Free · No signup · Nothing leaves your browser
Edge or luck?
Your strategy is up after 60 trades. Is that an edge, or a good run? Enter your numbers or paste your trade results. You get the after-cost average with an honest uncertainty range, the chance that luck alone explains it, and roughly how many trades you still need.
Not another statement analyzer. One question: luck or edge, after costs, with uncertainty.
Check your numbers
Now run the same question on your real export.
Averages typed from memory hide a lot. The Checker reads supported MT5 and TradingView exports offline on your desktop, counts commission, swap and fees once, stress-tests extra costs, and puts an uncertainty range beside every average. “Not enough evidence” is an allowed answer. Planned at $27 (US sales tax may apply by state) / 27,00 € incl. VAT, one-time, no subscription.
- No broker login, no upload: your file stays on your computer
- Cost stress: what happens if spread or slippage is worse
- Accepted and rejected rows shown, so you can check one known trade
Opens on Kit. Confirm your email, then get a few launch emails (max. 3). Unsubscribe anytime. No general newsletter. Orders are not open yet.
Not sure your export works? See supported formats and limits.
How the check works
Summary numbers. Every win counts as your average win, every loss as your average loss, minus the extra cost on every trade. Your after-cost break-even win rate is (loss + cost) ÷ (win + loss). The luck chance is the exact binomial probability that a strategy sitting exactly at break-even after costs wins at least as often as you did. The 95% range uses a Wilson interval on the win rate. The chance the true average is above zero uses a neutral Beta(½, ½) starting point.
Pasted trades. The cost is subtracted from every trade. We resample your own trades 2,000–10,000 times (bootstrap, fixed seed so the same input gives the same answer). The 95% range is the middle 95% of resampled averages. The luck chance is how often a zero-edge version of your trades (same spread of results, average shifted to zero) produces an average at least as good as yours.
Variants tried. Test 20 settings and keep the best, and one will usually look good by luck. We show the chance that at least one of that many zero-edge variants looks this good: 1 − (1 − luck chance)variants.
Trades still needed. Roughly how many trades an average and spread like yours needs before luck drops under 5% (adjusted for variants): (z × spread ÷ average)², with z ≈ 1.645 for one variant. It assumes your true numbers stay exactly as observed. They usually don’t.
What this cannot see
- Whether the trades are complete, or whether losing ones were left out.
- Overfitting, look-ahead bias or repainting in a backtest, beyond the variants count you enter.
- Streaks and trades that depend on each other, changing position size, regime changes and future costs.
- Summary mode ignores how much individual wins and losses vary, so its range is too narrow if your results vary a lot. Pasting real trades is better.
Educational statistics only. Not investment advice, a recommendation or a forecast. A low luck chance is not proof of a future edge; past results do not predict future results. Trading can lose money. Examples on this page are invented. Your inputs are calculated in this page only: no cookies, uploads or browser storage. We count anonymous page views with Cloudflare Web Analytics (cookieless, aggregate visit stats). The shareable link holds only summary numbers, never pasted trades.