Myth or Maths
Football's favourite claims, tested against the data. Some survive.
Not supported
- Data
- StatsBomb open data, men's and women's football
- Sample
- 1,557 penalties, 376 left-footed and 1,181 right-footed
- Method
- Scoring rates by foot with a luck margin, in play and in shootouts, with and without Messi, and by where the kick went
- Finding
- No difference: 74.2% for each foot
Not supported
- Data
- Scottish Premiership, 2000/01 to 2025/26
- Sample
- 82 international breaks, 485 first matches back
- Method
- Results after a break against all other matches, and against the bookmakers' odds, club by club
- Finding
- No difference beyond luck, league-wide, for the Old Firm or club by club
Not supported
- Data
- Premiership results and every in-season managerial change, 2000/01 to 2025/26
- Sample
- 80 changes with six league games either side
- Method
- Each change against sides with the same run and strength that kept their manager
- Finding
- Both bounce by 0.56 points a game; the change itself adds nothing
Intermediate
Supported
- Data
- Scottish Premiership results, 2000/01 to 2025/26
- Sample
- 26 seasons
- Method
- The table on Christmas Day against the final table
- Finding
- 18 of 26 sides bottom at Christmas finished bottom; every one that escaped was within a point
Beginner
Not supported
- Data
- Premiership results, cards and Bet365 odds
- Sample
- 297 derbies with odds
- Method
- Favourites' results against what the odds said, and cards per game
- Finding
- Favourites won about as often as the odds said; about three-quarters of a yellow card more a game
Beginner
Not supported
- Data
- Scottish Premiership results, 2000/01 to 2025/26
- Sample
- 23 promoted sides that stayed up
- Method
- Points in the first season against the second
- Finding
- 12 did better and 11 worse; over-performers fall back, promoted or not
Beginner
Not supported
- Data
- Premiership and Championship, 2000/01 to 2025/26
- Sample
- 10,471 matches
- Method
- Midweek against weekend results, upsets and goals
- Finding
- Same results, same upsets and, like for like, the same goals
Beginner
Not supported
- Data
- Premiership and Championship, 2000/01 to 2025/26
- Sample
- 10,471 matches, 26 seasons
- Method
- The share of 0–0s, season by season
- Finding
- One match in 14 ends 0–0, and the share hasn't budged
Beginner
Mixed
- Data
- Scottish Premiership results, 2000/01 to 2025/26
- Sample
- 25 seasons with a split
- Method
- Home games and fixtures after the split
- Finding
- The home games even out; the side 7th takes 2.6 more points from the last five than the side 6th, on average
Beginner
Not supported
- Data
- Premiership corners and shots, 2000/01 to 2025/26
- Sample
- 5,877 matches
- Method
- Results by who won the corner count, then corners and shots together
- Finding
- The side with more corners won only 44%; once you know the shots, corners add nothing
Beginner
Not supported
- Data
- Premiership results with shots, 2000/01 to 2025/26
- Sample
- 1,546 six-game spells
- Method
- Recent points and shot difference against the next six games
- Finding
- Shot difference predicts better; an 8-point form gap shrinks to under one
Intermediate
Not supported
- Data
- All four Scottish divisions, 2000/01 to 2025/26
- Sample
- Over 18,000 matches
- Method
- Home minus away goals a game, with a confidence interval
- Finding
- About a quarter of a goal, not one
Beginner
Mixed
- Data
- Scottish Premiership results, 2000/01 to 2025/26
- Sample
- 312 team-seasons
- Method
- The spread luck alone would give, against the real spread
- Finding
- Luck moves a team's points by about eight either way: enough for mid-table, not for a title
Intermediate
Supported
- Data
- Premiership shots and shots on target, 2000/01 to 2025/26
- Sample
- 5,877 matches
- Method
- Results by who had more shots, and more on target
- Finding
- More shots won twice as often as they lost; more shots but fewer on target lost over half
Beginner
Not supported
- Data
- Premiership and Championship half-time and full-time scores
- Sample
- 10,471 matches
- Method
- How each half-time lead ended
- Finding
- 2–0 leads were won 92% of the time and lost about one in forty
Beginner