coulomb/clay-borg
4
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

CB-WP-0024/0025: what play reported, split into a renderer and a search
Some checks failed
ci / check (push) Failing after 4s
Details

Browse source 
Five remarks from the maintainer's test games, checked against the code
before being written down — two had already reached ground-game on wrong
premises, so a claim now names the line that makes it true.

Three of the five turned out to be data the projection already carries,
drawn as text: solution_deck_len, solution_discard, and OutcomeView's
personal/mastery/winners. One control (`close — I have read this`) is
labelled as a reading but shuts the server down, and leaves a live-looking
page pointing at a dead port. One number does not exist at all: table.rs
loops run_game and keeps only the last summary.

CB-WP-0024 (S, chaos d8=6, declaration 7 of window 2) — the piles and the
other seats' plays as objects on the table, the ending control saying what
it does, and a tally that survives "play again".

CB-WP-0025 (L, chaos d8=6, declaration 8) — "could we have won" and "how
hard is this" are the same search asked twice. Tier L because the
information boundary is the whole design problem: a solver reading
GroundState sees the deck the rules hide, and would tell the maintainer he
could have won by playing a card he had no way to know was there. Also
unblocks GROUND-WP-0005, active with both tasks waiting on a measured
difficulty baseline.

Also: CB-WP-0022 T07's evidence file moves to CB-EV-0021 — CB-WP-0023
shipped CB-EV-0020 first.

loop-lint: no findings.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
tegwick 2026-08-05 12:59:08 +02:00
parent 43f94db42e
commit 8e9b3c19b7
3 changed files with 457 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

