Skip to content

Is your edge real — or a mirage?

Paste your backtest. We run the same honesty checks we use to kill our own signals — survivorship, overfitting (deflated Sharpe), single-trade dependence, regime robustness, and realistic cost — and tell you, plainly, whether the edge holds up. Pasting takes no signup, and the verdict is never an AI opinion — it is reproducible math.

This analyzes the hypothetical, past record you supply. It is an impersonal analytical tool, not investment advice, and it never predicts future returns.

Paste your backtest

Optional columns: date — unlocks the regime and out-of-sample checks — ticker, entry, stop.

“Variants you tested” = how many parameter sets, tickers, or re-runs you tried before settling on this one — it drives the overfitting haircut. Honest traders undercount; leave blank and we assume a conservative 10.

gauntlet --checks
free · no signup · no LLM
  • Survivorshipwas the universe picked with hindsight — today's winners only?
  • Overfitting haircutdeflated Sharpe: how much edge survives the variants you tried?
  • Single-trade dependencedoes the whole P&L hinge on one or two outlier trades?
  • Regime robustnessdoes the edge hold across bull / neutral / bear (needs dates)?
  • Realistic costdoes the edge survive spread + slippage at realistic bps?
  • Data integrityduplicate rows, impossible returns, gaps — flagged, not hidden.
$ paste your backtest above — six honesty checks run instantly.