The formula
win rate = winning trades ÷ decisive trades × 100
Note the denominator: decisive, not total.
The edge case: trades that finished at zero
A trade closed at exactly break-even is neither a win nor a loss. Three treatments exist, and they give different answers:
- Count it as a loss — punishes scratching a trade that stopped working, which is usually good behaviour.
- Count it in the denominator only — mathematically the same as calling it a loss.
- Exclude it entirely — treats it as a decision that produced no result.
Ours excludes it: wins and losses are counted separately and the rate is computed over their sum, so a scratched trade moves neither. On an account that regularly exits at break-even the difference runs to several percentage points, which is more than enough to change how a month reads.
Note that "exactly zero" is rarer than it looks once costs are included — see net P&L. A trade that closed flat on price still lost the commission, and with net figures it is a small loss rather than a scratch.
Why the number alone says nothing
Win rate is one of two halves. The other is the ratio between average win and average loss, and only together do they determine whether a method makes money:
| Win rate | Win : loss | Result |
|---|---|---|
| 70% | 1 : 3 | loses money |
| 30% | 4 : 1 | makes money |
| 50% | 1 : 1 | break-even before costs, loses after |
The threshold that connects them is the break-even win rate — see expectancy. Below it you lose regardless of discipline; above it, a low win rate is not a problem to fix.
Where it is genuinely useful
As a psychological forecast. A 35% win rate means losing runs of six or seven are ordinary rather than exceptional. Knowing that in advance is the difference between following a plan and abandoning it in the middle of a normal sequence.
As a change detector. A win rate that shifts sharply while your method has not is worth investigating — usually it points at market conditions or at execution, not at strategy.
Where it misleads
Chasing a higher win rate is the most common way traders damage a working method: cutting winners early raises the percentage and lowers expectancy at the same time. The number improves, the account does not.