4
workplans/CB-WP-0022-the-design-instrument.md
View file

@ -294,7 +294,9 @@ priority: high
state_hub_task_id: "4fb77911-486d-4d9e-a53d-c4e486c74ce2" state_hub_task_id: "4fb77911-486d-4d9e-a53d-c4e486c74ce2"
``` ```
`evidence/CB-EV-0020-*.md`. `evidence/CB-EV-0021-*.md`. *(Was CB-EV-0020 when written; CB-WP-0023
shipped that number first — `evidence/CB-EV-0020-solve-legality.md` — so
this one moves rather than collides.)*
- **Whether backfilling changed the design** — if all six findings fit the - **Whether backfilling changed the design** — if all six findings fit the
first taxonomy, say so and be suspicious of it. first taxonomy, say so and be suspicious of it.

182
workplans/CB-WP-0024-the-table-you-can-watch.md Normal file
View file

@ -0,0 +1,182 @@
---
id: CB-WP-0024
kind: product
title: "The table you can watch: the piles, the other seats' moves, the score that carries"
status: ready
---
# Purpose
```
structural tier S (renders state the projection already carries, inside
cb-render-html and cb-play — no new port, no canonical
interface, no dependency)
chaos d8 = 6 → no override
declared tier S
```
Declaration 7 of chaos window 2.
## Five remarks from play, and what is actually true of each
The maintainer played several games and reported five things. Four are
this pass. Each was checked against the code before being written down,
because **two findings have already reached `ground-game` on wrong
premises** (CB-WP-0023 T01, GROUND-WP-0004 T02) and the rule now is that a
claim names the line that makes it true.
| # | reported | what the code says |
|---|---|---|
| 1 | *"the button next to new game says 'I need to read this' — why? the UI is not closing"* | `doc.rs:825` renders `close — I have read this`. `hotseat.rs:248` reads it as **shut the server down**. The label describes a reading; the control ends a process. And nothing closes: a page cannot close a tab it did not open, so the tab sits there showing a dead game against a dead port. |
| 4 | *"following the other players is only possible in the log"* | Wrong by degree, not in kind. `player_card` (`doc.rs:385`) draws each seat, and `selections` are `Shown` for every seat once `step != Select` (`view.rs`) — so the data is there and is being drawn as words. What is missing is that a seat's *play* is not drawn **as a card on the table**. |
| 4b | *"visualise the draw stack and the discard stack"* | Both already project: `solution_deck_len` and `solution_discard` (`view.rs`). `body()` prints them as one line of text (`doc.rs:575`). Nothing is missing from the model; the whole gap is rendering. |
| 5 | *"if we calculate scores, show them — and a cumulative score over several games"* | Per-game scores exist and **are already drawn**: `OutcomeView.personal`, `.mastery`, `.winners`, `.total`/`.threshold` (`doc.rs:582–628`). Cumulative does not exist at all: `table.rs:241` loops `run_game` and returns **only the last summary**. |
Remarks 2 (*how could we have won*) and 3 (*difficulty*) are **not this
pass** — they need a search and a measurement, not a renderer. They are
CB-WP-0025.
**So this pass is smaller than the remarks make it sound.** One control is
mislabelled, one number is missing, and three things that already project
are being written as text where a picture would read better. That is why
it is S: no new state leaves the aggregate, and `view.rs` gains nothing
except where task T04 says so explicitly.
## Task: the ending control says what it does, and the page acts on it
```task
id: CB-WP-0024-T01
status: todo
priority: high
```
Two defects behind one button.
**The label lies about the mechanism.** `done` does not mean *I have read
this*; it means *stop serving*. Rename it to what it does — the session
ends, the process exits — and drop the first-person reading claim. The
`data-drop` key stays `done`: `hotseat.rs:248` matches on the body and the
tests in `hotseat.rs:524` drive it.
**The page does not change.** After the POST the server replies `closed`
and breaks its loop; the tab keeps showing a live-looking table with a
`play again` control that now hits a closed port. Replace the page body on
acknowledgement with a terminal state that says the session has ended and
this tab can be closed — and **disable `play again` at the same moment**,
because a control that can no longer work must not look like it can.
**Controls:**
- a test asserts the acknowledged page no longer offers `again` — not just
that it shows different words;
- `hotseat.rs`'s existing `play_again_deals_a_second_game` must still
pass unchanged, or the key was renamed and the server was not told;
- the label change must be mutation-visible: assert the rendered text, so
reverting the wording turns a test red.
## Task: the piles are objects on the table
```task
id: CB-WP-0024-T02
status: todo
priority: high
```
Draw the draw stack and the discard stack as stacks, with their counts on
top, replacing the text line at `doc.rs:575`. `problem_svg` and
`relations_svg` are the precedent for how this repo draws — follow them
rather than inventing a second idiom.
The maintainer asked for one thing the model does not yet do: **when the
