The RefutationEssays
How to tell a real edge from a mirage
Plain-English pieces on the methods we use to separate signal from noise — and the failure modes that fool everyone else. Education and data-journalism only; never signals or advice.
The method
Why most backtested edges are statistical mirages
Search enough parameters and something always looks profitable. How to tell luck from an edge.
The trapSurvivorship bias in backtesting, explained
The quiet assumption that inflates almost every strategy — and how to strip it out.
Checklist10 signs your trading strategy is overfit
A smooth curve, a Sharpe above 3, a magic parameter. Ten tells for the sceptic.
The harnessOut-of-sample testing: the harness that kills most edges
Two tiers, next-bar fills, and Leave-One-Regime-Out. A dollar or a headstone.
New refutations, straight to your inbox
One rigorous note at a time. Free refutations; paid members get the survivor research.
No signals. No spam. Unsubscribe anytime.