How much trading have you done?This only changes how things are explained. Every "why?" bubble and glossary entry on the page rewrites itself for you. The test itself is identical for everyone.
Thinridge · Independent strategy testing
Find out if your trading rule is real before you pay to trade it.
Describe the rule in five steps. We run it through 18 years of one-minute data, with real costs, against 5,000 coin-flip twins, and score it 0 to 100 on years you never looked at.
$49 a report. Every report, the same test. Nothing else to buy.
Because a rule that looked great on the last two years often stops working the moment real money is on it. A $50K prop-firm evaluation is $49 a month plus $149 when you pass (Topstep’s Standard Path, help.topstep.com, “Topstep Pricing”, read Sep 6, 2026), and it bills every month until you pass or stop. A $49 report first tells you whether the rule you’re about to bet on actually has anything in it.Because in-sample performance is uninformative about edge. The report isolates the one thing a rule claims to know (direction) against a matched random-side null, scores it on an era your development never touched, re-simulates costs to breakeven, and bootstraps the result. It is the test you'd run yourself if you had the data, the time, and no incentive to like the answer.
No. An eval measures you for a few weeks on one slice of tape, and ends the first bad week. This report measures the rule for eighteen years, every era, against five thousand coin-flip twins, before you bet. The audit, if you ever want one, measures how honest the rule’s history is. Three different questions. The eval only answers the first, and it charges you to ask.An evaluation is a short-window sample of trader-plus-rule under a trailing-drawdown constraint; it cannot distinguish edge from era. The report tests the rule’s directional information across the full history with a matched random-side null and an unseen-era holdout, then reports Livability so the trailing-drawdown question can be answered before the fee is paid. The audit adds a named examiner, a code read for look-ahead, and lineage taken under testimony rather than declared.
Test a rule
Describe your rule
Say it the way it was taught to you. The engine reads it back, asks only for what you left out, and opens the five steps already filled in, each with a "why?" beside it. Already have it written as code? Send the file instead.
Describe it, and the engine reads it back
Say the setup the way it was taught to you. Nothing is assumed: what you didn’t say becomes a question, and then the five steps open already filled in.
Bullish engulfing on ES, as taught. In the same window, 37 of 50 pure coin-flip strategies on daily ES were profitable too — the bull market writes résumés.
The Friday-only opening range breakout, found in public and declared as the best of five slices. The exam said yes — and the code gate, at honest lineage, still said no. Validated and tradeable are different words.
The ICT Silver Bullet, written down as a rule and run on eighteen years of NQ. Fourteen of fifteen unseen years lost, a coin beat it 4,628 times in 5,000, and the stop it is always drawn with erased its only profit.
Every report is the same test at the same depth. One price, paid up front when the engine is open, and you can see exactly what you’re buying in the six public teardowns above. There are no tiers, no subscriptions, and nothing is upsold inside the report.
The audit is a different question. The report tests the rule; the audit tests the trader: your own search history examined under testimony instead of declared, a code read for rules the form can’t hold, a human examiner, and a written record with a name on it. It is not a deeper version of the $49 test, and it cannot raise a score. $499.What the audit is, and how to request one →