Bankroll Management for Value Bettors: Staking Plans That Survive

Here is an uncomfortable truth: a bettor with a genuine 10% edge and reckless staking will go broke more often than a bettor with a 3% edge and disciplined staking. Variance doesn't care about your model. At overlay prices — typically 4/1 to 12/1 — losing runs of fifteen or twenty bets are not bad luck; they are a mathematical certainty you must plan for.

Level stakes: the honest baseline

Betting one unit on every flagged pick is simple, transparent, and easy to audit — it's how our public ledger settles every overlay. Its weakness is that it doesn't scale stakes to edge size or bankroll, so it leaves profit on the table in strong spots and doesn't shrink when the bank does.

For most people following a service, level stakes at 1–2% of bankroll per bet is the right starting point. It makes a twenty-bet losing run — which will happen — a 20–40% drawdown rather than a funeral.

Kelly and why you should divide it

The Kelly criterion computes the stake that maximises long-term bankroll growth: stake fraction = edge / (odds − 1). A 55% overlay at 9/2 says stake around 12% of your bank — which is insane in practice, because Kelly assumes you know your edge exactly. You don't. Model probabilities carry error, and full Kelly turns estimation error into ruin.

Serious operators use fractional Kelly — a quarter or an eighth of the computed stake. Quarter-Kelly keeps most of the growth rate while cutting drawdowns dramatically. If the maths above says 12%, quarter-Kelly says 3% — aggressive but survivable.

The rules that actually save you

Whatever plan you choose: define your bankroll as money you can lose entirely without consequence. Never chase — the next bet's stake is set by the plan, not by the last result. Recalculate percentage stakes weekly, not per bet. And keep a written record of every bet at the price you took, because self-deception is the most expensive leak in betting.