The one formula that tells you whether a strategy makes money, two examples that break intuition, and how leverage quietly destroys expectancy.
Expectancy equals win rate times average win, minus loss rate times average loss, measured in R (multiples of planned risk). A 40% win rate with 2R wins and 1R losses yields +0.2R per trade: profitable. A 70% win rate with 0.5R wins and 2R losses yields −0.25R per trade: a slow bankruptcy with a great-looking record.
Taking profit early and letting losers run produces a high win rate and a negative expectancy. It feels good every day and loses money every month. Many "signal services" sell exactly this shape, because the win rate is the number that sells.
High leverage caps your wins, because the first adverse wick scares you out, and uncaps your losses, because liquidation is −100% of margin. Both terms in the formula move the wrong way at once. The same setup at 5x and at 50x has different expectancy, and the difference is the leverage, not the setup.
Expectancy needs a sample. Below fifty trades, every opinion about your edge is noise, including the flattering ones. The fastest way to get fifty trades at realistic size without paying for them is a simulator with live prices, logged exactly as if they were real.