Patterns, honestly

After this lesson you'll know how to decide which chart patterns deserve your trust, and which are folklore wearing a serious name.

The zoo

Open any technical-analysis book and a zoo greets you: head-and-shoulders, double tops, cup with handle, flags, pennants, ascending triangles, falling wedges. Each has a confident description and a diagram drawn on a chart selected because the pattern worked there. This is the point in the curriculum where you need a way to sort the zoo, because you'll be offered patterns for the rest of your trading life, mostly by people selling something.

Three questions do the sorting.

One: is there a mechanism? Whose orders create this shape, and why would they? The patterns this school has already taught you pass easily. A base is accumulation you can explain participant by participant (course 3), and a volatility contraction is supply exhausting, visible in the shrinking pullbacks and the drying volume (course 6). When a pattern is a restatement of auction mechanics, it inherits the mechanics' credibility. When the explanation comes down to "this shape tends to resolve upward", with no account of whose orders make the shape, there's nothing underneath the drawing to trust.

Two: can it be defined precisely enough to test? "A cup with a proper handle" that three practitioners draw three different ways can't be tested, and untestable claims stay folklore forever, however old they are. A pattern defined by measurable conditions — range tightness, pullback depths, volume behaviour, a trigger price — can be checked against history. School VIII teaches how, along with the traps of doing it badly.

Three: does it survive honest testing? The published evidence on classical named patterns is thin and mixed, and you should hold the whole zoo loosely. What does carry evidence, in the academic literature and across the lineage's combined records, is the underlying behaviour the good patterns summarize: trends persist (the momentum effect, documented across decades and markets), and tight consolidations within trends resolve in the trend's direction more often than chance. The base-then-breakout structures this Academy teaches sit on that foundation, and even so, every trade wears a stop, because "more often than chance" loses plenty.

The working stance

Use pattern names as vocabulary, and mechanisms as the actual content. When another trader says "flag," you'll know they mean a brief tight pause after a sharp advance; useful shorthand. But your decisions rest on the questions underneath: which stage, what's the volume saying, where's the level that proves the idea wrong, how extended is it. Knowing fifty pattern names without the mechanics underneath them is worth very little, while a trader who has the mechanics and can't remember what the shapes are called is only awkward to talk shop with.

What it cost us to find out

This Academy's research desk spent a long stretch trying to rescue one named pattern: the opening-range breakout, a shape with a genuine mechanism behind it and a large literature of confident write-ups. We tested it about 1,300 ways across ten families of mechanism: entry rules, filters, universes, exit ladders, timeframes. The best result anywhere in that sweep reached a t-statistic of 2.50. Then we measured what a sweep of that size returns on data containing no edge at all, and the answer came back 2.80. Our search had surfaced less than chance would have.

That last number is the part worth carrying out of this lesson. Run enough variants of anything and the winner will look convincing, because you kept the best of many tries and reported it as though it were your only one. The claim we make from that work is deliberately narrow: this desk, on this data, could not make that pattern pay. School VIII, course 2 hands you the same measurement so you can run it on any pattern anyone asks you to believe, including ours.

Check yourself

1. Run "resistance becomes support" through the three questions. (Mechanism: regret and unfinished business becoming orders — yes. Definable: a broken level retested — yes. Evidence: consistent with documented trend persistence, and testable per stock. It sorts high.)
2. A course sells a proprietary pattern with an 85% success claim and no definition precise enough to test. Which question kills it, and why does the imprecision matter? (The second. An untestable claim can never be checked, which is precisely what makes it safe to sell.)
3. Why does this Academy still teach bases and contractions after this lesson's skepticism? (They're restatements of auction mechanics with testable definitions, sitting on the documented persistence of trends — they pass the three questions the zoo fails.)

The idea this lesson installs

A pattern earns trust through its mechanism and its testing, never through its name.

This completes School III. Next: School IV — Relative Strength, the Academy's core faculty.