Doubt your edge before the market does.

DoubtEdge audits your backtest exports and tries to break your strategy on purpose. Pre-registered criteria, placebo tests, plateau checks and walk-forward, explained in plain language by Claude.

Request early access In development. Early testers get it free.
Strategy report · XAUUSD 2hNet profit +184%
Example: a backtest equity curve marked up by DoubtEdge A smooth rising equity curve. Red annotations flag a forgotten take-profit setting, zero commission, a shortened date range, and a filter whose placebo test is chance-level. 20252026 only 21 months of data? TP1 = 50% left on (expected 0) commission = 0 filter vs placebo: p = 0.31, chance level
Illustrative example. The settings errors shown are ones DoubtEdge has caught in real exports.

A process that makes fooling yourself hard

Most tools help you find a strategy that looks good. DoubtEdge does the opposite: it tries to show why yours might not be.

  1. Upload

    Drop in one or more TradingView strategy exports (.xlsx with trades and properties).

  2. Pre-register

    Write down what counts as success before you see any result. The criteria are timestamped and locked.

  3. Audit

    Every export is checked for forgotten settings, shortened date ranges, zero costs and mixed symbols.

  4. Validate

    Deterministic Python checks run the validation suite. Claude never invents a number.

  5. Verdict

    Accept, reject or watch-list, with the reasons and the limits of each test written out.

What gets checked

Defaults come from a method used on a live multi-asset portfolio. You can change them; every change is written into the report.

Two-period acceptance

A change must improve net profit over max drawdown in both halves of the data, not just one.

Placebo test

If your filter blocks 30% of trades, block 30% at random 500 times. Does your filter beat chance?

Parameter plateau

Good parameters have good neighbours. A single sharp peak is a sign of overfitting.

Walk-forward

Pick parameters on past data, run them on the next unseen months, repeat. Measures the selection itself.

Profit concentration

How much of the profit comes from the best 5% of trades, and what is left without them.

Costs and portfolio fit

Real spread and commission, and whether a new strategy improves the portfolio you already run.

What DoubtEdge won't do

  • Buy or sell signals
  • Market predictions
  • Personal investment advice
  • Place orders
  • Hand you “the best” parameters

Where it comes from

DoubtEdge started as the checklist one systematic trader used on his own futures, FX and gold strategies. Written criteria first, two periods, placebo, plateau, walk-forward.

That process rejected far more ideas than it accepted, and caught export mistakes that would otherwise have gone live. DoubtEdge turns it into a tool anyone with a backtest can use.

Test it on your own strategy

We're looking for a small group of systematic traders who write their own TradingView strategies. Send a short note about what you trade.

Email sarper@doubtedge.com