Why does every pattern look obvious afterwards?
This is the lesson the rest of the course was built to support. Everything you've learned about shapes runs into one hard problem: human beings are pattern-detection machines that do not switch off in the absence of patterns.
Apophenia: seeing structure in noise
Apophenia is the tendency to perceive meaningful connections in random data. It's not a defect — it's the same faculty that lets you recognise a face in poor light or a predator in tall grass. Erring toward false positives was cheap; erring toward false negatives was fatal.
Price charts are close to an ideal trigger for it. They are noisy, they are visual, they have local structure (volatility clusters, so quiet follows quiet), and they're presented at exactly the scale where the eye works hardest. If you wanted to design an apophenia machine, you'd design a candlestick chart.
The demonstration is easy and worth doing yourself. Build a random walk — a few hundred rows of a coin flip, plus one or minus one, cumulatively summed — and plot it as a line. You will find flags. You will find triangles, double bottoms, head-and-shoulders shapes, clean support levels that "held" three times. All of them, in abundance, in data that by construction contains no information whatsoever.
This does not prove chart patterns are useless. It proves something narrower and more important: the fact that you can see a pattern is not evidence that the pattern means anything. Your eye would have found one either way.
Hindsight bias: the chart reorganises around the outcome
The second problem is worse, because it feels like insight.
Once you know how a chart ended, the preceding price action rearranges itself into an obvious story. The head and shoulders is unmistakable — because you can see the decline that followed it. The flag is textbook — because you can see the continuation. You are not remembering what you thought; you're reconstructing what you would have thought, using information you didn't have.
The antidote is physical: cover the right side of the chart. Put your hand over everything after a date, decide what you see, write it down, then reveal. Almost everyone finds this humbling the first time. The clarity that seemed to come from the chart was coming from the future.
Selection bias in every pattern book ever printed
Open any book of chart patterns. Every example works. Of course it does — the author chose them because they worked. Nobody publishes the two hundred flags that went nowhere, because a chapter of failures doesn't sell.
The same filter operates on social media, in trading courses, and in your own memory. You remember the pattern you spotted that paid off. The ones that didn't get filed as "market was weird that week."
The degrees-of-freedom problem
Finally, the arithmetic of choice. Before you draw a single line you have selected:
- The timeframe (daily, weekly, four-hour) — different candles, different patterns.
- The scale (linear or logarithmic) — every trendline changes shape.
- The start date of the window you're viewing.
- Where exactly to place the lines, and how much slack to allow.
Each choice multiplies the number of patterns you can find. With enough freedom, a pattern can be found in anything — which means finding one costs you nothing and therefore proves nothing. The only escape is to fix the choices in advance and then count. That's the next lesson.
Try it now
- Do it on the chart above, physically. Cover everything to the right of a date roughly a year in, write down what pattern you believe is forming and what it implies, then move your hand. Repeat at five different dates and keep score honestly.
- Now change one choice and watch the patterns change with it. The chart below is the same instrument over a shorter window — different bars at the edges, a different visible range, and therefore a different set of shapes you can find. Nothing about the market moved.
- Failure hunt, on yourself: find one instance from your practice so far where you were confident about a pattern and wrong. Write down what made you confident. That sentence is the bias, in your own handwriting.