I tested a “65% win-rate” Nasdaq 20/80 strategy — then compared it with 44 placebo versions
retained text
I tested the claim that NQ prices ending in **20 and 80** are unusually strong reaction levels. Rules: NQ futures NY morning 10-point stop 15-point target fade the 20/80 level Initial 5-minute test: 247 signals Resolved win rate: **46.4%** Claimed benchmark: **65%** 20/80 rank vs all 45 decimal-ending pairs: **23rd** Because many 5-minute trades were ambiguous, I reran it at 2-minute resolution: 119 signals Win rate: **41.0%** Even if EVERY ambiguous trade were counted as a winner: **50.4%** 20/80 placebo rank: **30th of 45** **29 arbitrary ending-pairs did better** A recent 1-minute spot check also produced **42.9%** and again ranked 20/80 **30th of 45**. So far I can’t find evidence that 20 and 80 themselves are special. This does **not** prove the trader’s full discretionary strategy doesn’t work — only that the measurable 20/80 level claim hasn’t survived these tests. I’d genuinely like people here to critique the methodology. If you spot a legitimate flaw, I’ll rerun it. What would you change?
document fetched 2026-10-05 00:38 · canonical https://reddit.com/r/algotrading/comments/1wxitxw/i_tested_a_65_winrate_nasdaq_2080_strategy_then