What it doesLets you type your strategy in normal English instead of filling boxes. It sets up the test for you and then explains the result.

When to use itStart here if you are not sure how to express your idea as rules.

Reading the resultThe numbers come from the same engine as every other tool — the AI only explains them, so it cannot make the results look better than they are.

What it doesYou pick the entry rule (for example: EMA 20 crosses above EMA 50), a stop-loss and a target. It then replays history bar by bar and records every trade those rules would have taken.

When to use itThis is the starting point for every idea. Everything below tests the strategy you build here.

Reading the resultProfit factor above 1.5 and a Sharpe above 1 are healthy. Fewer than about 30 trades means the result is mostly luck — widen your date range.

What it doesRuns your strategy hundreds of times with different settings (different EMA lengths, stops, targets) and shows which combination performed best.

When to use itAfter your basic idea shows some promise, to find sensible settings.

The trap to avoidThe single best result is usually too good to be true — it has been fitted to past noise. Prefer a setting where the neighbouring values also did well; that means the edge is stable, not a fluke. Always confirm with step 3.

What it doesRuns the exact same rules from the Strategy Builder across several pairs at once, splitting your capital evenly between them.

Why it mattersAn edge that only works on one symbol is usually luck. If the same rules make money across most pairs, the pattern is far more likely to be real.

Reading the resultLook at profitable pairs first — if only 1 of 5 made money, do not trust it. Then check whether each pair still beats buy & hold.

What it doesSplits history into chunks. It tunes the strategy on an early chunk, then tests it on the next chunk it has never seen — then rolls forward and repeats. This copies how you would really trade: decide today, find out later.

Why it matters mostThis is the single best test for whether an edge is real. A strategy can look perfect in a normal backtest purely because it was tuned on the same data it is being judged on.

Reading the resultIf out-of-sample results stay close to the tuned results, the edge is likely genuine. If they collapse, the strategy was overfitted and would probably lose money live.

What it doesTakes your actual trades and reshuffles their order thousands of times. Same trades, different sequence.

Why order mattersLuck decides whether your losing trades arrive spread out or all in a row. A strategy showing a 12% drawdown in the real order might show 28% if the losers had clustered at the start.

Reading the resultLook at risk of ruin and the worst drawdown across all runs, not the average. If a realistic bad run would wipe out your account, cut your position size — the strategy itself may be fine.

What it doesReplays your session changing one thing at a time — a wider stop, an earlier exit, skipping your worst hour — and shows what the result would have been.

When to use itAfter a backtest, to find which single change would have helped most.

Reading the resultTreat this as a source of ideas, not proof. Any change you adopt should be re-tested through steps 1 to 4 before you trust it.

What it doesReplays your strategy under prop-firm evaluation rules — profit target, daily loss limit and maximum drawdown — and tells you whether it would have passed.

When to use itBefore paying for a funded account challenge.

Reading the resultA profitable strategy can still fail a challenge, because one bad day can breach the daily loss limit even if the month ends green. Pass rate matters more than profit here.

What it doesArgues the case against your strategy and points out the weaknesses you would rather not look at.

When to use itWhen a result looks too good and you want a second opinion before risking money.

Reading the resultIt is deliberately sceptical. If it cannot find a serious hole, that is a genuinely good sign.

What it doesKeeps a ranked list of the strategies you have tested, so you can compare ideas over time.

Where it is storedSaved in this browser only — nothing is uploaded, and it will not follow you to another device.

Reading the resultCompare on Sharpe and walk-forward results, not on net profit alone. The highest-profit strategy is often the most overfitted one.