Your bets
How each staking plan did
| Staking plan | Median end bank | In profit | Bust | Worst drop |
|---|
Every plan is run on the same random results, so the differences come from the staking alone. Kelly stakes the fraction (p × odds − 1) ÷ (odds − 1) of your current bank, where p is your win chance; with no edge it stakes nothing. Chase losses doubles the stake after each loss and goes back to the base stake after a win (a martingale), capped at what is left in the bank. Bust means the bank fell under 5% of where it started. Worst drop is the typical (median) biggest fall from a high point. Paper betting only: it uses no money and does not predict any real result.
- Maths reviewed by Josh Miller
- Tested by Diana Westfield
Say you back a team at 2.10 and you’re right 49% of the time. That’s a real edge, about 3% per bet. Is it enough? This betting strategy simulator plays out thousands of seasons of those bets on paper and shows what happens to your bank under six staking plans: flat stakes, a fixed percentage of the bank, full, half and quarter Kelly, and doubling up after losses. It’s free and uses no money. Change the odds and the win rate and hit Run.
What the sports betting simulator does
Each run deals out a season of bets with your win chance, then settles every staking plan on the same results. Same luck, different staking. That’s the fair way to compare them, and it’s why the table can surprise you. With the defaults (bank £1,000, odds 2.10, 49% wins, 500 bets) you’ll typically see flat £20 stakes end in profit in about 72% of seasons, full Kelly suffer the deepest drops (a typical fall of over 45% from its peak), quarter Kelly barely wobble, and chasing losses go bust in about four seasons out of five, even with an edge.
The lines on the chart are twenty of those seasons with flat stakes. Some go up nicely. A few drift down for hundreds of bets before turning. That’s not bad luck in a special sense, just ordinary variance.
Comparing betting systems
| Plan | How it stakes | Good for |
|---|---|---|
| Flat stakes | The same amount every bet | Simple records, steady swings |
| % of bank | A fixed share of the current bank | Grows with wins, shrinks with losses, never quite busts |
| Full Kelly | (p × odds − 1) ÷ (odds − 1) of the bank | Maximum growth if your win rate is exactly right |
| Half / quarter Kelly | Half or a quarter of that | Most of the growth, far less pain, safer if you overrate yourself |
| Chase losses | Double after a loss, reset after a win | Nothing. It’s a martingale: small steady wins, then a wipe-out |
Kelly is the interesting one. It’s optimal only if your win chance is right. Most bettors overestimate their edge, and full Kelly on an imagined edge is a fast way down. That’s why serious bettors usually stake a fraction of it. Set the win rate a couple of points lower than you believe and watch what full Kelly does.
Sports betting strategy starts with the edge
Try one more run: set the win rate to the break-even figure shown under the box (47.6% at 2.10). Every plan now loses or treads water, and Kelly refuses to bet at all. No staking plan creates an edge. If you don’t know your real win rate yet, track your bets for a few hundred selections first; the CasinoCalcs bet tracker does it free. For the theory, see the Kelly criterion calculator, the betting variance calculator and the risk of ruin calculator. For roulette-style systems like Martingale and Fibonacci, there’s a separate betting systems simulator.
A simulator can’t tell you your real edge. Most bettors don’t have one. BeGambleAware.org offers free, confidential support.