THE METHOD SCOREBOARD

How well do timing theories actually predict crypto?

We took the classic market-timing methods — geometry, astronomy, Fibonacci — and scored them against real price history the way a skeptic would. Here's every result, edge or no edge.

How we score — and why most methods fail

A hit rate on its own is meaningless. Every method here is put through two tests. Monte-Carlo control: we score thousands of random dates or price levels and ask whether the real method beats them (p < 0.05); if random does just as well, there's no edge. Split-half replication: a genuine effect appears in both halves of the timeline independently — the single test that exposed our one promising-looking result as a false positive. Sample: 1,417 days of real XRP-USD closes, 152 detected turns. Full methodology and code live in the blog and xrp-astronomy/.

See the live version on your own coins

The Command Deck draws these same levels — Fibonacci, Gann, lunar, confluence — on any of the top-100 Coinbase coins, in real time. Practice pre-committing your watch dates and score yourself.