Expectancy and Win Rate
Win rate on its own says nothing. Expectancy combines it with the sizes of wins and losses to give the average result per trade, which is the number that decides everything.
MadStockAlerts Research · Updated August 28, 2026
What to take away
- Expectancy is the average outcome per trade, combining frequency and size.
- A high win rate with large losses can be negative expectancy.
- A low win rate with large wins can be strongly positive.
- Sample sizes needed to establish expectancy are much larger than most assume.
- Expressing results in R makes them comparable across position sizes.
- At a 40% hit rate, eight losses in a row should be expected more than once in a hundred trades.
MAD Academy Training Video · 0:46
Being Right Is Not the Goal
Expectancy combines how often you win with how much you win, and it is entirely possible to be right most of the time and lose money.
This lesson is part of a Stock Alerts + Tools plan.
The calculation
expectancy = (win rate x average win) - (loss rate x average loss)
| Method | Win rate | Avg win | Avg loss | Expectancy |
|---|---|---|---|---|
| A | 70% | $100 | $300 | -$20 |
| B | 30% | $500 | $100 | +$80 |
| C | 50% | $200 | $150 | +$25 |
Method A wins seven times out of ten and loses money. Method B is wrong seven times out of ten and is the best of the three. Win rate on its own is not merely incomplete; it is capable of ranking methods in exactly the wrong order.
Why high win rates are seductive
Being right feels better than being profitable, and a high win rate delivers that feeling on most days. Methods that produce many small wins and occasional large losses are therefore psychologically comfortable right up until the large loss arrives.
Method B is the harder one to run for exactly the opposite reason. Seven losses in ten, arriving in unpredictable order, produces long stretches in which the method appears broken and is working precisely as designed.
Sample size
Twenty trades tell you almost nothing. A method with a genuine 40 percent win rate produces stretches of eight consecutive losses regularly, and a run of twelve wins in twenty is entirely consistent with a losing method.
Establishing expectancy with any confidence takes well over a hundred trades, and most judgements about whether a method works are made on far fewer than that. This is the same problem the failed patterns article describes from the other direction.
Expectancy per unit of risk
Expressing outcomes in R, where one R is the amount risked, makes results comparable across differently sized positions. A method averaging +0.3R per trade is described completely by that number, whatever the account size or the individual position sizes were.
It also makes the connection to sizing explicit. Expectancy in R is a property of the method; the money it produces is that property multiplied by the amount risked per trade, which is the only part the trader controls directly.
Variance, and why a good method has losing runs
A method that resolves favourably forty percent of the time will produce a run of six consecutive losses roughly once in every twenty-one attempts at a six-trade sequence. That is not a failure of the method; it is what forty percent means. The runs are not rare enough to be remarkable, and they are long enough to feel like evidence.
| Hit rate | Chance of 5 losses in a row | Chance of 8 in a row |
|---|---|---|
| 60% | About 1 in 98 | About 1 in 1,526 |
| 50% | About 1 in 32 | About 1 in 256 |
| 40% | About 1 in 13 | About 1 in 60 |
| 33% | About 1 in 7.6 | About 1 in 26 |
Read across a hundred trades rather than across five, those figures say that a run of eight losses at a forty percent hit rate should be expected to happen more than once. A process that is abandoned on the first such run is a process that was never going to survive its own arithmetic.
The practical consequence is that a drawdown consistent with the method's own statistics carries no information about whether the method still works. Distinguishing an ordinary run from a genuine breakdown requires more data than the run itself contains, which is why the decision to stop is made in advance rather than during one.
Expectancy in R, and why the unit matters
Expressing outcomes in dollars makes trades incomparable, because a dollar outcome mixes the quality of the trade with how large the position was. Expressing them in R, where one R is the amount risked on that trade, removes the size from the measurement and leaves the part that describes the method.
expectancy in R = (hit rate x average win in R) - ((1 - hit rate) x average loss in R)
- one R is the amount risked on the trade, chosen before entry
- a result of +0.3R means each trade returns three tenths of what it risked, on average
The unit also makes the two levers visible separately. A method at a forty percent hit rate and an average win of 2.5R has the same expectancy as one at sixty percent and 0.9R, and they are completely different things to hold: the first spends most of its time losing, and the second produces a long series of small gains punctuated by losses that feel disproportionate.
Expectancy per trade multiplied by the number of trades gives the expected return in R over a period, which is where frequency enters. A small positive expectancy taken often is arithmetically equivalent to a large one taken rarely, right up until costs are subtracted, and costs scale with frequency.