and Home Team's shots and shots on target. Away Elo, Odds and Form can also be taken into account as well as historic scoring ratios.
🚩 CornersDifference	int	HomeCorners - AwayCorners.
🚩 CornersTotal	int	HomeCorners + AwayCorners.
🏋️ GameDominanceIndex	float	GDI, to some extent reflexing match ball possesion and the number of attacking chances. Derived from shot and corner data using formula (CornersDifference+ShotsDifference)/2), can be fine-tuned.
⛔️ CardPointsHome	int	HomeYellow + 2* HomeRed. Multiplier of 3 or 4 can also be used to get better results in Machine Learning.
⛔️ CardPointsAway	int	AwayYellow+ 2* AwayRed. Multiplier of 3 or 4 can also be used to get better results in Machine Learning.
⛔️ CardPointsDiff	int	CardPointsHome - CardPointsAway.
🧱 DrawLikelihood	float	Derived from EloDifference, Form5Difference and ImpliedProbDraw. Gives a weighted % likelihood of a draw, which is notoriously difficult to predict when predicting football matches. Used to counter-weight the clear-winner-biased Machine Learning methods.
🧤 CleanSheetProbHome	float	Derived from DefensiveRatingHome and ScoringEfficiencyAway. Gives a weighted % likelihood of Home Team keeping a clean sheet.
🧤 CleanSheetProbAway	float	Derived from DefensiveRatingAway and ScoringEfficiencyHome. Gives a weighted % likelihood of Away Team keeping a clean sheet.
🔍️ ExpectedGoalsHome	float	Derived from HomeTarget, ShotAccuracyHome, OddHome and recent home scoring trends (not to be confused with popular xG metric not included in this dataset).
🔍️ ExpectedGoalsAway	float	Derived from AwayTarget, ShotAccuracyAway, OddAway and recent away scoring trends (not to be confused with popular xG metric not included in this dataset).

---

🚀 POSSIBLE APPLICATIONS

⌛️ History book: Has a team ever won when 3-0 down at halftime? Browse historical data and find bizarre records, oddities, and against-the-odds performances of your favorites.

🔎 Hypothesis testing: Are more goals scored during late-night matches? Do Premier League players have better shot accuracy? Test your own hypothesis on this huge dataset.

📊 Match prediction: Who's going to win and by what margin? Predict match outcomes, total goals, or team goals before or during the actual game using historical data, team form data, and match stats.

🆚 Team comparison: United 2007/08 or City 2022/23? Compare teams' historical seasons, their points, game results, and form streaks during different periods of time.

🖼️ Data visualization: Elo strength over time, average goals scored per league, or anything else - create data visualizations and accurate graphs from any metrics possible.