draw stack runs out, the discard shows a shuffle**. Before drawing that,
establish whether it happens. Search the aggregate for the reshuffle: if
`solution_deck` is refilled from `solution_discard` anywhere, render the
moment; **if it is not, that is a finding, not a rendering task** — the
deck can be exhausted and the rules do not say what then. Raise it through
CB-WP-0022's register as *underdetermined* and report it to `ground-game`.
Do not implement a reshuffle. Inventing rules is what CB-WP-0023 exists to
stop.
**Controls:**
- the drawn counts come from `solution_deck_len` and
`solution_discard.len()`, asserted against the view, never recomputed;
- a discard of zero cards renders as an empty pile, not as a missing one —
an absent slot reads as "no discard exists in this game";
- the shuffle question is answered in the task record with the line number
that settles it, either way.
## Task: what the other seats played, on the table
```task
id: CB-WP-0024-T03
status: todo
priority: high
```
Each seat's revealed selection should be legible as a played card in that
seat's area, not only as a sentence in `selection_words` and a line in the
log. The log stays — it is the record, and CB-WP-0018 built it for a
reason. This adds the picture, it does not replace the text.
**The hiding rule is not negotiable and is not this pass's to touch.**
`SelectionView::Hidden` is what a face-down selection projects, and
`a_seat_never_sees_another_seats_face_down_selection` asserts it on the
serialized view. A rendering that draws something suggestive of the hidden
choice — a suit tint, a differently-shaped back, anything varying with the
hidden value — is a leak even though the model did not leak. Draw
face-down as one identical back.
**Controls:**
- a test renders a mid-`Select` view and asserts the HTML contains nothing
that varies with another seat's hidden selection;
- the same view after Reveal shows it — otherwise the assertion above
passes for a renderer that draws nothing at all (this is the shape
`view.rs`'s own test already uses; copy it);
- `make sim` still passes: 26 scenarios encode what the table looks like.
## Task: the score that carries across games
```task
id: CB-WP-0024-T04
status: todo
priority: medium
```
`play` (`table.rs:219`) already owns the session — one listener, a seed
that advances, `run_game` in a loop — and throws away every summary but
the last. Accumulate them and show the running tally on the ending page,
beside the per-game outcome that is already drawn.
**Decide what "cumulative" means before summing anything, and write the
answer down.** `OutcomeView` carries `personal` (per seat, `i32`),
`mastery`, `group_success` and `winners`. Summing `personal` across games
and counting `winners` across games are different games' worth of
different answers, and the game is co-operative in a way that may make
neither the right one. If the rules do not say, **it is underdetermined**:
pick the one the code can defend, say in the UI which one is shown, and
raise the question rather than silently canonising a choice.
**Controls:**
- two games in one session produce a tally that is not either game's
score alone — the weakest possible assertion, and the one that catches a
tally reset by `play again`;
- the tally survives the seed advance and resets only on a new process;
- the per-game outcome block is unchanged, and a test says so.
## Task: evidence
```task
id: CB-WP-0024-T05
status: todo
priority: high
```
`evidence/CB-EV-0022-*.md`. Short — this is an S pass.
- **How many of the five remarks were already implemented**, and what that
says about the table's legibility versus its content. Three of five were
data already in the view being drawn as text. That is a finding about
rendering, and it is worth naming.
- **Whether the deck can be exhausted**, as T02 settled it.
- **What the cumulative-score question turned out to be**, and whether it
went to `ground-game`.
- **Quote CB-WP-0023's cost by re-running the instrument.**

272
workplans/CB-WP-0025-could-we-have-won.md Normal file
View file

@ -0,0 +1,272 @@
---
id: CB-WP-0025
kind: product
title: "Could we have won: a path out of a lost game, and how hard the game actually is"
status: ready
---
# Purpose
```
structural tier L (creates a new capability — a search over game state,
and a measurement the engine does not currently take;
both produce numbers ground-game will tune against)
chaos d8 = 6 → no override
declared tier L
```
Declaration 8 of chaos window 2. Tier L: separate survey, **adversarial
review**, ADR, then spec, then code.
## Two remarks, and why they are one pass
> *"I had a game where we lost and in this case I would have liked to know
> if and how we could have won… the best path is not computable I guess so
> a path to win is fine."*
> *"Do we have difficulty estimations? If so we should show them. It will
> help tuning the game. I felt it was too easy but then we lost, so who
> knows."*
They are the same machine asked two questions. *Was this game winnable?*
is a search from a recorded state. *How hard is this game?* is that search
