How to calculate expectancy and profit factor (with a worked example)

The two numbers that tell you whether your trading or betting has an edge: expectancy per trade and profit factor. Formulas, a worked example, what counts as good, and the mistakes that make them lie.

Win rate is the most quoted and least useful number in trading. A 70% win rate with small wins and large losses loses money; a 35% win rate with 3R winners prints. The two numbers that actually answer "do I have an edge?" are expectancy and profit factor. Both take a minute to compute; the expectancy calculator does it live.

Expectancy

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

It is the amount you make, on average, every time you press the button. Use net numbers (after fees) and real dollars, not points.

Worked example. 120 trades last quarter: 52 winners averaging +$186, 68 losers averaging −$121.

Win rate = 52 ÷ 120 = 43.3%
Loss rate = 56.7%
Expectancy = 0.433 × 186 − 0.567 × 121 = 80.6 − 68.6 = +$12.0 per trade

At 120 trades a quarter that is about $1,440 of expected profit per quarter. Small, positive, and — this is the part that matters — it tells you that the plan works and the variance is just variance.

Profit factor

Profit factor = gross profit ÷ gross loss

Same example: gross profit = 52 × 186 = $9,672; gross loss = 68 × 121 = $8,228 → profit factor = 1.18.

How to read it, as rules of thumb rather than laws:

Profit factor	Meaning
under 1.0	You pay to trade. Stop sizing up; fix the leak first
1.0–1.3	Thin edge. Fees, slippage and one tilt session erase it
1.3–1.8	Solid, tradeable edge for a discretionary trader
above 2.0	Excellent — or a small sample. Check the trade count

Break-even win rate

The win rate below which your current win/loss sizes lose money:

Break-even win rate = average loss ÷ (average win + average loss)

Example: 121 ÷ (186 + 121) = 39.4%. Your 43.3% is above it, which is another way of saying expectancy is positive. If your break-even is 55% and your real win rate is 50%, no amount of "discipline" fixes it — the sizing does.

The three mistakes that make these numbers lie

Too few trades. Under ~50 trades, one outlier moves expectancy by more than the edge itself. Compute on the last 100–200.
Mixing setups. One setup at 2.1 profit factor and one at 0.7 average out to "fine". Compute per setup; that table is the whole point of a journal. Template: free trading journal CSV.
Counting tilt trades as trades. Entries taken on tilt are not your strategy, they are your leak. Compute expectancy with and without them; the gap is what the leak costs — see tilt: how to detect it in your own data.

For sports bettors

Same formulas with stakes: average win = average net profit on winning bets, average loss = average stake lost. Yield (net profit ÷ total staked) is the betting equivalent of expectancy per unit risked; a 3–5% yield over 500+ bets is a real edge, 15% over 40 bets is a good month. More in what to log in a sports betting bankroll tracker.

Sharpen computes expectancy, profit factor, break-even and drawdown on every import — overall, per setup, per instrument, per weekday and per hour — and flags the entries that drag them down.