Your trading rule, tested honestly
Most backtests are built to look good: no fees, the best settings picked afterwards, coins that died left out. Send us your rule in plain words and we test it the way we test our own, where 91 of the last 101 failed.
Launch price: US$9 per rule for the first ten orders, then the price goes up. You pay only after we confirm the rule can be tested.
How it works
Three working days from payment to report.
Describe the rule
When to enter, which coins, when to exit. One or two sentences is enough. We reply within a day to confirm it is testable, or to ask what is unclear.
We lock it
The exact rule and the bar it must clear go into a file whose fingerprint we send you before any result exists. Nothing can be adjusted after that, by us or by you.
You get the report
Results on data the rule never saw, after fees and funding, with the checks below and a plain verdict. If it fails, the report says why.
What the report contains
- Seen and unseen halves. The rule is judged on the half of the data it was not built on.
- Real costs. Exchange fees and funding on every trade, then again with costs raised.
- Luck check. The same trades at random moments, so you see what zero skill looks like.
- Fragility checks. The result without its five best weeks, one bar late, and year by year.
- The typical trade. Median and win rate, not only the average.
- The lock file and its fingerprint, so you can show anyone the test was fixed in advance.
What we can test
- Rules on crypto perpetual futures, 2020 to today, about 680 coins including the ones that were delisted.
- Signals from price, volume, funding, open interest, long/short ratios and order-book depth, or from dated public events.
- Holding periods from four hours to a few weeks.
What we cannot
- Rules that depend on judgment (“when the chart looks strong”).
- Stocks, forex, options, or holding periods of minutes.
- Anything that needs data we would have to pay for.
If we cannot test it we say so and you are not charged.
Send us a rule
No payment now. We read it, reply by email, and send a payment link only if it is testable. Your rule and its result stay private unless you tell us we may publish them.
A report measures how a rule behaved on past data. It is general information, not financial advice, and it is not a recommendation to trade the rule. Past results do not predict future results. See the ledger for how our own rules fared.