How to read the map: what each pattern is measured to mean
Most gamma commentary names shapes and moves on. We named ours, then graded every one of them against the control almost nobody runs: not 'how often did this level hold' but 'how often did it hold compared with any level the same distance from price'. Four of the seven descriptions turned out to be wrong, including one that meant the reverse of what it said. Here is what each pattern is measured to mean.
The map tells you HOW MUCH, not which way
Graded on next-day direction, every pattern reads about 52% higher. So does a board with no pattern at all, because equities drift up. Benchmark a direction rate against 50% instead of that baseline and you manufacture an edge that is not there.
Where the patterns genuinely separate is dispersion. Ordered by the range that followed, across a five-year split: the Vise ran about 1.1%, the Lid, Bookends and Fortress about 1.8%, a board with no pattern about 2.0%, Live Wire about 2.2%, and Open Water about 2.4% — the ordering held in both halves of the sample.
That is the read in one sentence: the name tells you how much price is likely to move, not which way it goes.
The Lid: the strongest thing on the map
A heavy positive level overhead with amplifying gamma underneath. In this shape the lid went unbroken 84.8% of following sessions against 75.5% for any level the same distance out, across 70,562 boards, in both halves of the sample.
It is also the only pattern edge that survives inside the session — six points or more over the next two hours. Every other level edge we found is an overnight effect. Note the edge belongs to the Lid pattern — the shape — not to every call wall on every board.
The Fortress means the opposite of what it looks like
One strike dwarfing the board looks like a destination. It is the reverse. Price touched a dominant Apex about 23% of following sessions against about 35% for any level the same distance away, on 50,653 boards — and the same shape shows up intraday.
The mechanism is the one that runs through this whole map: heavy structure damps movement. A still Apex that big is a reason the session stays quiet and stays away, not a magnet pulling price in.
The Vise is not a box, and Bookends are not intraday levels
The Vise marks the narrowest range we measure. It does not contain price: over five years, price finished between the two nodes under 2% of the time. Expect a slow session, not a box.
Bookends' upper edge holds about 78% of the next session against about 71% matched — but that edge does not reproduce intraday, where it inverts between halves. Read Bookends as an overnight reference for the upper edge only, not as levels to trade against during the day.
The midpoint between two nodes is not a level either. Neither is the node: asked whether price turned back once it reached them, both came in at about 51%, which is a coin flip. Levels are not bounce points.
Two more shapes are stand-down reads. The Whipsaw is a wide bracket — two heavy edges with a void between — and the edges are not bounce points any more than any other level, so there is nothing to be right against. Rainbow Road is nodes of both signs scattered with nothing dominant: structure everywhere and none of it holds. Both are the map saying no, for opposite reasons.
What moves the odds is the path, not size
Whether price reaches a level has almost nothing to do with how big that level is. It depends on distance, measured against the symbol's own range, and on what the path looks like. Two things on the path made a level harder to reach when we graded it: a low-gamma pocket between price and a nearby level, and a sign flip between price and a far one. We call either one a checkpoint.
A pocket is not a fast lane. The popular story says price travels quickly through thin gamma; what we measured is that a nearby level with a pocket in front of it gets reached less often. That is why the map reads the path rather than the size of the level, and why a level is described as in reach, never as a target.
Change is the only directional read
Where the size SITS predicts nothing. Where the book is being WRITTEN does. When exposure above price grew relative to below over the prior half hour, price traded higher thirty minutes later about 58% of the time against about 48% when the tilt ran the other way — measured across 27 sessions with the prior move held flat, so it is not momentum in disguise. It is thin, and it fades by the hour mark.
The Apex follows the same rule. A still Apex is not a magnet. An Apex being built within 1% of price was touched about 52% of the time against about 45% for the level the same distance on the other side; one being unwound was avoided — price crossed it about 7% of the time against about 16% normally. And when the Apex stepped to a new strike ahead of price, the move continued over the next half hour 54–59% of the time against 43–47% when it stepped behind. All three are about change, and all three held in both halves of the sample.
| Pattern | What it means | Measured |
|---|---|---|
| The Lid | Lid overhead holds; drift beneath it | 84.8% vs 75.5% matched; holds intraday too |
| Fortress | Price stays AWAY from the dominant Apex | Touched ~23% vs ~35% matched |
| Bookends | Upper edge, next session only | Upper edge ~78% vs ~71%; inverts intraday |
| The Vise | Slowest sessions — but not a box | ~1.1% range; inside under 2% of the time |
| Live Wire | Amplifying structure at price; size down | ~2.2% range |
| Open Water | Nothing big enough to lean on | ~2.4% range — the widest state |
| Springboard | Shelf below, thin air above — rare | Has gone both ways; rarely seen on our maps |
| Whipsaw | Two heavy edges, void between | Edges are not bounce points; stand down |
| Rainbow Road | Both signs scattered, nothing dominant | Stand down — a chop tape, not a setup |
Common questions
Does the biggest node pull price toward it?
Not while it sits still. Measured against a distance-matched control, a dominant Apex was touched about 23% of sessions versus about 35% for an arbitrary level the same distance away. The exception is change: an Apex being built within 1% of price was reached more often (about 52% vs 45%), and one being unwound was avoided.
Is the midpoint between two nodes a support level?
No, and neither is the node. Conditional on price reaching them, both turned price back about 51% of the time — indistinguishable from chance. A node's measurable value is range compression, not reversal.
Can the map tell me which way price is going?
Only through change. Growth of exposure above price relative to below predicted a higher price thirty minutes later about 58% of the time against about 48% the other way; an Apex being built nearby was reached more often; an Apex stepping ahead of price came before the move continuing. Where the size already sits describes how much price moves, not the direction.
Read the pattern name as a speed limit, not a direction. Two shapes carry a measured edge — the Lid pattern, which holds on both clocks, and the Fortress, which is reached less rather than more. Direction only ever comes from change: the book being written, the Apex building or unwinding, the Apex stepping ahead of price.
Keep reading
See it on a live board
FlowMonkey prices every level by whether your ticker can actually reach it today — measured on that symbol's own range, re-sorted as the session burns down.
See pricing