Bankroll management is the single most common reason recreational bettors lose money over time. Not lack of edge. Not bad luck. Position sizing. A bettor with a genuine 3% edge can still go bankrupt if they bet 25% of their bankroll on each position. The math of variance in a high-variance domain like football betting punishes oversized bets disproportionately.
The Kelly Criterion as Starting Point
The Kelly Criterion calculates the mathematically optimal bet size based on your edge and the offered odds. The formula is:
Kelly % = (bp − q) / b
Where b is the net decimal odds (decimal odds minus 1), p is your estimated probability of winning, and q is 1 minus p.
Example: You bet a market at decimal 2.50 (b = 1.50). You estimate the true probability at 45% (p = 0.45, q = 0.55). Kelly % = (1.50 × 0.45 − 0.55) / 1.50 = (0.675 − 0.55) / 1.50 = 0.083 or 8.3% of bankroll.
The Kelly tab in the odds calculator handles this math for you. Input bankroll, odds, and your estimated probability, and it returns the recommended stake.
Why Full Kelly Is Too Aggressive
Kelly optimises for long-term geometric bankroll growth, assuming your probability estimates are accurate. The problem is that your probability estimates are not perfectly accurate — they are themselves estimates with error bars. Overconfidence in your estimates at full Kelly stakes can produce catastrophic drawdowns that take years to recover from.
Professional bettors typically use a fraction of Kelly — half Kelly, quarter Kelly, or even smaller. The trade-off is that growth rate is slower (half Kelly produces about 75% of full Kelly’s expected growth), but the drawdown risk drops sharply (half Kelly has roughly a quarter of full Kelly’s maximum drawdown risk).
Practical Sizing Framework
| Edge confidence | Recommended Kelly fraction | Stake on a 5% edge |
|---|---|---|
| High (model-based, verified) | Half Kelly | ~2.5% of bankroll |
| Moderate (informed estimate) | Quarter Kelly | ~1.25% of bankroll |
| Speculative (rough estimate) | Tenth Kelly or 1% flat | ~0.5-1% of bankroll |
The Drawdown Reality
Even with disciplined sizing, drawdowns happen. A bettor running half Kelly with a 3% edge can still experience a 30% bankroll drawdown over a sample of 200 bets. That is the nature of football betting variance — binary outcomes, infrequent samples, high per-bet variance. The mental discipline to keep betting through a drawdown of that size, without doubling stakes to chase losses, separates long-term winners from everyone else.
The Flat-Stake Alternative
For bettors who do not trust their probability estimates enough to vary stake size, flat staking is a reasonable simplification. Pick 1-2% of bankroll per bet, apply it uniformly, and let the edge work itself out over volume. The expected growth rate is lower than Kelly’s, but the implementation discipline is higher. Most casual bettors who try Kelly end up overestimating their edges and effectively betting full Kelly on what they believed was half Kelly. Flat staking removes that failure mode.
Further reading on validating your edge: closing line value as a profitability predictor. For the broader strategic framework, see the Premier League betting guide.