KILLED · spurious
Hypothesis
Full-moon / new-moon phases (and other calendar folklore) correlate with crypto reversals.
Math
$$ \bar r_{\phi} \text{ by lunar phase bucket } \phi,\ \text{with FDR control} $$
Method
Returns bucketed by lunar phase, full-moon windows tested for reversal bias; false-discovery-rate controlled across many calendar hypotheses.
Results
| Raw lunar "effect" | p≈0.04 (uncorrected) |
| After FDR over calendar tests | not significant |
A textbook spurious-correlation honeypot. The lone "significant" calendar effect vanishes under false-discovery-rate control across the many calendar hypotheses tested. Included here as a discipline example: we test folklore so we can kill it with numbers, not opinion.
Test enough calendar quirks and one will hit p<0.05 by chance. FDR control is how you tell a discovery from a coincidence.