Most bankrolls die from one thing: betting like certainty exists.
Kelly Stake Sizer
If you actually have an edge, the next question is: how much should you risk?
Kelly sizing uses odds and your win probability to recommend a bankroll-based stake size—so you stop swinging too hard and start sizing like a process.
Kelly Stake Sizer
Enter your bankroll, the odds, and your estimated win probability. The calculator returns a suggested stake size based on Kelly staking logic.
One sentence that matters: Kelly doesn’t make a bad bet good. It sizes a bet after you already have a credible edge.
Features
Core outcomes
Turn edge into a stake size.
- Suggested bankroll fraction to wager
- Stake amount based on bankroll input
- Built-in “no edge” sanity check
Decision impact
Reduce overbetting and emotional sizing.
- Sizes based on probability, not confidence
- Helps smooth swings across many bets
- Encourages consistent bankroll rules
Process guardrails
Protects you from common sizing mistakes.
- Shows how sensitive size is to probability
- Prevents “double after loss” behavior
- Pairs cleanly with EV / fair line tools
Decision logic
Kelly is powerful—and that’s exactly why it can be dangerous when used carelessly.
Use these rules to keep it as a sizing tool, not a bankroll grenade.
Example (hypothetical): You estimate a small edge, but you know your probability is noisy—so you treat the Kelly output as a maximum, then size smaller for smoother swings.
If X → then Y rules
- If your edge is uncertain → treat the output as a ceiling, not a command.
- If the suggested stake is “huge” → your probability input is likely optimistic; stress test it.
- If your bankroll swings emotionally affect you → consider using a smaller fraction of the Kelly output (many people do).
- If you’re betting correlated outcomes → Kelly assumptions break; size smaller because variance is higher than it looks.
- If you can’t justify your win probability → do not use Kelly sizing; you don’t have a stable edge input.
Real-world scenarios
- “I’m hot right now” streak: Kelly stops you from scaling stakes purely on emotion.
- New model or strategy: use smaller sizing until your probability estimates are proven in real results.
- Same-game parlays: correlation destroys neat assumptions—Kelly outputs can be misleadingly large.
- Multiple books / line shopping: better odds increase Kelly size legitimately by increasing edge.
Practical truth: The hardest part of Kelly isn’t the formula. It’s not lying to yourself about probability.
Expanded math explanation
The Kelly idea
Kelly staking chooses a fraction of bankroll that (in theory) maximizes long-run growth when you have an edge.
For decimal odds, a common form is:
f* = (b×p − q) ÷ b
Where f* is the bankroll fraction to wager, p is your win probability,
q = 1 − p, and b is net odds (decimal odds minus 1).
What the parts mean
- p = your chance to win (your estimate)
- q = your chance to lose
- b = profit per unit staked (net odds)
- f* = recommended fraction of bankroll
If b×p − q is not positive, the formula yields zero or negative—meaning no bet (no edge).
Why Kelly can feel “too aggressive”
| What’s happening | Why it matters | What to do |
|---|---|---|
| Your probability estimate is optimistic | Kelly size increases fast with p | Stress test p; be conservative |
| Variance is high | Drawdowns feel brutal even with an edge | Consider smaller sizing |
| Bets are correlated | True risk is higher than assumed | Reduce stake sizes meaningfully |
Mini glossary
- Kelly fraction
- The percentage of bankroll the formula suggests wagering.
- Edge
- Your advantage over the break-even probability implied by odds.
- Net odds (b)
- Profit per unit staked (decimal odds minus 1).
- Variance
- Short-term swings that can cause losing streaks even with good bets.
- Correlation
- When outcomes are linked; it increases real risk and breaks neat sizing assumptions.
Takeaway: Kelly is math for sizing, not a license to bet bigger. If the inputs are shaky, the output is shaky.
Behavioral traps Kelly helps prevent
1) Chasing losses
Kelly uses bankroll-fraction rules. It doesn’t care that you lost yesterday, so it won’t “double you up” emotionally.
2) Overconfidence sizing
“I feel great about it” is not a probability estimate. Kelly forces the confidence into a number you can sanity-check.
3) Hot-hand illusion
Winning streaks make people believe they’re “due to keep winning.” Kelly centers you on edge and price instead.
4) Pain avoidance
After losses, people bet smaller even when they have value. Kelly helps keep sizing consistent across emotions.
5) Precision cosplay
Using a formula can make bad inputs feel scientific. If you can’t justify p, you’re not doing Kelly—you’re doing decoration.
How to use the Kelly Stake Sizer
- Enter your bankroll (the amount you’ve dedicated to betting, not bill money).
- Enter the odds for the wager.
- Enter your win probability as realistically as possible.
- Read the suggested stake and the bankroll fraction.
- Stress test your probability by moving it slightly up/down to see sensitivity.
- Decide your policy: treat the output as a max size if you prefer smoother swings.
- Be conservative with probability. Optimism makes Kelly dangerously large.
- Shop lines. Better odds increase edge and improve sizing legitimately.
- Don’t Kelly parlays blindly. Correlation and pricing complexity can mislead the output.
FAQ
What is Kelly staking?
Kelly staking is a bankroll sizing method that uses your estimated probability and the odds to suggest a fraction of bankroll to wager,
aiming (in theory) to maximize long-run growth when you have an edge.
Is Kelly staking safe?
It can be risky if your probability estimate is wrong or your bets are highly volatile/correlated. Kelly is math, not certainty.
Many bettors use smaller sizing rules to reduce drawdowns.
What if the calculator suggests 0?
That typically means you don’t have a positive edge at the given odds with your probability estimate. In other words: no bet by Kelly logic.
Do I need accurate probabilities for Kelly to work?
Yes. Kelly is extremely sensitive to probability. Overestimating your win rate is the fastest way to overbet.
Can I use Kelly for parlays?
You can, but be careful: parlay probabilities and correlation are tricky, and sportsbook pricing can add hidden margin.
If you use Kelly on parlays, treat the output as a strict maximum.
How does Kelly relate to EV?
EV tells you whether a bet is good (+EV or not). Kelly tells you how big to bet after you’ve identified a real edge.
Responsible use
Bankroll tools help with discipline, but they don’t remove risk. Bet within limits you can afford to lose,
avoid chasing losses, and follow your local laws. If you’re using Kelly outputs to justify oversized bets, pause and re-check your probability assumptions.
The sizing tool is only as honest as the inputs.