The reason win rate gets its own measure is that raw profit means very little without context. A player up a certain amount could have ground it out over months or spiked it in a single session, and those two stories point to completely different futures. A win rate fixes profit against the volume of play, turning a lump sum into a rate that can be compared, projected and trusted - once the sample behind it is big enough.
01Cash games: big blinds per hundred hands
Cash-game win rate is expressed in big blinds won per hundred hands, almost always written as bb/100. Counting in big blinds rather than currency keeps the figure comparable across stakes: a rate of five bb/100 means the same thing whether the big blind is small change or a serious sum. It measures pure edge, independent of how high the player is sitting.
The mechanics of that unit - how it is calculated, what a worked example looks like, and how it converts into an hourly rate - are covered in detail in the explanation of what bb/100 means. The headline point here is that a positive bb/100 is a winning player and a negative one is not, with the size of the number describing how strong the edge is.
02Tournaments: ROI and in-the-money percentage
Tournaments cannot use bb/100 in any useful way, because chips are not money and most entries end with nothing. Instead they are measured by return on investment - total profit divided by total buy-ins, shown as a percentage - usually alongside the in-the-money percentage, or ITM, which is simply how often the player reaches a paid finish.
These two figures tell different halves of the story. A high ITM with a flat ROI describes a cautious player who cashes often but rarely runs deep, while a modest ITM paired with a strong ROI describes someone who busts more but goes big when they connect. The full picture of how tournament returns are calculated sits in the guide to measuring MTT ROI.
A practical reminder: cash and tournament win rates are not interchangeable. A player can post a healthy bb/100 in cash games and a losing ROI in tournaments, because the skills, payout shapes and variance levels are genuinely different. Each format needs its own measured rate.
03What counts as a good rate
There is no universal threshold, because a good win rate is relative to the competition. In lower-stakes online cash games, a solid regular might hold a bb/100 somewhere in the low single digits up to the low teens; that edge typically compresses as the stakes rise and the opponents sharpen. In tournaments, even a few percent of positive ROI is a genuinely good result once it is built on a large enough sample, because the variance is so severe that small edges are hard to confirm.
The recurring theme is that the same raw number looks excellent against soft games and ordinary against tough ones. This is why realistic cash-game win rates are discussed in the context of the stake rather than as a single magic figure - the level decides what good actually means.
| Game type | Break-even | Solid | Strong |
|---|---|---|---|
| Micro-stakes cash | 0 bb/100 | 4-8 bb/100 | 8+ bb/100 |
| Low-stakes cash | 0 bb/100 | 3-6 bb/100 | 6+ bb/100 |
| Mid-stakes cash | 0 bb/100 | 2-5 bb/100 | 5+ bb/100 |
| Tournaments (ROI) | 0% | 5-15% | 15%+ |
The pattern is consistent: the softer and smaller the stakes, the higher a realistic win rate, and the figure compresses as the games toughen up. Cash rates are read in big blinds per 100 hands, while tournaments switch to ROI because their returns arrive in rare, large lumps rather than a steady drip.
04Sample size: when the number can be trusted
A win rate is only as honest as the sample behind it. Cash-game rates can take tens of thousands of hands to converge on their true value, and tournament ROI often needs thousands of entries before the noise fades. Below those volumes, a win rate measures variance far more than skill, which is why a few good or bad sessions reveal almost nothing about a true edge.
Treating the number with the right amount of doubt is the difference between using it and being fooled by it. A win rate over a tiny sample is a snapshot of luck; the same figure over a large one is a measure of ability. Seeing a rate next to the swings that surround it - the realm of poker variance - is what stops a short-term result from being mistaken for a settled truth.
Same true win rate, three sample sizes. Dashed line is the expected rate. Over a few thousand hands the result swings wildly and can sit below break-even; only as the sample grows toward six figures does the line settle into something worth trusting.
05What affects a win rate most
Two players at the same stake can post very different rates. The biggest levers are rarely raw card-playing skill alone:
- Game selection. Sitting in softer games lifts a win rate more reliably than almost any in-game adjustment.
- Rake. The house take quietly drags every rate down, and it bites hardest at the lowest stakes.
- Table count. Adding tables raises hourly volume but can thin out the per-table win rate if attention spreads too far.
- Stake level. The same player's rate compresses moving up, because the opponents get tougher.
- Player-pool strength. A soft regional or off-peak pool inflates rates that a tougher pool would erase.
- Positional discipline and tilt control. Leaks out of position and emotional play erode an edge that the raw strategy would otherwise earn.
- Volume. Not a skill, but it is what turns a noisy number into a trustworthy one over time.