The engine has been torn down and rebuilt piece by piece on the road to launch. This update is the last piece: the report card itself — what a score is made of, and where the grade lines sit.
Four things. A redundant subject is gone: how often a strategy trades used to earn points of its own, on top of already being counted in the small-sample discount and the evidence bar — being graded three times for the same virtue is a tax, not a measurement. Profit speaks louder: growth now carries more of the grade, because a strategy is ultimately a machine for producing return. The ceilings are honest again: the old scale's top end was calibrated for a different era — so many cards were hitting the maximum on the growth subject that the best strategies had become indistinguishable. The ceilings were re-drawn from the real distribution of today's warehouse. And there's now a floor: a strategy that trades too rarely to tell skill from luck (fewer than about thirty times a year), or whose yearly earnings don't at least match its worst fall, doesn't get a lower grade — it doesn't become a card at all. The Four Gates covers the floor in plain terms.
Every card in the warehouse was re-graded under the new scale. In last month's recalibration we promised scores would only go up, and we kept it — that fix removed an unfair tax, so upward was the honest direction. This time is different, and we'd rather say so plainly: some scores went down. When the ruler itself is redrawn, protecting every old number would mean the ruler is decoration — and a grade that can never fall isn't a grade, it's a souvenir. We're doing this now, before launch, precisely so that no one ever buys a card whose badge was grandfathered in by an obsolete formula.
Because we tested the scale itself before stamping it. Every strategy's final months are sealed away from breeding and only opened at graduation — which gives us thousands of honest measurements of how strategies performed on data they never saw. Held against those sealed results, the score does what a score should: the higher the grade, the more often the strategy stayed profitable on the unseen stretch — roughly half the time in the lowest bands, rising steadily to the vast majority at the top. The rebuild sharpens that relationship instead of leaving it to luck.
New formula, new distribution — so the grade lines moved to keep each seal meaning what it meant: Rare now starts at 66, Unique at 75, Epic at 80, Legendary at 86. The pyramid keeps its shape — Legendary remains roughly the top fraction of a percent of the warehouse — and card prices per badge are unchanged. As always, the record worth trusting most is Since Launch, written live on every card, every day.
A better scale grades the past more honestly — it doesn't predict the future. Every result on CoinLAB is hypothetical performance on historical data, and nothing here is financial advice. Always do your own research.