PROOF · proof · Updated 2026-11-05 · Derek William Frazier

Win Rate vs. Expectancy: Why a High Win Rate Can Still Lose Money

A high win rate doesn't make a strategy profitable, and a low one doesn't make it unprofitable. Expectancy is the number that actually decides that, because it weighs how often a strategy wins against how much it wins or loses each time — and a system that's right on 7 of 10 trades can still lose money if its losses are big enough, while a system that's right on only 3 of 10 trades can make money if its wins are big enough.

That's the answer to the search query. The rest of this page covers what expectancy actually is, how to calculate it from a handful of trades, why win rate alone is a trap, how profit factor relates to it, and what none of these numbers tell you on their own.

PLATE · RIGHT MORE OFTEN, STILL LOSINGTen trades each. The cyan line is the average outcome per trade — the expectancy.SYSTEM A · 70% win ratesmall wins, one big loss in tennet expectancy: negativeSYSTEM B · 30% win ratebig wins, small losses the restnet expectancy: positiveThe system that's right more often is the one losing money, on average, per trade.CYCLICAL MARKETS · THE CYCLITECNICAL METHODEducational. Not advice. No performance promise.

What is expectancy, exactly?

Expectancy is the average amount a system makes or loses per trade, once win rate and payoff size are both folded into one number. The formula:

Expectancy = (win rate × average win) − (loss rate × average loss)

Run it on two systems with very different win rates to see why the win rate on its own is close to meaningless:

Win rateAvg. winAvg. lossExpectancy per tradeProfitable?
System A70%20 units60 units(0.70 × 20) − (0.30 × 60) = −4No
System B30%70 units12 units(0.30 × 70) − (0.70 × 12) = +12.6Yes

System A is right more than twice as often as System B and still loses, on average, every time it trades. System B is wrong most of the time and still makes money, on average, every time it trades. The win rate alone told you nothing about which system you'd rather own — you needed the payoff sizes too. (These are illustrative, made-up figures to show the mechanism, not a claim about any real system's results.)

Why doesn't win rate alone tell you if a strategy is profitable?

Because win rate and payoff size trade off against each other, and a strategy's edge lives in the combination, not in either number by itself. Two broad archetypes show up constantly in real trading styles:

  1. High win rate, small payoff per win, occasional large loss. Common in styles that take profit quickly and let a losing trade run or get caught by a gap — scalping-style and some options-selling approaches lean this way. It feels good day to day: you're "right" most of the time. The risk sits in the tail — one outsized loss can erase a long run of small wins.
  2. Low win rate, large payoff per win, frequent small losses. Common in trend-following and breakout styles that cut losers fast and let winners run. It feels bad day to day: you're "wrong" most of the time, and the losing streak between wins can be long enough to test anyone's discipline. The edge sits in the few trades that pay for all the rest.

Neither archetype is inherently better. A profitable version and an unprofitable version both exist of each one — the only way to tell which you have is to actually compute the expectancy, not to eyeball the win rate and assume.

What is profit factor, and how is it different from expectancy?

Profit factor is gross profit divided by gross loss across every trade in the sample — one ratio for the whole set, rather than a per-trade average.

Profit factor = total $ (or points, or %) won ÷ total $ (or points, or %) lost

A profit factor above 1 means the sample made more than it lost in total; below 1 means the opposite. It's a useful quick check, but it has a blind spot expectancy doesn't: profit factor can look healthy purely because of trade count or one outsized win, without telling you what a "typical" trade looked like. Expectancy answers a more specific question — what should I expect from the next single trade, on average — which is the number that actually matters for deciding whether to keep taking the next one.

Both numbers can mislead if the sample is too small or too selectively chosen to trust in the first place — see how many calls it takes before a track record means anything and how a cherry-picked record hides this exact math by quietly dropping the losses that would have pulled the win rate and the expectancy both down.

Why does a losing system often look better day-to-day than a winning one?

Because the lived experience of trading is sequential, not statistical, and a high win rate produces more "I was right" moments per week even when it loses money overall. That mismatch between how a system feels and what it actually does is one of the most reliable ways traders fool themselves:

This is also why a hot stretch from either archetype doesn't prove much on its own — a run of wins from System A and a run of wins from System B both feel identical in the moment, and both revert toward the system's real average given enough trades. More on that specific trap: why every hot streak cools off.

How does position size interact with expectancy?

Expectancy per trade only becomes a real number — dollars, not a ratio — once it's multiplied by how large each trade actually is and how many trades get taken. A small positive expectancy, sized consistently and taken often enough, compounds into a real result over time; a large positive expectancy that's barely ever acted on, or sized so small it doesn't matter, doesn't. That sizing question is a separate topic from expectancy itself — see the four dimensions of trade risk for how size, stop, target, and time interact on a single trade.

One mechanical point worth stating plainly: expectancy describes the average, not any single trade. A positive-expectancy system can still produce a losing streak long enough to look broken, and a negative-expectancy system can still produce a winning streak long enough to look like genius. The average is only trustworthy once there have been enough trades to measure it — the same sample-size problem that applies to a win rate on its own.

What expectancy doesn't tell you

Questions traders ask

Is a 90% win rate (9 of 10 trades) always a good sign?

No. Winning 9 of 10 trades, with small wins and one rare large loss, can still add up to negative expectancy — the same shape System A above has at a lower win rate. The math doesn't care how satisfying the number looks, only how the wins and losses actually size out.

What counts as a "good" expectancy number?

There's no universal threshold — it depends on how often the system trades and what it costs to trade it. A small positive expectancy taken often, at low cost, can outperform a large positive expectancy taken rarely. The honest comparison is always expectancy net of real trading costs, not the gross number alone.

Can profit factor be misleading on its own?

Yes, in the same direction win rate can be misleading — a profit factor above 1 confirms a sample made more than it lost in total, but it won't tell you whether that came from a steady edge or from one or two outsized trades carrying an otherwise flat or losing sample. Checking expectancy per trade alongside it, on a sample large enough to trust, catches that gap.

Does a high-expectancy system guarantee I'll make money trading it?

No — expectancy is the average over many trades, not a promise about any individual trade or any individual trader's execution of it. Discipline, consistent sizing, and real trading costs all sit between a system's expectancy on paper and what any one person actually realizes trading it.

Where this fits the bigger picture: expectancy, sample size, and clean (not cherry-picked or curve-fit) data are the three legs a track record needs before any number from it means anything. The Sharpe ratio is the same idea one level up — it folds return and bounce together instead of win rate and loss size. Once a win rate and win/loss ratio are trusted, the Kelly Criterion is the formula that turns those same two numbers into a position size — and why full Kelly is rarely traded straight. How the method behind this site's own graded calls holds up against that bar is on the methodology page.

Keep reading

I Graded Four Years of My Own Public Calls. Here's the Scoreboard.How a Trading Educator Should Score Their Own CallsHow to Verify a Trading Guru's CredentialsMy 2022 Posts Beat My 2026 Posts. Badly.
Get our AI tools and everything we send out New AI trading tools and our Claude Code builds, plus session invites. All free.
Get them free