Bankroll management: staking, limits and the Kelly criterion
Stoopid Pigeon Editorial· · 8 min read
Most betting advice fixates on which side to back. Bankroll management is the quieter half of the equation: deciding how much to stake, when to stop, and how to keep one bad run from emptying the account. It can't turn a losing game into a winning one — nothing can — but it controls how fast money moves and how likely a player is to run out of it.
This guide covers what a bankroll is, the single habit that matters most, how to size each bet, the Kelly criterion explained without the jargon, where stop-losses and win targets fit, why chasing losses is the classic trap, and the maths reality that ties it all together. The figures here are illustrative — the point is the method, not a promise.
What a bankroll actually is
A bankroll is money set aside specifically for betting or gambling — an amount that can be lost in full without affecting rent, bills or anything else that matters. It is kept separate from everyday finances on purpose. Once money enters the bankroll, it is treated as already spent on entertainment; whatever comes back is a bonus, not income.
Defining it this way removes a lot of bad decisions before they happen. A player who only ever stakes from a ring-fenced, losable pot never has to ask whether a bet is “worth it” against the grocery budget — that question was settled when the bankroll was funded. If reading odds is still new, the guide to getting to grips with betting odds is a useful companion, and the Betting Insights section covers the wider picture.
The single most important habit
If there is one rule that does more than all the staking maths combined, it’s this: set a fixed budget and treat betting as entertainment, not income.
A fixed budget caps the worst case. It turns an open-ended activity — which can otherwise expand to fill whatever money is available — into something with a known cost, like a night out or a streaming subscription. The amount is decided in advance, when the mind is calm, rather than mid-session, when it isn’t.
Everything that follows — unit sizing, Kelly, stop-losses — is a refinement of this idea. None of it works if the underlying budget is money that was needed elsewhere. The wider guides collection covers the rest of the fundamentals.
Unit sizing: flat versus percentage staking
A “unit” is the basic building block of a stake. Rather than betting random amounts, a disciplined bettor stakes in units sized as a fraction of the bankroll.
Flat staking means betting the same amount every time, regardless of how confident the bet feels. A common approach is 1–2% of the bankroll per bet. On a £1,000 bankroll that’s roughly £10–£20 a wager. Flat staking is simple, hard to get wrong, and keeps any single result from doing much damage — the main reason it’s so widely recommended for beginners.
Percentage staking recalculates the stake as a percentage of the current bankroll. Bet 2% and the bankroll grows, the next stake grows with it; if it shrinks, the stake shrinks too. This is self-correcting — it can’t bet money that isn’t there — but it also means stakes drift down during a cold run and up during a hot one, which some bettors find harder to track than a fixed figure.
Neither method changes the odds. Both simply decide how much is exposed on each bet. Smaller units mean slower swings and a longer-lasting bankroll; larger units mean bigger swings in both directions.
The Kelly criterion, explained simply
The Kelly criterion is a formula for sizing a bet in proportion to the perceived edge. The idea: the bigger the advantage on a bet, the larger the share of the bankroll to commit — and the slimmer the edge, the smaller the stake.
In plain terms, the Kelly stake is roughly:
stake fraction ≈ edge ÷ net odds
So a small edge at long odds produces a tiny stake; a large edge at short odds produces a bigger one. The formula is designed to maximise long-run bankroll growth when the edge is real and correctly estimated.
There’s a catch, and it’s a big one. Full Kelly — staking the exact fraction the formula recommends — is high-variance. It produces wild swings and deep drawdowns, and if the edge estimate is even slightly too optimistic, it over-bets and risks serious damage. For that reason most people who use Kelly at all use a fractional version — half-Kelly or quarter-Kelly — staking half or a quarter of what the formula says. That sacrifices a little theoretical growth for a large reduction in variance and a far smoother ride.
The honest caveat: Kelly only helps if a real edge exists. Feed it an imagined edge and it simply sizes losing bets more aggressively. For how to judge whether an edge is real, see value betting explained.
Stop-losses and win targets
A stop-loss is a predetermined point at which a session ends after losses reach a set figure — say, the day’s budget is gone. A win target is the mirror image: a point at which to walk away ahead.
Both exist for the same reason. The moment that does the most damage to a bankroll is usually emotional, not mathematical — the decision to keep going “just one more” after a loss, or to push winnings back in because a run feels hot. Deciding the exits in advance takes that decision away from the version of the bettor least able to make it well. A stop-loss caps the downside of a single session; a win target is the only mechanism by which money actually leaves the table and stays out.
Why chasing losses is the classic trap
Chasing losses — raising stakes to win back what’s gone — is the most reliable way to turn a manageable loss into a serious one. It feels logical in the moment: a bigger bet could erase the deficit in one go. But each bet still carries the same negative edge, so larger stakes simply expose more money to the same disadvantage, faster.
