Harmonic patterns, Gartley, Bat, Butterfly, Crab, use specific Fibonacci ratios between swing points to define reversal zones. Charts with perfect geometry look persuasive.
The Texas sharpshooter fallacy describes someone who fires at a barn wall, then paints a target around the tightest cluster of bullet holes. Harmonic patterns invite the same thing. With enough swing points and tolerance in the ratios, some pattern can be found at almost any turning point.
A rigorous approach fixes the tolerances in advance, scans automatically for patterns across a long dataset, and measures what follows each detection, compared with a baseline of randomly chosen turning points. Scanners exist, and the results of such tests, where published, are usually unimpressive.
The patterns may still provide useful structure for some traders, pushing them to wait for confirmation and define risk clearly. That discipline can matter more than the pattern itself.
The risk lies in treating a beautifully drawn pattern as evidence of an edge.
One safeguard is a pre-registered list. Write down the pattern, the entry, the stop and the target before the trade, timestamp it, and record the result. After a hundred entries, the success rate is a fact, not a feeling.
Has an automatic harmonic scanner been tested with fixed tolerances, and what did the hit rate look like?
Harmonic patterns and the Texas sharpshooter
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
-
FXS Prop Desk
- Posts: 1691
- Joined: Wed Oct 07, 2026 9:52 am
Re: Harmonic patterns and the Texas sharpshooter
The sharpshooter comparison fits because the reversal zone is drawn after the swings exist. A fair test fixes the ratio tolerances in advance, scans history for every pattern that qualifies, and counts the failures too. Most published harmonic examples are the winners only, which is the painted target.
FXS Official Team | The Idea Lab & Strategy
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Harmonic patterns and the Texas sharpshooter
I spent most of one winter drawing Gartleys and Bats by hand, and my screenshot folder was full of beautiful reversals. Then I bought a scanner and set it to fixed tolerances of three percent on every ratio, and ran it across several years of fifteen minute EURUSD and AUDUSD data.
The hit rate was humbling. Reaching the first target before the stop came in at around forty six percent, and when I compared it to reversals taken at random swing points with the same stop and target distances, the difference was a couple of percentage points at most. Most of what I had been calling skill was me picking out the nicest drawings after the fact.
I still use the reversal zones occasionally, but only as an area to watch for a separate trigger candle, never as a reason to trade on their own. And I log every scanner hit, ugly or not, because the ugly ones are where the honest statistics live.
The hit rate was humbling. Reaching the first target before the stop came in at around forty six percent, and when I compared it to reversals taken at random swing points with the same stop and target distances, the difference was a couple of percentage points at most. Most of what I had been calling skill was me picking out the nicest drawings after the fact.
I still use the reversal zones occasionally, but only as an area to watch for a separate trigger candle, never as a reason to trade on their own. And I log every scanner hit, ugly or not, because the ugly ones are where the honest statistics live.