I ran a small experiment on myself a while back that I still think about. I generated a few hundred price charts from a random walk, no real market data anywhere in them, just a starting price and a coin flip at every step deciding whether the next candle went up or down. Then I flipped through them the way I would flip through a real watchlist, and I caught myself doing the thing. Clean double bottom on this one. Nice ascending triangle here. That one is coiling for a breakout. Every instinct I have built over years of staring at real charts fired exactly the same way it would on a live symbol. The charts meant nothing. My brain did not care.
That is apophenia, the human tendency to perceive meaningful patterns in random or meaningless data. It is not a trader flaw specifically. It is the same wiring that makes us see faces in clouds and hear our name in a noisy room. For most of human history it was a good trade, so to speak. The cost of imagining a predator that was not there was low, and the cost of missing a predator that was there was death. So we over-detect. We are pattern machines tuned to false positives, and financial charts are close to the worst possible environment to point that machine at, because price series genuinely do contain a lot of noise that looks structured after the fact.
Why price charts are a trap for this specific bias
A random walk is not visually random the way TV static is. It trends. It reverses. It puts in what look like clean tops and bottoms, it respects lines you draw on it for a while and then abandons them, and it produces stretches that look like textbook accumulation followed by textbook breakouts. This is a mathematical property of cumulative random processes, not a coincidence. If you sum up a series of small independent moves, you get long runs, apparent support and resistance, and formations that a chartist would name without hesitation.
So when someone shows you a beautiful head and shoulders on a real chart and it played out, the honest question is whether the pattern predicted the move or whether you are looking at one of the many arrangements that noise produces anyway, selected after the fact because it happened to work. Real markets are not pure random walks, there is real structure in there from flow, liquidity, and human behavior. But the amount of genuine signal is much smaller than the amount of pattern your eye reports, and separating the two by eye alone is close to impossible.
Run the test on yourself
The most useful thing I can hand you here is a way to calibrate your own pattern perception, because reading about the bias does almost nothing to reduce it. You have to feel your own confidence fire on data you know is fake. Here is a version anyone can do without much tooling.
- Generate synthetic charts. A random walk is enough to start. In a spreadsheet, put a starting price in the first cell, then each next cell is the previous value times one plus a small random number that can be positive or negative. Chart a few hundred rows. For a more realistic feel, use a lognormal step so the series cannot go negative and the volatility looks market-like.
- Mix in a few real charts with the tickers and dates stripped off, same axis scaling, so you cannot tell which is which. This is the important part. If you only look at fakes you will unconsciously brace for them.
- Flip through the blind pile and mark your calls out loud or on paper. Support here, breakout forming here, this one looks bearish. Commit to a direction as if you were about to size a trade.
- Then reveal which charts were synthetic. Count how often you named a confident setup on pure noise, and how often your directional call on the fakes was right, which over enough samples should land near a coin flip.
The first time you do this it is humbling. You will find that your confidence does not distinguish real from fake nearly as well as you assumed. That gap between how sure you felt and how well you actually sorted the piles is the exact size of the problem you are trying to manage.
What to demand before you act on geometry
The takeaway is not that technical analysis is worthless. It is that chart shape on its own is a hypothesis, not a signal, and it needs a second and ideally a third independent source before it earns a position. Independent is the operative word. Drawing three more trendlines on the same price series is not confirmation, it is the same data wearing a different hat.
The rule of thumb I use is that a pattern gets to graduate from interesting to actionable only when something outside the price geometry agrees with it. A few of the sources I lean on, roughly in order of how much I trust them:
- Volume and order flow. Does the breakout come with participation, or is it a drift on thin volume that noise produces for free. A move that nobody is trading is usually not a move.
- Positioning and flow data. What are large wallets, funds, or disclosed insiders actually doing around this level. Real money moving in the direction of your pattern is worth more than the pattern.
- Fundamentals or a catalyst. Is there a reason for the level to matter, an earnings event, a supply unlock, a policy shift, or is the line just a place your eye landed.
- Cross-market confirmation. Do related instruments, correlated names, or the broader risk backdrop tell the same story, or is this one chart alone in claiming something is happening.
If the geometry is the only thing you have, treat that as a warning, not a green light. The classic failure mode is the trader who finds a gorgeous setup, cannot articulate a single non-chart reason it should work, sizes up anyway because it looks so clean, and then blames execution when it fails. It did not fail on execution. It was a picture of noise that his eye promoted to a plan.
Building the doubt into your process
You cannot turn apophenia off, so the move is to build friction around it. A short pre-trade checklist does most of the work. Before I commit to a chart-based idea I make myself answer, in one sentence each, what non-price evidence supports this, what would have to be true for me to be wrong, and whether I would still take this trade if the exact same shape appeared on a chart I had just been told was randomly generated. That last question is the one that kills the most bad trades, because a surprising number of setups do not survive it.
This is also why I care about having flow, positioning, and disclosure data sitting right next to the price chart rather than in a separate tab I have to go hunting for. When the confirming evidence is one glance away, you actually check it. When it is friction, you skip it and let the geometry win by default. That is part of why we built Blockcircle to surface whale wallet activity, insider and political disclosure feeds, and market scorecards alongside the chart, so the second opinion is already there when the pattern-matching part of your brain starts making promises.
None of this makes you immune. I still catch myself naming triangles on random walks, and I have been doing this long enough to know better. The realistic goal is not to stop seeing patterns, it is to stop trusting them on sight. Treat every clean chart as a claim that has to be checked against something real before your money agrees with your eyes.