Flat Staking vs Percentage Staking: Which Fits

Staking is the part of betting that decides how much of your bankroll goes on each opinion. It does not change whether a bet is good. It changes how quickly a good bet compounds and how badly a bad run hurts. Two schemes dominate the amateur and semi-professional world: flat staking, where every bet gets the same amount, and percentage staking, where the amount scales with the current bankroll.

Flat staking, defined

Flat staking means one unit per bet, where a unit is a fixed slice of your starting bankroll, usually one or two percent. If you start with a thousand and define a unit as ten, every bet is ten until you decide to redefine the unit. The stake does not respond to a winning streak or a losing streak, which is exactly the point.

The advantage is measurement. When every bet is the same size, your profit and loss is a clean sum, your ROI is easy to compute and one bad bet cannot distort the picture. The disadvantage is that the scheme ignores bankroll growth and shrinkage. A unit that was two percent of your roll at the start becomes one percent after a good month and three percent after a bad one.

Flat staking also ignores confidence. A bet you consider marginal and a bet you consider strong get the same amount, which is either disciplined or wasteful depending on how much information your confidence actually carries.

Percentage staking, defined

Percentage staking recalculates the stake as a fixed fraction of the current bankroll before each bet. If you stake two percent, the first bet on a thousand is twenty, and the next bet is two percent of whatever is left. The stake shrinks automatically during a downswing and grows during an upswing.

This is the scheme that survives long losing runs. Ten losses in a row at a flat stake of ten percent would leave nothing; at two percent of the current balance, the same run leaves about eighty two percent of the roll intact. The cost is that recovery is slower, because the stakes are smaller exactly when you need them to be larger.

Percentage staking also makes the unit concept slippery. A unit is no longer a fixed number, so comparing your results to someone else's unit-based record becomes meaningless, and your own records need the actual stake stored rather than just the unit count.

Side by side

The two columns are close on a mild sequence and diverge sharply on an extreme one. That is the whole trade: percentage staking gives up some upside in exchange for surviving the tail.

Variance and drawdown

Drawdown is the distance from a peak bankroll to the following trough. It is the number that decides whether a staking plan is survivable in practice, because it is the number you feel. A scheme that produces a thirty percent drawdown once a year is comfortable. A scheme that produces a fifty percent drawdown is technically fine and psychologically brutal, and most people abandon their plan somewhere in the middle of it.

Flat staking with a small unit produces shallower drawdowns in absolute terms for the same edge, because the stake does not grow with the roll. Percentage staking produces proportionally similar drawdowns regardless of bankroll size, which makes the experience predictable. Predictability is worth more than it sounds when you are three weeks into a losing run.

The psychological difference

Flat staking removes a decision. There is no argument with yourself about whether this bet deserves more, because the answer is always the same number. That is a real benefit for anyone prone to chasing or to overconfidence after a good week.

Percentage staking introduces a decision every time, and the decision arrives at the worst possible moment, right after a loss, when the rational answer is to bet less and the emotional answer is to bet more to get it back. Bettors who choose percentage staking need to automate the calculation, or they end up with a hybrid that is really just flat staking with extra steps.

When each one fits

• Flat staking fits beginners, anyone recording results for the first time, and anyone who wants a clean ROI figure.
• Flat staking fits a bankroll that is not expected to grow, such as a fixed seasonal budget.
• Percentage staking fits a long horizon, where the bankroll is meant to compound over hundreds of bets.
• Percentage staking fits anyone who has survived a drawdown badly and wants the plan to enforce discipline.
• Neither fits a bettor who changes the unit size after a loss, because that is not a scheme at all.

There is no universal answer, but there is a wrong answer: mixing the two without writing down which one you are using. A record where some bets are flat and some are percentage cannot be analysed, and analysis is the only reason to keep records.

A hybrid that most people should use

Percentage staking with a floor and a cap gets most of the safety and most of the simplicity. Stake a fixed percentage of the current roll, but never less than a minimum unit and never more than a maximum unit. The floor stops the stake from becoming trivial after a bad run. The cap stops a hot streak from turning into an oversized position that one loss can erase.

Set the floor and cap in advance, in writing, and change them only at a scheduled review, never in the middle of a session. The point of a staking plan is that it makes the decision before you are emotional. A plan you renegotiate at midnight is not a plan.