Trading math

Win rate vs expectancy: why 70% winners can lose

Win rate is the most quoted number in trading and the most misleading. Expectancy is the one that tells you if a strategy works.

Win rate is the most quoted number in trading, and the most misleading. A trader who wins 70% of the time can lose money, and one who wins 35% of the time can do very well. The number that actually tells you whether a strategy works is expectancy.

A 70% win rate that loses money

Ten trades: 7 winners of $50 each and 3 losers of $150 each.

Wins: 7 × $50 = $350. Losses: 3 × $150 = $450. Result: −$100.

The trader was right 70% of the time and still lost. This is a common pattern: taking profits quickly and letting losers run.

Expectancy: the average result per trade

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

For the trader above: 0.7 × $50 − 0.3 × $150 = −$10 per trade. Now compare a trader with a much lower win rate:

Win rate 40%, average win $300, average loss $100.

Expectancy = 0.4 × $300 − 0.6 × $100 = +$60 per trade.

Losing 6 trades out of 10 feels bad, but the strategy earns money over time. Expectancy combines both halves of the picture, how often you win and how big wins are compared to losses, into one number.

Expectancy in R

It is easier to compare strategies when you measure results in R, where 1 R is the amount you risked on the trade. A win of 3 R means you made three times your risk. Then:

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

A strategy with +0.3 R expectancy earns 30% of your risk per trade on average. At 1% risk per trade, that is about +0.3% of the account per trade, before fees.

The win rate you need for each reward : risk

Average win : average lossBreak-even win rateExpectancy at 40% win rateExpectancy at 60% win rate
0.5 : 166.7%−0.40 R−0.10 R
1 : 150.0%−0.20 R+0.20 R
1.5 : 140.0%0.00 R+0.50 R
2 : 133.3%+0.20 R+0.80 R
3 : 125.0%+0.60 R+1.40 R

The risk reward calculator does this for a single trade plan: enter entry, stop, target and your win rate.

How many trades before your win rate means anything?

This is the part most traders skip, and as a data analyst it is the one I find most important. A win rate from 20 trades is mostly noise. Using a standard 95% confidence interval for a true win rate of 50%, here is how much your measured win rate can swing just by chance:

Number of tradesMargin of error (95%)A true 50% can look like
20±22 points28% – 72%
50±14 points36% – 64%
100±10 points40% – 60%
200±7 points43% – 57%
500±4 points46% – 54%

After 20 trades, a coin-flip strategy can easily show 30% or 70%. After 100 trades you still have a ±10 point band. That is why changing your strategy after a handful of losses usually doesn't help: you are reacting to noise, not information.

Practical takeaway: judge a strategy on at least 50 to 100 trades taken with the same rules, and track average win and average loss next to the win rate.

How to improve expectancy

Examples are simplified and ignore fees. The confidence intervals use the normal approximation for a binomial proportion. This is educational content, not financial advice.