THE STANDARD / HOW WE GRADESAME TEST FOR EVERY STRATEGY
What a verdict on this site actually means
A stamp here is not an opinion. Each one is earned by hitting a fixed, written rule, and every strategy goes through the identical test. That is the whole point: the results are comparable to each other, and you can check our working.
01 The three verdicts
Failed
Rule: lost money outright, OR beat buy-and-hold on 3 or fewer of the 10 coins.
It didn't just disappoint. It lost money, or it's clearly worse than doing nothing on almost every coin. The famous name on the tin doesn't save it.
No edge
Rule: beat buy-and-hold on 4 to 6 of the 10 coins.
A coin flip. It wins on some coins, loses on others, and you can't tell the result apart from luck, except that you paid fees the whole way for the privilege. This is the most common and most seductive outcome, because it feels like discipline.
Survived
Rule: beat buy-and-hold on 7 or more of the 10 coins AND stayed ahead in the out-of-sample period.
Rare. It beat doing nothing across most coins, and it kept beating it on data the rules were never fitted to. Not a money printer — usually a low win rate and an ugly ride — but a real, measured edge rather than a bull-market coincidence. On this whole site, only a couple clear it.
02 The test, the same for all of them
Every strategy is run through the identical harness. Nothing is hand-tuned per strategy, because the moment you do that you're fitting to the answer.
What it trades
Spot only, long or flat. No leverage, no shorting. The plainest possible version of the rule, so we're judging the idea, not a risk setting.
Data
Daily candles over each coin's full Binance history, on 10 major coins, with a warm-up before the first trade.
No look-ahead
A signal read at one day's close is filled at the next close, a full day later. The strategy can never trade on information it couldn't have had at the time. This one rule quietly kills most backtests you see online.
Costs
A 0.06% fee on every entry and exit. Buy-and-hold is charged one entry fee too, so the comparison is fair.
The benchmark
Every result is measured against simply buying the coin once and never touching it. "Made money" is not the bar. "Beat doing nothing, after fees" is.
The honest test
The most recent ~45% of each coin's history is held back as out-of-sample. A strategy can look brilliant on the years it was built around; the real question is whether it still works on data it never saw. That's what separates an edge from a lucky fit.
03 What a green "Survived" does NOT mean
A survivor cleared a high bar, and no more than that. Being straight about the limits is the point of the site, so here they are.
The 10 coins are today's majors, which is survivorship bias baked in (the coins that died aren't in the list). They're also heavily correlated, so "beat on 7 of 10" is not 7 independent tests. There's no live slippage, no funding, no shorts, spot long-only. And "held up in the past" is never a promise about the next cycle. Past performance is not the future. A survivor is a strategy that didn't fall apart the way the famous ones did — that's all a green stamp claims.
04 Why we publish the rules at all
Because a verdict you can't check is just an opinion, and the internet has enough of those. Every number on this site comes out of the same script under the rules above. Soon you'll be able to run your own strategy through the exact same test and get the exact same kind of verdict, so you never have to take our word for it.