Numbers about the past, not promises.

What If makes it easy to check an idea against history. It does not make the idea safe. Read this before you put money on anything you learn here.

Last updated 19 October 2026

The past is not the future

A trade that was profitable 80% of the time since 2017 can lose this year. Crypto markets have been through only a couple of full cycles, so even nine years of data is a short history.

Small samples

Consecutive days overlap: a condition that held for 300 days in a row is really one stretch of the market, not 300 chances. We count independent episodes and warn when there are fewer than ten. When you see that warning, treat the percentages as anecdotes, not odds.

The spread matters more than the average

An average of +116% can hide a worst case of −92%. Look at the median, the worst entry and the typical dip before exit, not just the headline percentage. If you could not sit through the worst case, the odds do not matter.

Data limits

We use daily closing prices from Binance spot and Hyperliquid perps. Perp prices track spot closely but not exactly. Newer coins have little history; the site tells you how many days it has. Data can arrive late or contain errors.

The rules can misread you

An AI model turns your sentence into rules. It can misunderstand a word, a number or a time frame. Always read the “Understood” line; if it is not what you meant, rephrase.

Your decisions are yours

What If never trades. Whatever you do on an exchange, with leverage or without, is your decision. Only risk money you can afford to lose completely. Laws and taxes on crypto trading differ by country and change often; you are responsible for both. See the terms.