Bankroll Management for Long-Term Value Betting

Flat, fixed-percentage staking beats confidence-based sizing every time. Here's why, and what a sensible starting percentage looks like.

Last updated 2026-09-17

Finding genuine value bets is only half of what makes value betting work over time. The other half — arguably the more important half for anyone's long-term results — is how much of your bankroll you stake on each bet. Get this wrong, and even a genuinely profitable strategy can wipe out an account during a normal losing streak.

Why flat, fixed staking beats "confidence-based" staking

A common beginner instinct is to bet more when you "feel" more confident about a bet and less when you don't. The problem: your confidence is a poor substitute for the actual statistical edge, and confidence-based sizing tends to correlate with recent results rather than the underlying quality of the value identified — you feel confident after a few wins and cautious after a few losses, which is precisely backward from what a disciplined process should do.

Flat staking — betting the same percentage of your bankroll on every bet that clears your value threshold, regardless of how the last few bets went — removes this bias entirely and lets the actual expected value of your bet selection do the work over time.

A reasonable starting percentage

There's no universally "correct" number, but a common, conservative starting point for value betting is staking between 1% and 3% of total bankroll per bet. This sizing is deliberately small: it means even a longer-than-expected losing streak (which, as covered in earlier posts, is entirely normal for value betting) only erodes a modest portion of total capital, giving the strategy room to play out over the volume it actually needs to prove itself.

Why bankroll size, not stake size, should scale with results

As your bankroll grows through consistent positive expected value over time, your stake per bet should scale up proportionally (since it's a percentage, this happens naturally) — but the percentage itself shouldn't change based on a hot or cold streak. Increasing your percentage after a win because "it's working" reintroduces the same confidence-based bias flat staking is meant to eliminate, just in the opposite direction.

Separating value betting capital from arbitrage capital

If you're running both value betting and arbitrage strategies, it's worth keeping the capital allocated to each separate, at least conceptually. Arbitrage capital needs to be spread across many funded bookmaker accounts, ready to deploy on short notice, with an expectation of small, consistent, guaranteed returns. Value betting capital needs to be sized for variance and drawdowns that arbitrage capital, by design, never experiences. Mixing the two without a clear split makes it hard to judge whether either strategy is actually performing as expected.

Handling drawdowns

A losing streak that eats into 10–20% of your value betting bankroll is well within normal variance for the strategy, especially early on with a smaller sample size. The discipline that matters most here is resisting the urge to abandon the process (or double stake sizes to "win it back") during a drawdown that's statistically unremarkable. Tracking results over a rolling few hundred bets, rather than reacting to any short stretch, is what separates bankroll management that actually protects long-term returns from bankroll management that just feels reassuring in the moment.