The Martingale-style instinct (double up until a win recovers everything) runs straight into the bankroll wall or the bet limit during any normal losing streak — and losing streaks are normal. One bad run undoes a long string of small recoveries. Chasing doesn’t fix a deficit; it accelerates it.
That's how much any staking system changes the house or bookmaker edge: nothing. Good bankroll management decides how fast money is lost and how likely a bettor is to go broke — it does not make a losing game profitable.
The maths reality
Here’s the part that ties everything together, and it’s worth stating plainly: staking strategy controls variance and risk of ruin — it does not change the edge.
If a game has a negative expected value (as casino games and most bookmaker markets do for the player), no staking pattern makes it positive. Flat staking, percentage staking, Kelly, Martingale — all of them distribute the same underlying disadvantage differently across time. What good staking buys is control: smaller, steadier units lower the chance of ruin and stretch a bankroll over more bets, so the entertainment lasts longer for the same money. What it can never buy is a positive return on a negative-edge game.
This is also why Kelly is not magic. It optimises growth only when the edge is genuinely positive. On a negative-edge bet, the Kelly stake is zero — the formula’s own advice is don’t bet. For where these edges come from in the first place, the house edge explainer is the place to start, and baccarat tips shows the same logic applied to one specific game.
Responsible-play tools
Most regulated operators provide built-in controls, and they work alongside personal limits:
- Deposit limits — cap how much can be added over a day, week or month.
- Time limits / session reminders — flag how long a session has run.
- Loss limits — cap losses over a set period.
- Cooling-off and self-exclusion — temporarily or permanently block access to an account.
For independent support and self-assessment tools, BeGambleAware is a widely used resource. Setting these limits when calm — not mid-session — is the same principle as a pre-set budget, enforced by the platform rather than willpower.
Quick reference
- A bankroll is money set aside only for betting that can be lost in full.
- Treat it as entertainment, not income — a fixed budget is the rule that matters most.
- Flat staking of about 1–2% per bet is simple and durable; percentage staking self-adjusts to the bankroll.
- The Kelly criterion sizes a stake to the edge (roughly edge ÷ net odds); most people use half- or quarter-Kelly to cut variance.
- Set a stop-loss and a win target before the session, not during it.
- Don’t chase losses — bigger stakes only expose more money to the same edge.
- No staking system changes the house or bookmaker edge; it only controls how fast money is lost and the risk of ruin.
Frequently asked questions
What is a betting bankroll?
It's money set aside specifically for betting or gambling — an amount that can be lost in full without affecting essentials like rent or bills. Keeping it separate from everyday finances is the point: once funded, it's treated as already spent on entertainment.
How much should I stake on each bet?
A common flat-staking approach is about 1–2% of the bankroll per bet. Smaller units mean slower swings and a longer-lasting bankroll; larger units mean bigger swings in both directions. The figure is a guideline, not a rule.
What's the difference between flat and percentage staking?
Flat staking bets the same amount every time. Percentage staking recalculates the stake as a percentage of the current bankroll, so stakes shrink during a cold run and grow during a hot one. Percentage staking is self-correcting; flat staking is easier to track.
What is the Kelly criterion?
It's a formula that sizes a bet in proportion to your perceived edge — roughly stake fraction equals edge divided by net odds. The bigger the genuine advantage, the larger the suggested stake. It's designed to maximise long-run bankroll growth when the edge is real and correctly estimated.
Why do people use half- or quarter-Kelly?
Full Kelly is high-variance and punishes any overestimate of the edge with deep drawdowns. Staking half or a quarter of what the formula recommends gives up a little theoretical growth in exchange for a large reduction in variance and a smoother ride.
Can good staking make a losing game profitable?
No. Staking strategy controls variance and the risk of ruin, but it does not change the house or bookmaker edge. On a negative-edge game, every staking pattern distributes the same disadvantage differently — none turns it positive.
Does the Kelly criterion always help?
Only if a genuine edge exists. On a negative-edge bet the Kelly stake is zero — the formula's own advice is not to bet. Feed it an imagined edge and it simply sizes losing bets more aggressively.
What is a stop-loss and a win target?
A stop-loss is a predetermined point at which a session ends once losses reach a set figure. A win target is the point at which to walk away ahead. Deciding both in advance removes the in-the-moment decisions that tend to cause the most damage.
Why is chasing losses such a problem?
Raising stakes to win back losses exposes more money to the same negative edge, faster. Doubling-up systems run into the bankroll wall or bet limit during a normal losing streak, and one bad run can undo a long string of small recoveries. Chasing accelerates a deficit rather than fixing it.
What responsible-play tools can help?
Most regulated operators offer deposit limits, time and session reminders, loss limits, cooling-off periods and self-exclusion. Independent resources such as BeGambleAware provide support and self-assessment tools. Setting these limits while calm works the same way as a pre-set budget.