Bankroll variance & risk of ruin simulator
Set your edge, stake size and time frame, and see the range of bankroll paths your plan will actually produce — median, 5–95% band, drawdowns and risk of ruin included.
Want the live prices this simulates against?
PinnWire is the real-time Pinnacle odds API behind the setup: REST snapshots, SSE drop alerts and a raw WebSocket, with no-vig fair prices on every drop record. Feed a bot the sharp reference price while this page tells you what its edge is likely to do to your bankroll.
What the simulator does
Expected profit is one number, and it hides the entire story. This tool runs a Monte Carlo simulation: it plays your staking plan thousands of times over, betting on independent events that win with the probability your edge implies, and shows the spread of bankroll curves that come out. Each run is a fresh shuffle of wins and losses — which is what variance actually is.
The win chance comes from your edge and your average odds: win chance = (1 + edge) / decimal odds. A 2% edge at 1.95 average odds means a 52.31% chance per bet. Stake sizing then decides how much of the ride your bankroll feels. With percentage staking each stake is a share of the current bankroll, so it compounds; with flat staking the dollar amount stays fixed and the bankroll takes the full arithmetic swings.
How to read the chart
The green line is the median run — half of the simulated runs finished above it and half below, so treat it as what a typical path looks like. The dashed line is the mathematical expectation, the average across all runs, and it sits higher because a handful of lucky runs pull the average up. The gap between the two is variance drag, and it widens as you stake more or bet longer.
The shaded bands show where the middle 50% and the full 5–95% of runs landed. And because a bankroll chart flatters you by only showing the final number, the stat cards underneath report the drawdowns: the typical peak-to-trough dip and the dip that only one run in twenty exceeded. Those are the numbers that decide whether you can actually stay in the seat for the whole period.
Where the inputs come from
The edge is yours to support: it is the gap between your model's probability and the price you take, and an honest number here is the difference between a useful simulation and a fantasy. The average odds should match the market you actually bet into. PinnWire supplies the sharp side of that comparison — a real-time Pinnacle odds API with REST snapshots, a dropping-odds feed for price moves and no-vig fair prices on every drop record, so a bot can compute edge against the market's reference price rather than a stale copy. Details are in the API documentation, and the other free tools build on the same data.
What it does not model
It assumes a constant edge and constant odds on every bet, independent outcomes, and no market frictions. It does not know about limits, market suspensions, line movement, correlated bets, parlays, or pushes and voids. A simulation cannot tell you whether your edge is real — it only tells you what a given edge would produce, and an edge estimate is never a guarantee. No staking plan turns a bad model into a good one.
Bankroll variance & risk of ruin FAQ
What does risk of ruin mean in this simulator?
Risk of ruin is the share of simulated runs in which the bankroll ever dropped below half of its starting value at any point in the period. In flat-stake mode the line is whichever is higher: half the starting bankroll, or one stake. It is estimated from the simulated paths, not a closed-form formula.
Why is the median line lower than the dashed expectation line?
The dashed line is the mathematical mean, which is pulled upward by a small number of very lucky runs. The green median line is the middle run. With compounding stakes the gap between them is variance drag: the drag grows with stake size and with the number of bets, which is why aggressive staking can lose money in the typical run even with a real edge.
How many bets before results look like the expectation?
Slowly, because the standard deviation of your result grows with the square root of the number of bets while the expectation grows linearly. At a 2% edge and average odds of 1.95 your win chance is 52.31%; over 1,000 flat-stake bets the expectation is +20 units but the standard deviation is about 31 units, so roughly a quarter of runs are still behind after a thousand bets.
Should I stake a percentage of my bankroll or a flat amount?
Percentage staking compounds: the stake shrinks as the bankroll shrinks, so a bad run cannot wipe you out in one bet, but the same percentage of a smaller bankroll recovers more slowly. Flat staking keeps the stake constant, so a losing run consumes a growing share of the bankroll and recovery takes a proportionally bigger edge. Run both in the simulator with your own numbers.
How does the simulator turn my edge into wins and losses?
Each simulated bet wins with probability equal to one plus your edge, divided by the average decimal odds you enter. A 2% edge at 1.95 odds therefore wins 52.31% of the time and loses 47.69%. It assumes the same edge and the same odds on every bet, so an edge of 15% at 1.5 odds, which would imply a win chance above 100%, is rejected as impossible.
Does the simulator include the Kelly criterion?
Yes, as a reference. Full Kelly at your edge and odds is edge divided by (odds minus 1): a 2% edge at 1.95 odds is 2.11% of the bankroll per bet. The panel shows your chosen stake as a multiple of full Kelly, and the preset buttons set quarter, half or full Kelly for you. Most practitioners stake below full Kelly because its drawdowns are larger than they expect.
Does anything I type get sent anywhere?
No. The simulation runs entirely in your browser with JavaScript, there is no signup and no API key, and no analytics script runs on the page. The numbers never leave your machine.