A 94% WIN RATE THAT
TELLS YOU ALMOST NOTHING.
Win rate is the most quoted number in retail trading and the least informative one. You can raise it to almost anything you like by moving your take-profit closer and your stop further away. What you cannot do is raise it without paying for it somewhere else.
Rather than argue that abstractly, here are our own numbers. Four of the seven strategies we ship win more than nine trades in ten — and every one of them loses about eight times as much on a loser as it makes on a winner.
| Strategy | Win rate | Average win | Average loss | Profit factor |
|---|---|---|---|---|
| Mean Reversion | 94% | £47 | £397 | 1.73 |
| Aggressive Survivable | 93% | £48 | £401 | 1.66 |
| Trend Follower | 93% | £50 | £409 | 1.54 |
| Swing Rider | 72% | £121 | £249 | 1.27 |
| Stretch Fade | 54% | a normal-looking payoff | 1.14 | |
| Momentum Rejoin | ~25% | wins are many times the losses | 1.23* | |
Measured Feb 2023 – Jul 2026 at shipped settings; corrected 28 July 2026 after four figures were found to have been measured against the wrong configurations. *Momentum Rejoin was retired on 29 July 2026 — never validated, and re-testing moved it further from significance rather than closer. It is kept in this table because the shape of its payoff is the point being made.
Read that table top to bottom and the win rate falls from 94% to 25% while the payoff improves at almost every step. The two columns are trading against each other, which is exactly why neither means much alone.
The arithmetic nobody quotes
Mean Reversion wins 94% of its trades. It also loses roughly eight times as much on a loser as it makes on a winner. Put those together and one losing trade undoes about eight winners.
Eight wins at £47 is £376. One loss at £397 is £397. That is a losing sequence, and it arrives on schedule.
The strategy is comfortably profitable — its profit factor is 1.73 across 2,727 positions, and it was positive in all four years tested. But none of that is because it wins 94% of the time. Anyone shown "94% win rate" and nothing else would form a wildly wrong impression of what holding it feels like, and would be just as wrong about a strategy at 1.73 as about one at 1.10.
What it feels like to hold
Here is the number that actually describes the experience. Mean Reversion's worst peak-to-trough drawdown was around £3,434 per lot. Its losses are not frequent — they are large and they cluster, because the conditions that break a mean-reversion trade tend to break several at once.
By year its profit factor reads 1.77, 1.47, 2.16, 1.47 — positive throughout, but with a spread wide enough that any single year would have given a badly misleading impression of the other three. Swing Rider, on the same measurement, has a losing year sitting inside a positive overall figure. A win rate tells you nothing about either of those facts.
The strategy with the worst win rate carries the least tail risk
Stretch Fade wins 54% of the time, barely better than a coin toss, and makes 1.14 — the lowest profit factor of anything discussed here. It was nonetheless positive in all three test windows, its 90% confidence range sits entirely above breakeven, and ten of its twelve pairs make money.
It is not a better strategy because it wins less — on the numbers, Mean Reversion is plainly the stronger system. But its edge does not depend on the rare large loss staying rare. A 94% strategy is one bad tail away from a very different profit factor, and that tail is the whole risk. A 54% strategy with a positive expectancy has already priced its losses in and takes them constantly. Which of those you can actually hold is a question about you, not about the ratio.
The honest caveat, which we publish on the strategy itself: Stretch Fade enters on resting limit orders, and no backtest can prove those fill in the real world at the modelled price. That is exactly why it is marked experimental and demo-only rather than promoted on the strength of that 1.14.
What to ask instead
- Average win against average loss. A win rate without this is half a sentence. Together they tell you the shape of the strategy.
- Maximum drawdown against total profit. Whatever fraction of its gains it handed back at the worst moment is what you would have had to sit through.
- The year-by-year breakdown. An average across four years hides the year it lost, and you have to live through years, not averages.
- How many trades. A 94% win rate over 40 trades is noise. Over 2,727 it is a description.
We publish all four for every strategy we ship — including the one measuring 1.05, which we tell you not to fund. In July 2026 we also published four of these figures against the wrong strategies and had to correct them; that article covers both what the numbers are and how we got them wrong.
These are backtested and simulated results, shown after out-of-sample validation. They describe how each system behaved historically. They are not a forecast, not a promise, and not a return you should expect. Our evidence standard →
See all seven, with their numbers
Profit factors describe what an automated strategy did. They say nothing about the risk stack around it or the work you do yourself — which is where most of this platform actually lives. See what our own risk guards cost you, measured the same way, and why the desk argues with your analysis instead of handing you a signal.