Win Rate vs Expectancy: Why a 70% Win Rate Can Still Lose Money

"It wins 7 out of 10 trades."

That's one of the most common ways people sell a strategy, and one of the most common ways people fool themselves. Win rate feels like the score. It isn't.

A strategy can win most of the time and still lose money. Another can lose most of the time and still come out ahead. The number that tells them apart is called expectancy. This post covers what it is, how to work it out, and how to check it without kidding yourself.

What win rate leaves out

Win rate only counts how often you win. It says nothing about how much you win or lose each time.

Picture two strategies:

• Strategy A wins 70% of the time. Each win makes $100. Each loss costs $300.
• Strategy B wins 35% of the time. Each win makes $300. Each loss costs $100.

A looks great on a results page. B looks like a mess. Now look at 10 trades of each:

• A: 7 wins × $100 = $700. 3 losses × $300 = $900. Down $200.
• B: 3.5 wins × $300 = $1,050. 6.5 losses × $100 = $650. Up $400.

The "worse" strategy is the one making money.

Expectancy in one line

Expectancy is the average amount you expect to make or lose per trade, over many trades:

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

For A: (0.70 × $100) − (0.30 × $300) = $70 − $90 = −$20 per trade.

For B: (0.35 × $300) − (0.65 × $100) = $105 − $65 = +$40 per trade.

Many traders write it in "R," where 1R is the amount you risk on a trade. That way you can compare strategies of any size. A risks 3R to make 1R. B risks 1R to make 3R.

Two things to include, every time:

• Fees and slippage. They come out of every trade, win or lose. A strategy with tiny wins can be wiped out by fees alone.
• Real averages, not targets. Use the wins and losses the test actually produced, not where you hoped your take-profit and stop would fill.

We ran it 100 times

Ten trades is a story. We wanted to see what happens over many runs, so we ran a quick simulation. Both strategies paid a small fee on every trade (0.05R). Each one got 100 separate runs.

Over 200 trades per run:

• Strategy A (70% win rate): 4 of 100 runs net positive (simulated).
• Strategy B (35% win rate): 100 of 100 runs net positive (simulated).

Over just 20 trades per run:

• Strategy A: 17 of 100 runs net positive (simulated).
• Strategy B: 75 of 100 runs net positive (simulated).

Two lessons sit in there. First, over enough trades, expectancy wins and win rate doesn't. Second, short records lie in both directions. With only 20 trades, the losing strategy looked fine about one time in six, and the better strategy looked bad one time in four.

The catch with low win-rate strategies

Strategy B made money in every long run. But it's not easy to live with.

In the 200-trade runs, B's median longest losing streak was 10 trades in a row (simulated). The worst run hit 19 in a row (simulated). A's median worst streak was only 4.

That's why high win-rate strategies are so popular even when they lose. Losing slowly with lots of small wins feels better than winning with long dry spells. If you'd quit after eight losses in a row, a strategy like B isn't really your strategy, no matter what the math says. Size it so a long streak is boring, not scary. We cover that in Position Sizing for Beginners and What Is Maximum Drawdown?.

How to check expectancy honestly

1. Look at win rate and the win/loss sizes together. Never accept one without the other. If someone only shows win rate, ask for average win and average loss.

2. Put fees in first. Work out expectancy after fees, not before. See How to Backtest Without Fooling Yourself.

3. Watch for a few huge trades. If one or two big wins make up most of the gains, take them out and look again. Ask whether you'd really have caught them.

4. Don't trust a short record. Our 20-trade runs got the answer wrong often. More trades on new data is the cure.

5. Test forward. Lock the rule and run it on market data it has never seen, in paper trading. Then compare its win rate and win/loss sizes to the backtest. See Backtest vs Forward Observation and In-Sample vs Out-of-Sample Testing.

Where AI fits

AI tools like ChatGPT, Claude, Grok, or Cursor can write strategy rules and run quick backtests in minutes. When you ask for results, ask for these every time: win rate, average win, average loss, expectancy after fees, number of trades, and longest losing streak. If the tool only reports win rate, the report isn't finished.

How Pickled Lizard tests ideas like this

Pickled Lizard is an AI trading and research platform. You bring the idea, or your own AI does, and we test it the same way we test everything:

1. Backtest on the full data set we have, with fees and next-candle timing.
2. Paper forward testing on new market data, with real fees and timestamps. The idea shows as "still collecting" while it builds a record.
3. Verified only when its forward results reach 80% of its own backtest trade rate and 80% of its backtest win rate.

Results show on the public paper trading leaderboard. Pickled Lizard isn't an exchange and doesn't hold your money. It's a non-custodial aggregator that connects to your own exchange accounts. Our servers are in Europe. The beta is free, and live trading is off during the beta.

Want to see how your idea holds up on new data?

Test it free at pickledlizardz.com/join

Quick recap

• Win rate tells you how often you win, not whether you make money.
• Expectancy = (win rate × average win) − (loss rate × average loss), after fees.
• In our simulation, a 70% win-rate strategy was net positive in only 4 of 100 runs, while a 35% win-rate strategy was net positive in 100 of 100 (simulated, 200 trades each).
• Short records mislead, and low win-rate strategies come with long losing streaks. Size for them and test forward.

New here? Start with What Is a Quant Trading Strategy?.