Kelly bet-sizing
Find the growth-optimal fraction of capital to risk on a repeatable bet from your win rate and your reward-to-risk — and see why full Kelly is so violent that disciplined bettors use a fraction of it.
Runs entirely in your browserRuns entirely on your device — nothing you type is sent anywhere. Educational only, and not investment advice.
How to use this calculator
Each step names a control you will find on screen above.
- Win probability
How often the bet wins, as a percentage. This is the number you almost always overestimate; a backtest flatters it badly.
- Reward-to-risk
How much you win per unit risked when it works — 2 means you make twice what you would lose. Kelly needs both this and the win rate; a high win rate at a poor payoff can still have no edge.
- Read the fraction
Full Kelly is the theoretical growth-maximising stake; half Kelly is what most people who have lived through full Kelly actually use. A zero or negative result means the bet has no edge — stake nothing.
Worked example: 55% win rate at 2:1
A bet that wins 55% of the time and pays twice the amount risked when it wins.
What to enter
- Win probability
- 55%
- Reward-to-risk
- 2.0
What it shows you
- Edge per ₹1 risked
- ₹0.65
- Full Kelly
- 32.5% of capital
- Half Kelly
- 16.25% of capital
Where this is taught
A calculator gives you a number. These explain what the number means and when it misleads you.
- Technical Analysis13 minHow much to bet: Kelly, half-Kelly, and why full Kelly ruins peopleThere is a mathematically optimal fraction to risk on a favourable bet. It is larger than anyone can tolerate, and the reason why is worth understanding properly.
- Technical Analysis11 minFrom signals to a systemThe six components every complete system needs, and why most people never write theirs down.
- Technical Analysis12 minRisk of ruin: the arithmetic that decides if you surviveA positive-expectancy system can still destroy an account. The variable that decides it is size — and the relationship is far less forgiving than it looks.
- Technical Analysis13 minThe average trade you did not getA strategy report says 200 trades and an average of +2.5% each. You compound that and get a number nobody has ever earned, and the report’s own equity curve ends far below it. Nothing has been faked. The average trade is simply not a figure you are allowed to compound.
- Technical Analysis11 minThe case for slowing downMost retail traders operate on timeframes that maximise noise, cost and stress while minimising their actual advantage. The weekly chart fixes all three.