Two different rules share the name
The following illustrative labels distinguish the calculation timing; they do not recommend either rule:
| Illustrative label | Calculation timing |
|---|---|
| Dynamic-bankroll 1% | Recalculate one percent from the eligible bankroll at the declared snapshot |
| Fixed-base 1% | Calculate one percent from a frozen reference balance until a scheduled reset |
With a GBP 500 reference bankroll, both begin at GBP 5. After a GBP 5 loss, dynamic-bankroll staking produces GBP 4.95, while fixed-base staking remains GBP 5.
The difference grows after a long sequence. Always publish the denominator, reset timing, rounding, and treatment of open bets.
What the rule does
| Claim | Assessment |
|---|---|
| One loss is limited to about 1% of the declared bankroll | True under full-loss settlement and no extra liability |
| Ten open 1% bets risk only 1% | False; nominal open stake is about 10% before dependence |
| The rule proves the bet has value | False |
| The rule is mathematically optimal | Unsupported without an objective and probability model |
| A 100-bet bankroll cannot be lost | False; losses, exposure, top-ups, and infinite play invalidate that shortcut |
Add an open-exposure calculation
In an illustrative exposure check, six 1% back bets are open from the same GBP 500 snapshot. Nominal stake exposure is:
Illustrative nominal open stake = 6 * GBP 5 = GBP 30, or 6% of the snapshot bankroll
If a market can create liability beyond the displayed stake, use maximum loss instead. If bets share a team, match, or tournament outcome, a simple sum does not describe dependence.
Why one percent is not universal
The Kelly framework links stake to estimated probability and payoff, while modified Kelly research shows that probability uncertainty can materially reduce model-derived fractions. Drawdown-constrained work makes risk tolerance an explicit model input rather than assuming one fixed fraction suits everyone (Busseti, Ryu and Boyd).
Those sources do not validate the 1% convention. They explain why a universal percentage cannot be inferred from stake arithmetic alone.
Affordability still comes first
One percent of an unaffordable or borrowed bankroll remains unaffordable. Build the external household budget first using accurate income and outgoings; MoneyHelper provides a public budgeting process.
Define reset and rounding behavior
Write whether one percent is calculated before each attempt, after each settlement, or from a scheduled snapshot. If two bets are placed before either settles, they should normally use the same snapshot under an after-settlement rule. Do not let database row order determine stake size after the fact.
Set a rounding rule and minimum-stake response. For a GBP 73.40 bankroll, one percent is GBP 0.734. Rounding to GBP 0.73 and rounding to GBP 0.75 are different policies. If an operator minimum exceeds the calculated stake, the reproducible action is pass unless a predeclared rule says otherwise; rounding up silently increases the intended exposure.
Next step
Use Percentage Staking for the next part of this topic.
Continue learning
- Next guide: Variance in Betting
- Related guide: Level Stakes vs Variable Stakes
Assumptions and limitations
Examples assume losing back bets forfeit only the displayed stake. The rule does not cover exchange liability, correlated positions, multiple accounts, unsettled funds, or changing personal circumstances. It can standardize exposure, but it cannot guarantee survival, profit, or control over harmful gambling.
GambleAware
GamCare
Gambling Therapy
GAMSTOP