Guide
How to read your win rate when break-evens exist
The same twenty trades can show 50% or 62.5%. Which convention your journal uses, and why it matters.
Updated · 5 min read
Win rate looks like the simplest statistic in trading: winners divided by trades. It stops being simple the moment you have trades that closed at zero. Depending on how a journal treats them, the same twenty trades can report a win rate of 50% or 62.5%, and both can be called correct. If you compare your numbers across tools, across months or against a strategy’s claims, you need to know which one you are looking at.
Three ways to count a break-even
Take twenty trades: ten winners, six losers and four that closed flat.
| Convention | Formula | Win rate |
|---|---|---|
| Break-evens left out | 10 ÷ (10 + 6) | 62.5% |
| Break-evens in the denominator | 10 ÷ 20 | 50.0% |
| Break-evens counted as wins | 14 ÷ 20 | 70.0% |
The second convention is the most common default, because it is what “winners divided by trades” gives you. But look at what it does: a scratched trade drags the rate down exactly as hard as a full stop-out. A trader who scratches a lot is shown a worse win rate than one who takes the same trades to their stops.
The third convention flatters you. A trade that went nowhere is not a win.
The first one answers the question most traders are actually asking: of the trades that resolved, how many went my way? That is the rule Traquil uses everywhere, with the number of break-evens shown beside it so nothing is hidden.
Expectancy does not care, which is the point
Whichever win rate you use, your average result per trade is the same, because a break-even adds zero to the total either way. If your ten winners average +1.5R and your six losers −1R, then over all twenty trades:
That is why win rate should never be read alone. It is one half of the picture. The other half is how big the wins are compared with the losses, and expectancy in R puts the two together.
When is a trade really break-even?
Two details change the count more than people expect.
Fees.A trade that exits at the entry price but paid $4.50 in commission is a $4.50 loss, not a break-even. Count on the net result, after fees, and a lot of “scratches” become small losses. That is the honest count.
A threshold. If your average winner is $300, is a +$3 exit a win? Many traders would say no, it was a scratch. Some journals let you set a band, for example ±$10, inside which a trade counts as break-even. That is a reasonable choice, as long as you set it once and leave it alone. A band that grows whenever the week looks bad is just a way to edit your statistics.
Break-even win rate: a different thing with a similar name
“Break-even win rate” usually means something else: the win rate at which a strategy neither makes nor loses money, given its reward to risk.
At 1R winners you need more than 50%. At 2R winners, more than 33.3%. At 3R, more than 25%, before fees. Comparing your actual win rate with this number is a quick way to see how much margin your strategy has. Note that it should be compared with the win rate that leaves break-evens out, since the formula assumes every trade is either a full win or a full loss.
A checklist for reading your win rate
- Find out which convention your journal or platform uses.
- Make sure results are net of fees before anything is counted.
- Look at how many break-evens there are; a lot of them is itself worth investigating.
- Read win rate beside average win and average loss, never alone.
- Prefer expectancy in R for the question “is this working?”
The risk of ruin calculator shows the break-even win rate for any reward to risk, and what your actual win rate implies for a drawdown limit.
Education, not financial advice. Rules differ between firms and change over time; your firm’s own rulebook is the authority. Trading carries risk, including the loss of capital.