🏆 Strategy League

Weighted towards what leverage can't fake: Sharpe, Sortino, profit factor and expectancy don't change when you lever up — CAGR does, so it's no longer the heaviest term. Calmar stops counting past a 60% drawdown, or a near-blowup scores like a great year. Expectancy is per-trade edge over average loss: it catches strategies whose average trade barely clears the fee.

Caught peeking at future candles by lookahead-analysis — which re-runs the strategy rather than reading it — and you're out of the League. The static trust check only reads code, so it scales instead: 30% for a leak or an overwritten candle column, 70% for a v2 signal that never fires. And of the top 20, fewer than one look like flukes once significance is corrected across all 1886 — though that test only asks whether profits beat random sign-flips, so it can't see overfitting. Full detail →