A wall publishes. The next session's close lands somewhere. There are only four places it can land relative to the wall — and which one happens decides whether the publication held or didn't.
That's the whole framework. Four discrete outcomes, scored mechanically the same way every time, banked alongside the raw measurements so the rate is reproducible. This is the layer underneath everything Runir publishes about itself.
The four outcomes
A wall isn't a single price line — it's a tight zone around the published level, sized to the stock's typical daily range. Inside that zone, the wall is considered defended. Outside it, the wall is considered exceeded.
Every published wall gets graded the next completed session into one of four classes:
- Untested. The session's high (for a call wall) or low (for a put wall) never reached the zone. The wall was published, price went elsewhere, the level was irrelevant.
- Rejected. Price touched the zone, then closed back away from it. Dealer hedging defended the level in the most visible way — the wall acted exactly as the framework expected.
- Pinned. Price closed inside the zone. The wall held not by repelling price but by absorbing it. Dealer flow pinned the close to the level.
- Broke. Price closed clear past the zone. The wall failed.
Three of those count as a held wall: Untested, Rejected, and Pinned. One counts as a break. But only Rejected and Pinned represent the wall actually being tested — Untested walls were never challenged. Runir publishes the rate as Held over Tested, so untested walls don't inflate the headline number. A service that grades every wall it never tests as a “hit” is juicing its own number; we don't.
Why regime decides which class is likely
Dealer hedging behavior switches depending on which side of the gamma flip spot is on. In the pinning regime (above the flip), dealers sell into rallies and buy into dips — their flow dampens volatility. In the chasing regime (below the flip), dealers buy into rallies and sell into dips — their flow amplifies it.
The four classes populate differently in each regime:
- Pinning regime. Pinned and Rejected dominate. Walls do roughly what walls are supposed to do — they act as resistance the close respects.
- Chasing regime. Broke shows up disproportionately. Dealer flow is amplifying the move that's already breaking the wall, so the same setup that would've pinned in a different regime breaks decisively in this one.
Knowing the regime tells you which buckets to weight in your prior — not which way the next session will go. The framework gives you a base rate, not a forecast.
Why this is the mental model that matters
The wrong question is “will this wall hold tomorrow?” That question has no honest answer in advance. Anyone who tells you they know is selling something.
The right question is: of walls like this one — same regime, same persistence, same distance from spot — how often have they historically held? That has an answer. The answer is a frequency, not a guarantee. But a frequency, applied consistently across many decisions, is the entire difference between a methodical trader and one running on vibes.
The four-class taxonomy is what makes the frequency honest. Every grade traces back to raw measurements that anyone can re-derive. The classifier doesn't take a side, doesn't round in our favor, and runs the same way on Monday that it runs on Friday. You can audit the rate; you can split it by regime or by persistence; the math still works.
What it doesn't tell you
Frequencies are not certainties. A wall that pinned for six sessions can break on the seventh — and historically, those are exactly the sessions where traders relying on a wall as a guarantee get hurt. The framework gives base rates. Base rates are valuable because they help size decisions correctly across many trades, not because they promise what one trade will do.
At small sample sizes, frequencies are noise. A 75% hold rate at n=8 is statistically indistinguishable from a coin flip. Runir doesn't publish per-ticker rates until the sample is meaningful — the cutoff and the live universe-wide rate live on the methodology page, with the sample size next to the number.
And the four classes are descriptive of what happened, not prescriptive of what to do. The framework reports outcomes. Sizing a trade against those outcomes is the trader's job — nobody can do that step for you and remain honest.
How to read the card
Open the gamma board and the universe is ranked by regime — pinning names at the top, chasing names at the bottom. Each per-name page surfaces a reliability section once enough sessions have been scored for that ticker. The Friday Accountability Card aggregates the week's class counts across the whole universe so you can see the base rates compounding in real time.
The next stop is the gamma flip explainer — the regime call that decides which class dominates — or back to the call wall if you came in cold and want the level the framework grades against.