Your report
The lines that matter, and what each one means
Real lines from a real report. Press "why?" on any of them.
SCORE 90 / 100 VERDICT VALIDATED
Six checks, added up. Seventy or more with no weak spot is Validated.
Think of it like a safety inspection with six stations. A car can't pass by acing the brakes if the steering fails. Validated means the rule cleared every station reasonably and most of them well. Marginal means it's not nothing, but something specific is broken, and the report names it. Rejected means don't. Protocol v1.1 bands: 70–100 Validated (no gate, no pillar below 50%), 40–69 Marginal, under 40 Rejected. Gates override points: insufficient sample, look-ahead, reproduction failure, era-manufactured edge. The examiner may lower a verdict with a written reason, never raise one.Unseen era p = 0.0030 (15 / 5000)
On years you never looked at, only 15 of 5,000 coin-flip twins did as well as your rule.
The smaller this number, the less likely your result was luck. 0.05 means one twin in twenty matched you, which is the usual bar. 0.003 means three in a thousand. Above 0.20 or so, your rule is indistinguishable from flipping a coin on the same days. Empirical p: share of 5,000 random-side seeds with net ≥ real net, computed on the unseen era only. It is fatter-tailed than a normal approximation would suggest and we report the count, not a fitted z. A lineage-adjusted p appears on the next line.Breakeven 17.0 ticks/side
Slippage would have to reach 17 ticks on every entry and exit before the rule stops making money.
Every time you trade, you lose a little to the gap between buying and selling prices. One tick is normal; three is a rough day. A rule that breaks even at 17 ticks has a lot of room. A rule that breaks even at 2 ticks is a rule that works on paper and not in your account. Searched by full re-simulation at integer ticks per side with linear interpolation between rungs; for bracketed rules, exits are re-resolved at each rung. Under 4 ticks fails the code-offer gate.Largest year 2022 = 44.7% of unseen-era net
Almost half the profit in the honest test came from one year.
A rule can be real and still be one you'd never have stuck with. If most of the money comes from one wild year, you'd have spent the other years wondering why you bother, and most people quit before the good year shows up. This line is the difference between "it works" and "you could have traded it." Era-consistency exam: calendar years in the unseen era as windows; share profitable, PF retention unseen/seen, and max single-year share of unseen net (over 40% costs points). Reported beside top-trade and top-5 share, negative rolling 12-month share, and DD95/actual from order shuffle.Lineage-adjusted p (variant #3) = 0.0090
You said this was your third version, so the odds are corrected for having picked the best of three.
If you roll a die three times and keep the best roll, the best roll isn't a fair picture of the die. Same with strategies. The more versions you tried, the more we have to discount the winner. This line does that math using the number you gave us. Multiplicity: 1 − (1 − p)^n with n = effective variant. Sidak-style, conservative for correlated variants, and the best available correction when the search history is self-reported. The code-offer gate reads this line, not the raw p.CODE OFFER => OFFERED
Four conditions cleared, so the rule comes with a Python file you can run yourself.
If a rule is Validated, beats the coin flips on the honest years even after the lineage correction, has room for slippage, and doesn't depend on one trade, we hand you the rule as a program. It's the same code we used to test it, so what you run is what we scored. It is not a trading robot and it doesn't place orders. Gate: Validated; unseen lineage-adjusted p ≤ 0.05; unseen bootstrap 95% CI excludes zero; breakeven ≥ 4 ticks/side; top trade ≤ 25% of net. The file is generated by lifting the exact signal and simulation functions from the engine, then verified against the engine on the same data to the cent before delivery. Standard library only.