Win rate fooled me for years. This is what finally fixed it.
For a long time, I judged strategies mostly by **win rate**. Higher win rate = better strategy. Or so I thought. After digging into a bunch of backtests recently, what kept showing up was the opposite: * Some of the …