run over many deals and counted. Building the second without the first
gives a win-rate with no witness; building the first without the second
gives one anecdote per game.
**"I felt it was too easy but then we lost, so who knows" is the finding.**
The maintainer cannot calibrate the game from play, and that is precisely
the gap `ground-game` is currently blocked in: **GROUND-WP-0005**
*(Difficulty tiers (threshold ratio) and optional Pressure dial)* is active
with **both its tasks in `wait`**. Tiers cannot be set without a measured
baseline, and clay-borg is the thing that can measure. This pass is what
unblocks it — which is the design-instrument aspect CB-WP-0022 is in the
middle of stating, arriving with a concrete demand.
## What already exists, so the survey does not re-find it
- **The state is replayable.** `cb-game-runtime` records sessions as
scenarios; `replay.rs` and `make replay-test` already re-run them.
A search does not need new persistence.
- **The move space is enumerable.** `legal_commands` exists and, since
CB-WP-0023, is narrow enough to be worth trusting — SOLVE is offered
only where it can act, so the branching factor is real rather than
inflated by inert moves.
- **Bots exist.** `games/ground/src/bot.rs` has `GreedyPolicy` and
`RandomPolicy`, wired through `bot_policy` (`table.rs:208`). A win rate
over N seeds is reachable with what is already there — the question is
whether that number *means* anything, which is the survey's problem.
- **The threshold is public.** `OutcomeView.total` / `.threshold` /
`.group_success`. Difficulty has a denominator already.
## What makes this hard, and must not be waved through
**The game is not perfect-information and the search must respect that.**
A path computed with the deck known is a path the players could never have
found. `view.rs` hides the deck, other seats' hands, and face-down
selections *by rule* (GR-S02/S04, GR-R02/R04). A retrospective solver
running on `GroundState` sees all of it. So the ADR must decide, in
words, **which of these three the tool answers**:
- *was this deal winnable by an omniscient player* — cheap, honest,
and answers a question nobody asked;
- *was it winnable from what the seats could see* — the question actually
asked, and the expensive one;
- *did a reasonable line exist* — a bounded search from the losing seat's
information, which may be the only affordable honest answer.
Getting this wrong produces a feature that tells the maintainer he could
have won by playing a card he had no way to know was there. **That is
worse than not shipping it.**
**And a difficulty number is a claim about a distribution.** One win rate
over one bot policy over N seeds is not "the difficulty"; it is that
policy's win rate. Whatever the spec adopts must name its policy, its N,
and its seed range, or `ground-game` will tune tiers against a number
whose meaning drifts the next time a bot improves.
## Task: survey
```task
id: CB-WP-0025-T01
status: todo
priority: high
```
`research/CB-RES-0008-*.md`, with `tier: L` and the chaos roll recorded
(`loop-lint` checks both).
Per §Step 1 the survey is done when it can name a **benchmark-to-beat**
per dimension — a number or a reproducible comparison, not an impression.
- **Retrospective solvers in games with hidden information.** The prior art
is real and should be named: determinized search (perfect-information
Monte Carlo) and its known failure — *strategy fusion*, where a
determinizing solver claims lines that require knowing which world it is
in. That failure is exactly the trap in §What makes this hard. Bridge
and Skat post-mortem tools are the closest analogues; poker solvers are
the well-studied case and the wrong shape.
- **"A path to win" as a product, not a proof.** The maintainer already
conceded optimality (*"the best path is not computable I guess"*). So
the target is a **witness**: one concrete line of play that reaches
`group_success`, or a defensible *no line found within bound B*. Name
what a witness must carry to be checkable.
- **Difficulty as a measured quantity in co-operative games.** Pandemic and
its relatives set difficulty by a dial with a published win rate. The
benchmark-to-beat is: can we produce a win rate whose confidence
interval is tight enough to distinguish two threshold settings?
- **Cost.** Search over an event-sourced aggregate with full `validate` on
every branch has a per-node price. Measure it on our machine, on our
scenarios — the runnable-baseline option applies here, since a search
that cannot finish while the player is still looking at the page is a
different feature.
## Task: adversarial review
```task
id: CB-WP-0025-T02
status: todo
priority: high
```
Tier L requires it. Exactly one round: challenge, then response, trail in
`history/`, unpolished. Require an attempt at:
- **that the honest version is unaffordable** — that a search respecting
the information rule is too expensive or too weak to find anything, so
the shipped tool will quietly become the omniscient one with a
reassuring label;
- **that a witness misleads more than it helps** — being shown a line that
needed a card you could not know about teaches a wrong lesson about the
