R-Multiple or Win Rate? Why Most Traders Track the Wrong Number

A study of 25,000 retail accounts found 65% had win rates above 50% but 82% were net negative. Win rate is a vanity metric. Here's what to track instead.

There's a study that gets cited a lot in trading education. Twenty-five thousand retail accounts, around four million trades. The headline finding: 65% of those accounts had win rates above 50%. And 82% of those same accounts were net negative.

That's the entire problem with win rate as a metric.

If you're tracking win rate as your primary number, you're tracking a thing that doesn't predict whether you make money. You can have a 70% win rate and lose money. You can have a 35% win rate and make money. The number that determines which side you land on is somewhere else.

Here's the version of this conversation that actually goes somewhere useful.

Why win rate is a vanity metric on its own

Win rate tells you the percentage of trades that finished green. It says nothing about how big the wins were or how big the losses were. A trader with a 70% win rate who's making 0.5R on wins and losing 2R on losses is bleeding capital. A trader with a 35% win rate making 3R on wins and losing 1R on losses is compounding nicely.

Most retail traders end up in the first category for a specific reason. The instinct to "lock in" wins kicks in too early, and the instinct to "give it room" kicks in too late on losers. So winners get small, losers get big. The win rate stays high because most trades hit your too-tight target. The PnL bleeds because the few losers eat through ten winners.

Tracking win rate alone reinforces this exact problem. You see "70%" and feel like you're winning. The PnL number tells the actual story, and it's negative.

R-multiple is better but incomplete

R-multiple measures each trade as a multiple of your initial risk. If you risked 1% to make 2%, that's +2R. If you risked 1% and lost 1%, that's -1R. Aggregate R across your trades and you get a unitless number that's comparable across position sizes, accounts, and time periods.

This is a real upgrade over win rate. R-multiple distribution shows you whether your wins are bigger than your losses. A median of +0.4R across 100 trades means something. "70% win rate" doesn't.

But R-multiple by itself still misses things. A trader with average +1.5R wins and -1R losses at 40% win rate has positive expectancy on paper. If they hit a streak of seven losses in a row (which happens often at 40% win rate), they might be down 7R on the year and quit before the strategy plays out. So R-multiple needs to sit alongside expectancy and drawdown.

Expectancy: the number win rate and R together produce

Expectancy is the formula that combines win rate and R-multiple:

Expectancy = (win rate × average win in R) - (loss rate × average loss in R)

So a 40% win rate trader with +2R wins and -1R losses has:

(0.40 × 2) - (0.60 × 1) = 0.80 - 0.60 = +0.20R per trade.

That trader makes 0.20R on average across every trade they take. Over 200 trades, that's +40R, regardless of how the streaks distributed. Over 500 trades, +100R. The number compounds.

Compare a 70% win rate trader with +0.5R wins and -2R losses:

(0.70 × 0.5) - (0.30 × 2) = 0.35 - 0.60 = -0.25R per trade.

This trader is losing 0.25R per trade despite winning 70% of the time. Over 200 trades, -50R. Over 500, -125R.

Expectancy is the number that actually predicts whether your strategy makes money. Most retail journals show win rate and PnL prominently. The good ones show expectancy. The exceptional ones show expectancy with confidence intervals so you know whether your sample is big enough to trust the number.

The third number nobody tracks: exit efficiency

Here's the one that gets skipped even by traders who already track expectancy.

MFE (maximum favourable excursion) is how far the trade went in your direction before you exited. MAE (maximum adverse excursion) is how far against you it went before turning around or stopping out.

The ratio of where you exited vs where the trade peaked tells you whether your exits are leaving money on the table or whether your stops are too tight.

If your average winner exits at 70% of MFE, you're capturing most of the move. Decent.

If your average winner exits at 40% of MFE, you're leaving more than half the available R on the table. That's a fixable problem, usually a too-close target or a fear-of-giveback exit habit.

If your average loser exits at 90% of MAE, your stops are positioned where the market actually wants to flip. Tighten them and you'll get stopped on every wick. That's usually a sign your stop placement methodology is wrong, not your sizing.

MFE/MAE analysis turns vague intuitions ("I exit too early") into specific data. You can see the actual gap and work on closing it.

What to actually track

Three numbers, not one:

Expectancy in R , with a note on sample size. Anything under 50 trades is too noisy to trust. Anything over 200 trades is solid.

R-multiple distribution . Not the average, the histogram. Are most of your wins clustered around 1R, or are there a few 4R outliers? The shape matters because outliers are how strategies survive losing streaks.

Exit efficiency: average MFE-capture ratio for winners, average MAE-distance for losers. This tells you which side of your strategy needs work.

Win rate is fine to glance at. Don't make decisions based on it.

The trap most traders fall into

Once you start tracking expectancy, the temptation is to maximise it by doing one of two things: making your stop tighter (which increases R per win mathematically) or making your target wider (same idea). Both can collapse your win rate so badly that the expectancy goes negative anyway.

The healthier framing is: optimise the setup, not the math. If you can find conditions that produce a higher MFE relative to MAE on average, your expectancy improves naturally without changing how you size or exit. Custom field tagging in the journal helps with this. Tag every trade with the conditions present, filter by those conditions later, see which combinations have the best MFE-to-MAE ratio.

That's the real value of a good journal. Not "your win rate is X". The diagnostic question of "which setups in which conditions actually produce positive expectancy with clean exit efficiency, with enough samples to trust it".

If your current journal can answer that question in under a minute, it's working. If it can't, it's why you're not improving despite logging every trade.

Why most retail dashboards stop at win rate

Win rate and PnL are easy to compute and they make the user feel something on every login. Expectancy with confidence intervals, MFE/MAE distributions, and rule-followed-vs-rule-broken segments require structured data and a dashboard that surfaces the right cuts.

TradeSave+ was built around exactly this list. Expectancy with sample size and confidence chips. R-multiple histograms by setup, by regime, by session. MFE-capture and MAE-distance per filter. Plus the Edges and Leaks analyser that ranks every (factor, value) bucket by PnL impact with statistical confidence so you can see which conditions produce the cleanest exit efficiency without building 30 custom views by hand.

7-day free trial if you want the diagnostic answers ranked for you.

One last thing

The 25,000-account study isn't about retail traders being bad. It's about retail traders tracking the wrong number. You can fix this in a week of journal restructure. Drop win rate from your dashboard. Replace it with expectancy. Add R-multiple distribution. Add MFE-capture ratio. Watch what changes.