Football Statistics
By Thomas DuboisData & Football AnalystHey, it's Thomas. Here's a statistical concept with a slightly intimidating name, but one that can genuinely boost your betting game: Regression to the Mean. Once you grasp it, you'll view "incredible streaks" with fresh eyes, often giving you a significant edge over the market.
The Principle, Explained Simply
Regression to the Mean is the idea that after an extreme performance, results tend to revert to the average.An analogy: if you flip a coin 10 times and it lands on tails 9 times, the next 10 flips won't be 9 tails. They'll trend back towards 50/50, the normal 'mean'. That extraordinary sequence was a stroke of luck, not a new rule.
In football, it's the same: a team that scores an abnormally high number of goals or goes on an exceptional winning streak will statistically revert to its true level. The extraordinary never lasts. Never.
Why This Is a Goldmine for Betting
Here's the practical benefit. The odds market often overreacts to streaks. A team winning 6 matches in a row becomes a systematic heavy favourite, with very low odds. If this streak is due to luck (and not genuine improvement), the team will regress, and betting against them can offer value.
Conversely, an unlucky team (one that plays well but loses) is undervalued by a market that only looks at results. They will rebound. Again, an opportunity.
Signal #1: The xG Discrepancy
How do you know if a streak is about to regress? My favourite tool: compare actual results to underlying stats, especially xG.
| Situation | Diagnosis |
|---|---|
| Winning frequently, low xG created, conceding high xG | Overperforming: will regress (come back down) 📉 |
| Losing, but dominating on xG | Underperforming: will rebound 📈 |
| Results aligned with xG | True level, no regression expected |
Regression ≠ Bad Form
An important nuance not to confuse. Not every slump is a regression:
- Genuine bad form can reflect a real problem (injuries, a tactical crisis, a failed managerial change).
- Regression, on the other hand, is a return to normal after an abnormal sequence due to luck.
The key to distinguishing them: again, xG. If poor results are accompanied by good xG, it's bad luck that will correct itself. If the xG is also poor, it's a genuine decline, not a regression. Don't mix the two; it's a classic mistake.
The Trap: Don't Bet Against a Streak Blindly
Crucial warning. Don't bet against a streak just because it's long. A streak can reflect genuine improvement:
- a new coach who has transformed everything,
- a game-changing signing who transforms the team,
- new collective maturity.
In such cases, the team's 'average' has genuinely risen, and the streak isn't just luck. Always verify if the performance is supported by underlying stats (xG). Betting against a streak without this analysis is like gambling against chance without information, as risky as blindly following it.
A Word of Caution
As with all betting, let's be clear-headed. Regression to the Mean is not a certainty on any given match: a lucky team can remain so for a few more games. It's a statistical trend over time, a tool to identify mispriced odds, not a guarantee of an immediate turnaround. Patience and value remain your best allies.
In Summary
- Regression to the Mean: after an extreme performance, results revert to the average.
- The market overreacts to streaks: an opportunity to bet against overperformers and on the unlucky.
- Signal #1: the discrepancy between results and xG (overperformance = will regress).
- Regression ≠ bad form: xG helps distinguish between luck and genuine decline.
- Trap: don't bet against a streak blindly (it might be deserved).
- Statistical trend, not immediate certainty. Bet responsibly.
Frequently asked questions
How does Regression to the Mean help with betting?
It helps you identify teams whose current results are either too good or too bad to be sustainable. A team that's overperforming due to abnormal success in front of goal will likely come back down: the market overvalues them, and you can bet against them. Conversely, an unlucky team will rebound.
How do you spot a team that's about to regress?
Compare their actual results to their underlying statistics, especially xG. A team that's winning frequently but creating little xG and conceding a lot of xG is overperforming: they're getting lucky, and that luck will run out. The discrepancy between results and xG is the strongest signal of impending regression.
Does Regression to the Mean guarantee a turnaround?
No, it's not a certainty on any given match, but a statistical trend over time. A lucky team can remain so for a few more games. It's a tool to identify mispriced odds, not a guarantee of an immediate result. Patience and value remain essential.
What's the difference between Regression to the Mean and bad form?
Bad form can reflect a genuine problem (injuries, a crisis), whereas regression involves a return to normal after an abnormal sequence due to luck. The key is xG: if poor results are accompanied by good xG, it's bad luck that will correct itself, not a genuine decline.
Should you always bet against exceptional streaks?
Not blindly. A streak can reflect genuine improvement (a new coach, a game-changing signing) and not just luck. Always verify if the performance is supported by underlying stats. Betting against a streak solely because it's long, without analysis, is risky.
Data analyst who moved into sport. I break down the stats so you don't have to.