game, and the tool would be better refusing to answer;
- **that the difficulty number is a bot benchmark wearing a difficulty
costume**, and `ground-game` will tune the game against our bot rather
than against play;
- **that this is CB-WP-0022's job** — the design instrument is being built
right now, and a difficulty measurement is a finding-producing tool. The
strongest counter is that the register records findings and this
*produces* them, but the reviewer should press whether that is a
distinction worth a separate capability.
## Task: decide
```task
id: CB-WP-0025-T03
status: todo
priority: high
```
`decisions/ADR-0013-*.md`. (ADR-0012 is CB-WP-0022 T03's.) At minimum:
- **which question the solver answers**, from the three in §What makes
this hard, and what it is called in the UI — the name must not overclaim;
- **the information boundary**: whether the search runs on `GroundState`
or on a `GroundView`, and if on state, what stops it using what the view
hides. Note that running on the view makes the rule structural rather
than a promise, and that this is the cheapest guarantee available;
- **the bound**: depth, node budget, or wall clock, and what *no path
found* means against it — a bounded search that says "unwinnable" is
lying, and the wording must say "none found within B";
- **whether difficulty ships as one number or a small table**, and what it
is a function of: policy, seat count, threshold, seed range;
- **where it lives** — a new crate, a mode of `cb-play`, or a tool under
`tools/`. The tier was declared L on the assumption of a new capability;
if the ADR concludes it is a mode of an existing one, say so, and the
over-declaration is a chaos-window data point worth recording.
## Task: specify
```task
id: CB-WP-0025-T04
status: todo
priority: high
```
`specs/` — extend `MetricsAndScenarios.md` or add a capability spec as the
ADR directs, with metrics, because a spec without them is prose.
Candidates, to be argued not adopted:
- **witness checkability** — every path the tool emits replays through the
existing scenario runner and ends in `group_success`. Target 100%, and it
is a hard gate, not a metric: a path that does not replay is a bug that
says the opposite of the truth;
- **search cost** — nodes and wall clock at the chosen bound, on the
recorded games we have;
- **difficulty resolution** — the smallest threshold difference the
measurement can distinguish, with its N. This is the number
`ground-game` needs, and stating it as *"we can tell 5 from 7 but not 7
from 8"* is more useful than a win rate with no error bar.
Per `ground-game`'s ruling (GROUND-WP-0004 T02), **any arithmetic finding
this produces ships a runnable reproduction and a row-level table** — never
a summed figure. A difficulty number is arithmetic, and it is exactly the
kind that has already gone wrong twice.
## Task: build the witness
```task
id: CB-WP-0025-T05
status: todo
priority: high
```
The search, the bound, and the replayable path. Wire it to the ending page
so a lost game can be asked the question — the page CB-WP-0024 T01 is
already reworking, so land that first or expect a conflict.
**A game that was won is not asked the question.** The feature exists for
a loss.
**Controls:**
- every emitted witness replays to `group_success` through the existing
runner — asserted, not spot-checked;
- a deal constructed to be unwinnable returns *none found*, and the test
says which construction makes it so;
- the information boundary is mutation-provable: relax it, and a test
naming *that* boundary goes red. If it cannot be mutated, it was a
comment rather than a rule.
## Task: measure the difficulty, and hand it to ground-game
```task
id: CB-WP-0025-T06
status: todo
priority: high
```
Run the measurement, ship it as a `make` target beside the other
instruments, and show the result in the game — the maintainer asked for it
to be visible, and a number in a file will not calibrate anything.
Then send it to `ground-game` **against GROUND-WP-0005**, which is active
with both tasks waiting on exactly this. Per CB-WP-0022 T06, it lands as a
file in their repo under their workplan, not only an inbox entry — *the
message that sat unread for four days is the baseline to beat*.
**Controls:**
- the number regenerates from a single command, and `facts.toml` carries
it if anything else quotes it (§Single source of fact — `make
facts-check`);
- the report carries the row-level table the ruling requires;
- **the seed range and policy are in the number's name**, not in a
footnote.
## Task: evidence
```task
id: CB-WP-0025-T07
status: todo
priority: high
```
`evidence/CB-EV-0023-*.md`.
- **Was the game winnable**, for the maintainer's actual lost game. That is
the acceptance test with a face on it.
- **What the honest search cost against the omniscient one**, since the
review will have pressed hardest there.
- **Whether the difficulty measurement moved ground-game**, or sat.
- **What tier L cost against what it caught** — third full-weight L pass in
the project, and the second in this chaos window.
- **Quote CB-WP-0024's cost by re-running the instrument.**