Your game
The edge you’re assuming. We treat it as given to show the swings around it — your real rate is itself a range.
How long a stretch to simulate.
How wild the swings are. ~100 is typical for online 6-max No-Limit Hold’em.
The range of outcomes
Where you most likely finish (95% of runs)
−3,698 to +8,698
big blinds over 100,000 hands — centred on +2,500.
- Chance you end in profit78%
- Typical worst downswing−2,758 bb · 27.6 buy-ins
- Rough-run downswing (1 in 20)−5,396 bb · 54.0 buy-ins
- Share of the ride below break-even31%
The range of outcomes still crosses zero, so a break-even or losing result over this sample is entirely consistent with being a long-term winner. That is variance, not a verdict on your game.
How this is calculated
Your cumulative result after a sample is modelled as a sum of independent 100-hand blocks, each with mean equal to your win rate and standard deviation equal to the value you enter. After b blocks that sum is approximately normal with mean win rate × b and standard deviation SD × √b — which gives the shaded 70% and 95% bands and the headline finishing range.
The downswing depths and the time spent below break-even have no tidy formula, so we run 5,000 simulated samples and read the results off them (sampling the path in up to 1,000 steps so a long sample stays fast — this very slightly understates the deepest swings). It assumes roughly normal results and a fixed, known win rate; real poker has neither perfectly, so treat the numbers as an honest guide to the shape of variance, not a prophecy.
Why the result is a cloud, not a line
Two players with the exact same skill, playing the exact same number of hands, can finish thousands of big blinds apart. The shaded fan above is that spread: the middle line is your expected result, and the bands are where most runs actually land. The longer the sample, the wider the absolute spread grows — variance doesn’t shrink, it just gets swamped by skill as a share of the result.
The number that matters most for your bankroll isn’t the finish — it’s the downswing depth. A 2.5 bb/100 winner can still run 30+ buy-ins below their peak on an ordinary sample. That’s not a leak; it’s the cost of admission. Plan your bankroll for the rough runs, not the average one.
Notice how much of the ride sits below break-even even when you end up ahead. That’s exactly why a short losing stretch proves nothing — and why StackWise won’t call you a winner off a small sample. See how wide your own win-rate range is.
See your real variance, session by session.
StackWise tracks your actual swings and pairs every figure with a confidence range — so you know the difference between a downswing and a leak. Free to start, no account.
Sources & method
Cumulative results are modelled as a normal random walk in 100-hand blocks (mean = win rate, SD = standard deviation per 100), with downswing depth and break-even time from a 5,000-run Monte Carlo. Method and typical standard-deviation values from Primedope, Phil Galfond, and ThePokerBank.