The formula, and the ambiguity inside it
win rate = wins ÷ (wins + losses) × 100
Simple, except for one question nobody agrees on: what happens to trades that closed at break-even?
Fifty-four wins, sixty-six losses and five scratches gives 45.0% if the scratches are excluded and 43.2% if they are counted. Neither is wrong; they answer slightly different questions. What is wrong is quoting a figure without saying which convention was used — which is why this calculator shows both rather than picking one for you.
The gap widens for anyone who routinely moves stops to break-even. A trader with twenty scratches in a hundred trades is choosing between two numbers several points apart.
The number is meaningless alone
This is the part worth internalising: a win rate cannot be good or bad by itself.
| Win rate | Reward per unit risked | Result |
|---|---|---|
| 30% | 1 : 4 | Profitable |
| 30% | 1 : 2 | Loses money |
| 70% | 1 : 0.3 | Loses money |
| 70% | 1 : 0.5 | Profitable |
Every ratio comes with a threshold attached, and the only useful reading of your win rate is the distance between it and that threshold. That distance is your edge. The calculator computes it once you enter the reward figure, and the break-even win rate page covers what happens when trading costs are added on top.
This is also why "improve your win rate" is unhelpful advice on its own. Taking profits earlier raises the win rate and lowers the reward per trade, usually by more — the number goes up and the account goes down.
Why high win rates attract the wrong attention
A 70% or 80% win rate is easy to produce and easy to sell. Widen the stop, take profits quickly, and most trades close green. The losses that remain are large, infrequent, and arrive together.
The equity curve from such a system looks excellent for months, which is exactly long enough to attract capital and confidence. What it hides is the shape: many small wins funding one severe loss, which is the same payoff structure as a martingale arrived at by a different route.
When you see a win rate quoted without the reward ratio and the largest loss beside it, the omission is usually the point.
Sample size
A win rate from thirty trades is a wide guess. At an underlying 45%, thirty trades produce observed rates anywhere between roughly 27% and 63% purely by chance — which is the entire distance between an excellent system and a failing one.
The expectancy calculator puts a confidence range on this directly. Until the sample reaches a few hundred trades, treat your win rate as an estimate with a wide error bar rather than as a property of your trading.