Research
52,841 matches from six top European leagues since 2000, checked against their betting odds.
When the odds said 67%, it happened 70% of the time. When they said 12%, it happened 10% of the time.
Each dot is a group of results with similar odds. Dots on the dashed line mean the odds were exactly right.
| Odds said | Happened | Results |
|---|---|---|
| 12% | 10% | 13,211 |
| 19% | 18% | 13,210 |
| 23% | 22% | 13,210 |
| 26% | 25% | 13,210 |
| 28% | 28% | 13,210 |
| 28% | 28% | 13,210 |
| 30% | 30% | 13,211 |
| 33% | 32% | 13,210 |
| 38% | 38% | 13,210 |
| 44% | 45% | 13,210 |
| 52% | 53% | 13,210 |
| 67% | 70% | 13,211 |
Favorites won slightly more often than their odds said, and long shots slightly less (calibration slope 1.10, 95% range 1.07 to 1.12; 1.00 is perfect).
Betting $100 on every heavy favorite (−200 or shorter) lost $1 on average. Betting $100 on every long shot (+900 or longer) lost $27.
Average result of a $100 bet on every result in each odds range, at a typical bookmaker's price
| Odds | Won or lost per $100 |
|---|---|
| −200 or shorter | −$1 |
| −200 to +100 | −$4 |
| +100 to +150 | −$4 |
| +150 to +200 | −$9 |
| +200 to +300 | −$7 |
| +300 to +400 | −$12 |
| +400 to +600 | −$11 |
| +600 to +900 | −$20 |
| +900 or longer | −$27 |
At the best price across bookmakers, heavy favorites returned +$1 per $100 and long shots −$11.
The average bookmaker's margin fell from 13.0% in 2000 to 6.5% this season.
The bookmaker's built-in cut on an average match, by season
| Season | Cut |
|---|---|
| 2000/01 | 13.0% |
| 2001/02 | 12.1% |
| 2002/03 | 11.5% |
| 2003/04 | 11.7% |
| 2004/05 | 11.2% |
| 2005/06 | 10.4% |
| 2006/07 | 9.9% |
| 2007/08 | 9.1% |
| 2008/09 | 8.2% |
| 2009/10 | 7.6% |
| 2010/11 | 6.9% |
| 2011/12 | 6.8% |
| 2012/13 | 6.3% |
| 2013/14 | 6.1% |
| 2014/15 | 5.6% |
| 2015/16 | 5.6% |
| 2016/17 | 5.1% |
| 2017/18 | 5.2% |
| 2018/19 | 4.8% |
| 2019/20 | 5.0% |
| 2020/21 | 4.9% |
| 2021/22 | 4.5% |
| 2022/23 | 4.5% |
| 2023/24 | 4.5% |
| 2024/25 | 4.9% |
| 2025/26 | 6.4% |
| 2026/27 | 6.5% |
Long shots lost more than favorites in 25 of 27 seasons.
Result of a $100 bet on every heavy favorite or every long shot (smoothed over 3 seasons)
| Season | Favorites | Long shots |
|---|---|---|
| 2000/01 | −$5 | −$2 |
| 2001/02 | −$2 | −$18 |
| 2002/03 | −$1 | −$16 |
| 2003/04 | −$8 | −$11 |
| 2004/05 | −$8 | −$25 |
| 2005/06 | −$4 | −$36 |
| 2006/07 | −$5 | −$22 |
| 2007/08 | −$5 | −$11 |
| 2008/09 | −$1 | −$22 |
| 2009/10 | −$4 | −$24 |
| 2010/11 | −$6 | −$11 |
| 2011/12 | −$4 | −$11 |
| 2012/13 | −$3 | −$16 |
| 2013/14 | $0 | −$16 |
| 2014/15 | −$3 | −$12 |
| 2015/16 | −$4 | −$24 |
| 2016/17 | +$3 | −$23 |
| 2017/18 | −$2 | −$16 |
| 2018/19 | −$2 | −$15 |
| 2019/20 | −$3 | +$1 |
| 2020/21 | −$1 | −$1 |
| 2021/22 | −$2 | −$8 |
| 2022/23 | −$3 | −$3 |
| 2023/24 | $0 | −$26 |
| 2024/25 | −$3 | −$22 |
| 2025/26 | −$7 | −$23 |
| 2026/27 | −$7 | −$26 |
When the odds shortened before kick-off (chance up 3+ points), a $100 bet at the opening odds made +$11 on average. When they drifted, it lost $15.
Result of a $100 bet at the opening price, split by whether the chance moved 3+ points before kick-off
| Before kick-off | Won or lost per $100 |
|---|---|
| Chance rose | +$11 |
| Chance fell | −$15 |
31,790 matches, 2012 to 2026. Closing odds were more accurate than opening odds: error score 0.9800 against 0.9823 (lower is better).
Calibration slope (1.00 is perfect): 1.09 in the Premier League, 1.06 in the Championship.
| League | Matches | Margin | Slope | Betting $100 on everything |
|---|---|---|---|---|
| Premier League | 9,806 | 5.7% | 1.09 | −$8 |
| Championship | 12,680 | 6.4% | 1.06 | −$7 |
Teams losing by 2+ at half-time scored 15% more in the second half than the odds predicted. Teams winning by 2+ scored 5% fewer.
Second-half goals compared with what the pre-match odds predicted, by the team's score at half-time
| At half-time | Goals vs expected |
|---|---|
| down 2+ | +15% |
| down 1 | +10% |
| level | -2% |
| up 1 | -5% |
| up 2+ | -5% |
Slight favorites leading 1–0 at half-time were given a 76% chance to win and won 72% of the time (10,249 matches). Games level at half-time were given a 38% chance of a draw and ended level 40% of the time.
In the second half, the team that was behind was expected to score 43.7% of the goals and scored 47.0%. This held in 12 of 12 checks.
How much more of the second-half scoring the trailing team got than the odds expected
| Check | Extra share |
|---|---|
| All matches with a lead at half-time | +8% |
| Lead of one goal | +7% |
| Lead of two or more | +10% |
| One goal in the first half (1–0) | +7% |
| Three or more first-half goals | +10% |
| No red card in the match | +8% |
| Home team trailing | +4% |
| Away team trailing | +11% |
| 2000–2010 | +8% |
| 2010–2020 | +8% |
| Premier League | +7% |
| Championship | +7% |