In this guide
Profit factor is a totals ratioExpectancy is an average outcomeOne synthetic sample, two viewsNeither metric shows the pathSample context mattersSide-by-side comparisonProfit factor is a totals ratio
Profit factor is gross profit divided by gross loss. If a fictional sample produces 75R of gross gains and 50R of gross losses, profit factor is 1.50. The ratio does not require the number of observations, which is also one of its limitations: a sample of ten outcomes and a sample of one thousand can display the same ratio.
Side-by-side comparison
| Metric | Core formula | Unit | Main blind spot |
|---|---|---|---|
| Profit factor | Gross profit ÷ gross loss | Ratio | Does not show sample size or sequence |
| Expectancy | Win rate × average win − loss rate × average loss | Average outcome per observation | Depends on the assumptions or sample estimates supplied |
| Maximum drawdown | Largest decline from a running peak to a later trough | Percentage | Describes path depth, not average profitability |
Expectancy is an average outcome
Expectancy combines win rate, average win and average loss, or equivalently divides a sample’s net result by its number of observations when all values use the same unit. In a 100-observation fictional sample with 75R of gross gains and 50R of gross losses, the net result is 25R and the sample-average result is +0.25R per observation.
One synthetic sample, two views
Suppose 100 hypothetical outcomes contain 50 wins averaging 1.5R and 50 losses averaging 1R. Gross gains are 75R and gross losses are 50R, so profit factor is 1.50. Net result is 25R, so expectancy measured directly from that completed sample is +0.25R per observation. The two metrics are mathematically consistent but express different units.
Neither metric shows the path
Both profit factor and average expectancy compress a sequence into summary numbers. Neither tells you whether losses arrived gradually or in one severe cluster. Two samples with the same totals can have very different maximum drawdowns because drawdown depends on order.
Sample context matters
A precise ratio does not make a small or unrepresentative sample reliable. Trade count, observation rules, costs and outliers are separate pieces of context. PreBreakout therefore reports the arithmetic without assigning quality grades or benchmark labels.