Methodology
Why High Win Rate Strategies Still Lose Money
A practical guide to reading win rate with expectancy, profit factor, average win/loss, and max drawdown before trusting a crypto backtest.
Traseq··6 min read
A practical guide to reading win rate with expectancy, profit factor, average win/loss, and max drawdown before trusting a crypto backtest.
Start with a no-code crypto spot strategy, lock the version, run the backtest, and keep the result traceable for comparison.
A high win rate only tells you how often a backtest closed trades with gains. It does not tell you whether the winning trades were large enough to cover the losing trades, fees, slippage, and drawdowns.
Traseq is a no-code crypto spot research workspace, not live trading, exchange execution, a broker terminal, or a performance guarantee. This guide is about reading historical backtest evidence, not predicting future returns.
A strategy with a high win rate can still lose money when the average loss is much larger than the average win. The fast check is expectancy:
Expectancy = (Win Rate x Average Win) - (Loss Rate x Average Loss)
If expectancy is negative, the strategy lost money per trade in the tested sample, even if most trades were winners. Read Win Rate with Profit Factor, average win/loss, trade count, fees, slippage, and Max Drawdown before trusting the result.
Win Rate is the share of closed trades that ended with a gain:
Win Rate = Winning Trades / Total Trades
It is useful context. A low Win Rate can mean long losing streaks and a harder strategy to follow. A high Win Rate can mean frequent small wins and a smoother feeling trade log.
But Win Rate does not measure payoff size. Two systems can both win 60% of trades while producing opposite results:
Same Win Rate. Different economics.
Expectancy combines frequency and size. It answers the question Win Rate cannot answer: what happened on the average trade?
System A: high Win Rate, negative expectancy
90%$50$50010%(0.90 x $50) - (0.10 x $500) = $45 - $50 = -$5 per trade
This system wins nine times out of ten and still lost about $5 per trade in the sample. One large loss erased ten small wins.
System B: low Win Rate, positive expectancy
35%$300$12065%(0.35 x $300) - (0.65 x $120) = $105 - $78 = +$27 per trade
This system was wrong most of the time and still made money per trade in the sample because the average win was much larger than the average loss. That is why trend following vs mean reversion often looks different when you compare Win Rate against payoff size.
Profit Factor compares total gross gains with total gross losses:
Profit Factor = Gross Profit / Gross Loss
A Profit Factor above 1.0 means gross gains exceeded gross losses in the tested sample. Below 1.0 means losses exceeded gains. It is the scale-free cousin of expectancy.
Use this rough read:
Do not treat the table as a forecast. A high Profit Factor can still be fragile if it came from too few trades, one outlier, unrealistic costs, or a tuned window.
The Backtesting Basics demo runs three system templates on real BTC/USDT 1h candles from 2024-09-01 to 2025-05-30. The window includes a trend leg into a BTC all-time high and a later pullback into a range, which makes it useful for seeing why headline Win Rate is not enough.
The highest Win Rate did not produce the highest Return. RSI had the best Win Rate, but the lower-win-rate SMA trend filter had the stronger sample return and a deeper drawdown. Donchian's middle Win Rate still net-lost. The lesson is not that one template is universally better; it is that Win Rate alone did not explain the outcome.
Use this order before giving Win Rate much weight:
For the broader metric review order, read Backtest Metrics: Win Rate, Profit Factor, Sharpe Ratio and the docs guide to reading backtest results.
Want to see the mismatch yourself? Open the Backtesting Basics demo, switch between the three templates, and compare Win Rate against Profit Factor, Return, and Max Drawdown.
Yes. If the average losing trade is much larger than the average winning trade, the strategy can have negative expectancy. A 90% Win Rate with $50 average wins and $500 average losses lost about $5 per trade in the sample.
Expectancy is the average profit or loss per trade in a tested sample. It is calculated as (Win Rate x Average Win) - (Loss Rate x Average Loss). Positive expectancy means the average trade was positive in that sample; negative expectancy means it was not.
Profit Factor is usually more useful for profitability screening because it includes payoff size. Win Rate tells you how often trades were winners; Profit Factor tells you whether total gains exceeded total losses in the tested sample.
Trend-following and breakout strategies often take many small losses while waiting for fewer larger winners. A low Win Rate can be normal for that style, but only if the larger winners are large enough to offset the losses and costs.
Read Win Rate with average win, average loss, expectancy, Profit Factor, trade count, fees, slippage, Max Drawdown, and out-of-sample behavior. Win Rate is behavior context, not a verdict by itself.
| Profit Factor | How to read it |
|---|
Below 1.0 | Losing in the tested sample |
1.0 to 1.3 | Thin and cost-sensitive |
1.3 to 1.5 | Candidate, but needs validation |
Above 1.5 | Stronger sample result, still needs robustness checks |
| Template | Win Rate | Profit Factor | Return | Read |
|---|
| SMA(200) Trend Filter | 18.4% | 1.51 | +31.70% | Lowest Win Rate, highest Return, deepest drawdown |
| RSI Mean Reversion | 49.1% | 1.26 | +21.44% | Highest Win Rate, positive but not the top Return |
| Donchian Breakout | 38.9% | 0.86 | -12.89% | Middle Win Rate, losing sample |
Dec 12, 2025