Four things happen to your strategy
You describe the strategy
Say it the way you were taught it; the engine reads it back and asks only for what you left out. Or fill in the five steps yourself. No coding. You see the strategy written back in plain words, and you lock it.
Because the most common way people fool themselves is by changing the strategy a little, running it again, and keeping the best result. Locking means the strategy you described is the strategy that gets tested. If you change it later, that's a new test, and the report says so. Submission freeze. The locked configuration is the only thing simulated; any change is a new submission counted in your lineage, and the lineage count feeds a multiplicity adjustment on the reported p-value. Without freeze, the test is just an optimizer with extra steps.We replay it on real history
One-minute bars back to 2008 on most markets. Your strategy trades every day exactly as written, with commission and slippage charged on every trade.
Because every trade costs money even when you're right: a fee to the broker, and the small amount you lose because you buy at the asking price and sell at the bid. Lots of strategies make money on paper and lose money after costs. We charge them up front so the number you see is the number you'd keep. For scale, a large prop firm's published round-turn is $3.78 on NQ, ES and RTY, $4.32 on gold and $1.22 on the micros. The engine charges $4.20 plus a tick each way on every market: a little more than that on the index futures, a little less on gold, and about three times as much on a micro, so a micro's result here is the harsher one. $4.20 round-turn commission plus one tick adverse slippage per side, market orders filled at next bar's open; a limit or stop entry fills at its own price. Then re-simulated at two and three ticks, and searched to the breakeven slippage. For bracketed strategies, slippage changes which exits fire, so each rung is a full re-simulation, never an arithmetic subtraction. Reference: a large prop firm's published schedule, $3.78 RT on ES/NQ/RTY, $1.22 on MES/MNQ/M2K, $4.32 on GC, $4.02 on CL (read Sep 3, 2026, exchange and NFA fees included); a sim fills the target on touch with no slippage, which the engine does not.We compare it to coin flips
Five thousand copies of your strategy trade the same days at the same times with the same costs, but pick long or short by coin flip. If your strategy can’t beat them, it shows no skill in picking the side. They test the side it takes, not when it trades or when it sits out: they trade exactly when it trades. A strategy that only ever buys, or only ever sells, gets coin flips on random days with its own side instead, so for it they test the timing.
Because markets mostly went up for fifteen years, so almost anything that bought looks smart. The coin flips get every advantage your strategy had except the one decision it claims to be good at: which direction. If your strategy beats nearly all of them, that decision is worth something. If it doesn’t, its choice of side added nothing the coin flips didn’t have. A random-side null: identical eligibility, entry and exit timestamps, and cost model, with direction drawn per trade. It isolates directional information from timing, frequency and drift. The p-value is the empirical share of 5,000 seeds that match or beat the real net. It's computed on the full window and separately on the years before you first used it (2008 to 2022 on the shelf), and that one governs. True, and the point. The test gives your strategy perfect discipline, honest costs, and no hindsight, then asks whether its one decision — direction — beats a coin flip. If it can’t under those conditions, execution won’t save it. If it can, the report’s lines on holding it say what that would take. A backtest is the floor, not the ceiling, and a strategy that fails the floor is not a strategy. The engine measures directional information under idealized execution with a conservative cost model. Live execution adds variance and, usually, cost; it does not add information. A strategy that fails the random-side null under these conditions shows no skill in its choice of side; its timing and exits may still matter, and the report shows what the coin flips themselves made. The holding lines (max months to new high, worst 12 months, trailing-drawdown survival) are reported so the reader can judge whether a strategy that passed is holdable, which is a separate question from whether it is real.We score it and tell you plainly
A score from 0 to 100 and one of three words: Passed, Close or Failed (reports sent before October 23, 2026 said Validated, Marginal or Rejected). Passed needs a score of 70, no check under half its points, no automatic fail, and two more checks: it beats the coin flips by 1 in 20 counting every try, and its range of likely results on the years before 2023 stays above zero. Close is a score of 40 to 69, or a pass by score that misses one of those two checks; Failed is under 40, or failed automatically whatever the score (too few trades, its money all from 2023 on, or money made before 2023 and lost since). Then the lines that explain it, in plain words, in your email within minutes.
Six checks, each worth a fixed number of points: is the data clean, does the strategy work on the years before you first used it, is it steady year to year, does it beat the coin flips, does it survive costs, and does it fall apart if you nudge a setting. A strategy has to do reasonably well on all six. One great number can't rescue five bad ones. The scoring (Protocol v1.1): six weighted examinations (data integrity 5, out-of-sample 25, era consistency 20, Monte Carlo 15, cost stress 20, parameter sensitivity 15). A pass needs 70+ with no pillar under half its points and no gate tripped, and, since October 25, 2026, two more checks: the coin flips counted for every try (the lineage-adjusted p) at 1 in 20 or better, and the bootstrap's range for the years before you first used it above zero; a score of a pass that misses either is Close, and the report says which; gates include insufficient sample and era-manufactured edge. The exact thresholds are ours; the structure and every input are printed on your report.- Fills at the next bar’s open (a limit or stop entry fills at its own price), one tick against you per side, $4.20 a round turn. A target limit has to be traded through, not touched.
- All times are US Eastern. The usual time out, 15:30, is 2:30 Chicago, forty minutes inside a typical prop-firm close-out; a strategy may also hold for days.
- The years before you first used it govern. The coin flips run on the years before the year you say the strategy first existed; a result that only exists in the years since stops it.
- Every look at the data is counted. Your declared lineage is printed on the report and priced into the probability. Nobody gets to show you the winner of thirty.
- The exit you describe is the exit that runs. “No stop” on a public teardown is the baseline; the bracketed version is measured next, as its own strategy.
- Reproducible from the strategy text. Same data plus the printed strategy lands within tolerance of the printed number; every report prints the strategy as the engine read it.
- Every run is free, and no money comes from anyone who earns when you trade. No broker, no prop firm, nobody selling signals, courses or strategies; not as an ad, not as a sponsor, not as a referral link.
- If a sponsor ever appears, it will be named here, in its own marked place. Never on a result page or a share card, and never with a say in a score, a verdict or which results are published. Every verdict stays published either way.
- Today there is none. FirstRate Data, named on every page, is where we buy the prices; it is not a sponsor. The only kind we would take: a company that makes something traders use and does not earn when you trade.
If you trade it by eye. A score is a verdict on the written steps, taken every time with no judgment. What you do on top of them, the days you skip, the trades you cut short, the ones you take off-plan, is not in the score. It is in your fills, and the trade audit prices it in dollars against the same written rule: audit your trades →
The same method without the reasons, short enough to quote: the method, on one page →
How many of the shelf’s results luck alone would account for, worked out: what luck alone would pass →
Before the searches and the teardowns, we wrote down what we expected, and published how we did, misses included (the 300 runs of the shelf carry no predictions): the record of our predictions →
Test your strategy, free Every run is free.