Ask a trader how their system is doing and you'll usually hear a win rate. It's the wrong number. A 70% win rate is a disaster if the 30% losses are four times the size of the wins, and a 35% win rate is a business if the winners run. The number that settles it is expectancy.
The formula
Expectancy is what you expect to make, on average, per trade:
Expectancy = (Win rate x Average win) - (Loss rate x Average loss)
Say you win 40% of trades. Your winners average $300, your losers average $150. Then: (0.40 x 300) - (0.60 x 150) = 120 - 90 = +$30 per trade. That's your edge, expressed in the only unit that matters. Take that trade 200 times and the math says roughly +$6,000, before costs.
Flip it. You win 70% of the time, winners average $100, losers average $260: (0.70 x 100) - (0.30 x 260) = 70 - 78 = -$8 per trade. A 70% win rate that loses money. This is the single most common way a trader fools themselves.
Win rate is a feeling. Expectancy is a number. Only one of them pays.
Why it decides everything else
Expectancy is what position sizing multiplies. Size is the lever; expectancy is what the lever acts on. A positive expectancy sized too big still blows up (risk of ruin). A negative expectancy sized perfectly just loses slowly and politely. Get the sign right first, then argue about size.
It's also the honest test of a trading edge. An edge isn't a setup you like or a pattern you can name. An edge is a repeatable situation with positive expectancy after costs. If you can't state the expectancy, you don't have an edge — you have a preference.
Costs are part of the number
Expectancy before costs is a marketing figure. The real one subtracts commissions, the spread you cross, and the slippage between the price you wanted and the price you got. On short-horizon strategies that gap routinely turns a positive number negative — see where most edges quietly die. Compute it after costs or don't bother computing it.
One number, honestly gathered
Expectancy is only as good as the trades you put into it. Two rules keep it honest:
Count every trade. Not the ones you remember. Not the ones that fit the thesis. The scratch trades, the fat-finger, the one you closed early out of boredom — all of it. A record with the embarrassing trades removed will show an edge that does not exist.
Know your sample. Expectancy computed over twelve trades is a rumour. Over four hundred it's evidence. The distinction matters more than most traders want it to — see how many trades before you believe it.
Reading it like an operator
Once you have expectancy per trade, the questions get sharper. Expectancy by setup tells you which plays to keep (attribution). Expectancy by time of day tells you when to stand down. Expectancy by market regime tells you what conditions you actually trade well. Every one of those is the same formula, applied to a slice.
That's the whole discipline: define the situation, gather enough trades, compute the number after costs, and let it decide — instead of the last trade, which is always the loudest voice in the room.