Level Stakes vs Proportional Staking on Greyhounds — Which Wins Long-Term?

What's the difference between level stakes and proportional staking?

**Level stakes:** the same unit on every bet. £10 today, £10 tomorrow, £10 next month. Simple, traceable, and the cleanest way to measure whether a system is genuinely profitable.

**Proportional staking:** stake size scales with bankroll. If your bankroll grows 20%, your unit grows 20%. If it shrinks 20%, your unit shrinks 20%. More complex, but theoretically maximises long-run growth when you have a real edge.

When should I use level stakes?

**When you're testing a new system.** A 50-pick sample on level stakes gives you a clean ROI number. Mixing in proportional changes mid-test makes the result harder to interpret. The Historic Results page shows how level-stake outcomes settle over longer windows.

**When you're a recreational bettor.** The simplicity is a feature, not a bug. You won't lose track of your stake size, you won't compound mistakes, and you can compare your results directly against the system's reported ROI.

**When your edge is small or unproven.** Proportional staking only outperforms level stakes when the edge is real and reasonably large. With marginal edges, the extra variance from scaling stakes can hurt more than it helps.

When should I use proportional staking?

**When you have a proven, sizeable edge.** A system with 12%+ ROI over 200+ picks is sizeable enough to benefit from proportional growth. Below 5% ROI, the math gets noisy — you may grow modestly when winning but lose proportionally faster when variance bites.

**When you're committing to long-term play.** Proportional staking compounds returns. Over 1,000+ bets at a sustainable edge, the difference between flat and proportional stakes can be substantial. If you're betting for the next year, proportional can pay.

What about Kelly Criterion?

Kelly is a specific form of proportional staking — it tells you the optimal bet size as a fraction of bankroll based on your edge and the odds. The full-Kelly formula is mathematically optimal but impractical (it produces stake sizes that swing wildly, which most bettors can't psychologically tolerate).

**Half-Kelly or quarter-Kelly is more realistic.** You get most of Kelly's growth advantages with a lot less variance. For a 25%-strike-rate system at average 4.0 SP, quarter-Kelly typically suggests ~2-3% of bankroll per bet — which is broadly aligned with the standard 1-2% bankroll management advice.

What's the simplest middle ground?

Level stakes within fixed periods, then recalibrate. Set a unit for the next 50 bets at 1-2% of current bankroll. Bet level stakes through those 50 bets. After 50, recalculate the unit based on new bankroll size and continue.

This approach gives you most of proportional staking's growth benefits with most of level staking's simplicity. The recalibration cadence (50 bets, 100 bets, etc.) is up to you — the cleaner the cadence, the easier it is to track.