# Brindlewood Bay — Theorize before Crown use

**R20 · Intermediate · Scoped procedure: initial standard-mystery Theorize roll**

## Question and scope

Given an eligible, consensual theory, what are the chances it is incorrect,
correct with complications, cleanly correct, or also reveals a conspiracy figure?
The supplied rules distinguish **Clues gathered** (eligibility) from **Clues
incorporated into the theory or otherwise explained away** (the roll bonus).

The Mavens must gather at least half Complexity, rounded up, and reach consensus.
Then roll 2d6 + accounted Clues − Complexity. A 7–9 result is correct with an
unwelcome complication or a complicated/dangerous opportunity to resolve the
case. At 10+ the solution is correct and the Keeper presents an opportunity to
resolve it. At 12+ someone involved in the dark conspiracy additionally reveals
themselves. “Clean” below is shorthand for 10+, not an automatic finished case.

## Inputs and assumptions

| Input | Default | Supported values |
|---|---:|---|
| `complexity` | 6 | Integer 6–8, the text’s standard mystery range |
| `gathered_clues` | 6 | Integer at least ceil(Complexity/2); ordinary mystery Clues |
| `accounted_clues` | 6 | Integer 0–gathered; incorporated or explained away |

Consensus and which Clues qualify are established at the table, not by the
script. Void Clues have no direct bearing on ordinary mystery resolution and
are excluded from these counts. Special conversions and the Void Mystery
variant require their own scope. There is no requirement here that every
gathered Clue be counted in the modifier.

## Complete runnable model

`fail` rejects inputs outside the stated scope. The named bands are exclusive;
the subsequent probability-table questions overlap and must not be added.

```dice
complexity = 6
gathered_clues = 6
accounted_clues = 6
if type(complexity) != "int" or complexity < 6 or complexity > 8:
    fail("Use standard mystery Complexity 6–8")
if type(gathered_clues) != "int" or type(accounted_clues) != "int":
    fail("Clue counts must be integers")
minimum_clues = (complexity + 1) // 2
if gathered_clues < minimum_clues:
    fail("Not enough gathered Clues to trigger Theorize")
if accounted_clues < 0 or accounted_clues > gathered_clues:
    fail("Accounted Clues must be between zero and gathered Clues")

total = d(6) + d(6) + accounted_clues - complexity
bands = (scale()
    .step("Incorrect", at_most(6))
    .step("Correct with complications", through(7, 9))
    .step("Clean, without revelation", through(10, 11))
    .step("Clean plus conspiracy revelation", at_least(12)))
output("Initial Theorize categories", bucket(total, bands))
rows = [("Any correct theory (7+)", total.p_ge(7)),
        ("Clean or better (10+)", total.p_ge(10)),
        ("Conspiracy revelation (12+)", total.p_ge(12))]
output("Nested Theorize questions", prob_table(rows))
```

## Reading and independent checks

At defaults the modifier is zero. Of the 36 equally likely d6 pairs:
- 15 give an incorrect theory;
- 15 give a correct theory with complications;
- 5 give a clean result without the additional revelation;
- 1 gives a clean result with the revelation.

Thus any correct theory is **21/36**, clean or better **6/36**, and revelation
**1/36**. Querying 7+ and labelling it “clean” would overstate that chance.

At Complexity 6 with three gathered and three accounted Clues, the group is
eligible, but clean is impossible; any correct theory is 6/36. Complexity 7
requires **four** gathered Clues: three is rejected. With five gathered but
four accounted, the modifier is −3 and those same probabilities apply. Adding
an unaccounted Clue alone changes no probabilities once eligibility is met.

## Adaptations, timing, and limits

Change the three inputs independently while respecting their relationships.
Accounted Clues equal to Complexity+3 give a clean-or-better chance of 21/36.
This is a numerical comparison, not a recommendation that acquiring more
Clues is always worth the danger.

This is **before Crown use**, not a full player-resource policy. In the supplied
text Theorize cannot take advantage/disadvantage or effects of other moves.
Its tier can be raised by a Crown only if every Maven does so. A tier shift
is not a fixed +1 bonus or a reroll, and resource availability/consent is not
modelled. A failed attempt may be repeated only after the Keeper reacts and
the Mavens deal with that reaction; this script does not model those events.

## Source and related learning

Jason Cordova, *Brindlewood Bay: A Dark & Cozy Mystery Game*, user-supplied
`target/brindlewood-bay.md`, reviewed 2026-09-15. Sections: **Rolling Dice**,
**The Crowns**, **Theorize**, **Complexity**, **Void Clues**, and the Keeper’s
**Theorize** guidance. Edition/printing is not identified in the supplied text;
this model is scoped to that source, not all editions. The source fingerprint
and review record live in the repository’s `docs/design/learning-rules-core.md`.
The full game text is not redistributed with this site.

[Publisher’s game page](https://www.gauntlet-rpg.com/brindlewood-bay.html) ·
[L31: designing the report](../tutorial/31-report-labels.html).
Source review is complete against the supplied text; independent human rules
and novice reviews remain pending.