---
id: CB-WP-0039
kind: product
title: "A seat that does not regulate"
status: done
state_hub_workstream_id: "8387cea3-fcaa-4d34-89f4-527074db1514"
---

# Purpose

```
structural tier  M   (adds a policy to the panel every future difficulty
                      and design measurement is reported against)
chaos            d8 = 2  → no override
declared tier    M
```

**Declaration 11 of chaos window 3.**

## Why

[CB-EV-0030](../evidence/CB-EV-0030-h1-measured.md) §4 named this as the
measurement it had not run, and it is the one that decides whether H1's
verdict is about **H1** or about **greedy**.

H1 assumes rising Stress pushes a table toward the DARVO arm. It does not,
**because greedy regulates**: `Action::Ground if gated => 100` makes
GROUND the top choice the moment the stress gate bites, so Stress
plateaus at 3 and neither the gate (4) nor the arm (5) is ever reached.

> **Every policy in that panel was greedy-family.** So "H1 does nothing"
> may be "H1 does nothing *to a seat that already manages its Stress*" —
> and H1 was written for the seat that does not.

## The one preference that varies

`Reactive` is `GreedyPolicy` with **GROUND demoted below ATTACK**, and
nothing else changed. It is the player who chases Problems and ignores
their own state — which is not a straw man: it is the commonest way a
person plays a co-operative game badly, and the behaviour GROUND's whole
Stress economy exists to punish.

**One preference, per ADR-0018 and the same discipline as
`attack-value.rs`'s `Attacker`.** A policy that differed in several places
would tell us a variant is different without saying which difference did
the work.

## Task: the policy, and the measurement

```task
id: CB-WP-0039-T01
status: done
priority: high
state_hub_task_id: "2c509cdb-aedb-4dce-a21f-f16bcdf17d80"
```

**Controls:**
- **exactly one preference differs from greedy**, and the artifact says
  which;
- **both variants, same instrument, same run** — otherwise the comparison
  is between harnesses;
- **the game count is reported**, because CB-EV-0030 §3's harness defect
  was silently discarding games and printing zeros that read as losses;
- **the answer may be that H1 still fails**, and that is reported as
  plainly as the other outcome.

**Done 2026-08-08, then CORRECTED after adversarial review**
([CB-REV-0001](../reviews/CB-REV-0001-h1.md)).
[CB-EV-0031](../evidence/CB-EV-0031-a-seat-that-does-not-regulate.md).

> **This task's own control — "exactly one preference differs from
> greedy" — was false when it was marked done.** The policy differed in
> five places, including `SpendFreedom`, which made the seat burn its
> Freedom token in round one of every game: a second change to the exact
> mechanism under study. **Every number below was measuring that.**
>
> `GreedyPolicy::rank` is now public and the policy delegates to it,
> overriding one match arm — so the control is structurally true rather
> than asserted in a comment.

**It changed CB-EV-0030's verdict**, and that part survives: H1 does reach
the unregulated seat. DARVO arms **2 per seat per game**, scaling with the
table — the earlier flat "400 in every cell" was the broken policy.

**Criterion 3's verdict is corrected and was wrong**: reactive wins
**66 at 2p against greedy's 68**, not "nothing at any seat count". It wins
nothing at 3p and above, and so does greedy.

**The baseline observation survives only in a narrower form.** Peak Stress
*held* across the **1,600** baseline games was **2** — the starting value,
never exceeded. The original said **1** because the metric maximised over
`StressSet` event payloads and the starting value is written by `setup`,
never by an event: **wrong subject**, in the pass that was written to
correct a wrong-subject error.

And the generalisation was refuted by this repo's own harness: the rank-95
policy *is* a reckless policy and it drives Stress to 5 under **baseline**.
What survives: under the two policies here, neither of which selects ATTACK
under baseline, **no Stress is ever added** — ATTACK is the sole inbound
pressure.

**The H1-B suppression claim is WITHDRAWN.** It was inferred from a
constant that the broken policy produced, and disabling H1-B under the
corrected policy changes the arm count by **exactly zero** in every cell.
The inference was also invalid on its own terms: final Stress does not
record who armed, because `DarvoEnded` resets the stage and REVERSE gives
its owner −2.

**The pass hedged the wrong variable** — it disclaimed "the number 2" and
defended "the direction", and the direction is what failed.