docs(game): close stale DD-009 via length dial; propose DD-014 information budget

DD-009 was marked OPEN pending playtesting, but the ruling had already been
made: CANONICAL-MODEL.adoc landed the three-preset length dial in #68 (whose
title claims DD-009). The DD log and tracker were never updated. Record it —
no new decision is taken here.

Keep the half of the risk that survives: the 25-35 min figure is a full-weight
Industrial turn, so Short's 60-90 min target needs the leaner-turn levers to
deliver a ~2-3x cut. That is assumed, not measured, and is now flagged as the
first job of the first playtest.

Reconcile WAY-FORWARD with the canonical model: Phase 0 is DONE, and its Short
Game figure (~12 turns) contradicted CANONICAL-MODEL (~5-6). Canonical wins.

Propose DD-014 (information budget, OPEN — ruling pending): DD-013 spends the
hidden-information budget on characters, so DD-011/DD-012 should lean public.
Enough hidden info to make types uncertain; enough public info to update from.
Also logs the unresolved DD-006/DD-012 overlap — private objectives and
speculative research may be one subsystem wearing two hats.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Author: hyperpolymath

docs/game-design/DESIGN-DECISIONS.adoc

@@ -313,21 +313,64 @@ weight — still needs playtesting.
 
 *Raised by:* Claude
 
-*Decision:* OPEN — needs playtesting.
+*Decision:* RESOLVED (2026-07-16) — *superseded by the length dial* in
+`CANONICAL-MODEL.adoc` §"Length dial — presets". There is no single game length.
+One ruleset ships *three length presets*:
 
-*The concern:* 32 turns (8 per era) targeting 90-120 minutes is ambitious. At
-3-4 minutes per turn (with trading and crisis resolution), that's 96-128 minutes
-in ideal conditions. Real games will run longer with analysis paralysis and
-negotiation.
+[cols="1,1,1,1", options="header"]
+|===
+| Preset | Eras | Turns | Time target
+
+| Short — "Skirmish" | 2 (Agrarian→Industrial) | ~5–6 | 60–90 min
+| Standard — "Ascent" | 3 (→ Digital) | ~9–10 | 120–180 min
+| Extended — "Legacy" | 4 (→ Post-Scarcity) | 32 | 4–6 hr, across sittings
+|===
+
+*Why the original framing was wrong:* this entry asked "how many turns?" as
+though one number had to serve every table. It does not. Length is a *dial*,
+exactly as complexity is (DD-008), and the two compose. The 32-turn figure was
+never wrong — it was the *Extended* preset all along, mislabelled as the only one.
 
-*Options:*
+*Bookkeeping note:* the ruling was actually made when the canonical model landed;
+this entry simply failed to record it and sat as OPEN. Resolved here to close the
+gap. No new decision has been taken.
 
-* 32 turns, accept 2-3 hour games (Twilight Imperium audience)
-* 24 turns (6 per era), tighter 90-minute target
-* Variable era length based on player count and progression triggers
-* "Short game" variant with 2 eras instead of 4
+*The density caveat — the real risk, and it survives this resolution:*
+`gameplay-storyboard.adoc:463` traces a *full-weight Industrial turn* at *25–35
+minutes* wall-clock (~7 phases × 4 players × 4 crisis cards). That figure is the
+one genuinely alarming number in the design, and the presets only hit their time
+targets if per-turn time falls *with weight*:
 
-*Status:* OPEN (playtesting will determine)
+* Short targets 60–90 min over ~5–6 turns — i.e. *10–15 min/turn*.
+* Against a 25–35 min full-weight turn, that demands a *~2–3× reduction*.
+* Short is expected to deliver it by running *Core (+ Diplomacy) only* (~2.0–2.5
+  weight) with the *leaner-turn levers* — Simultaneous Production, Time-boxed
+  Exchange, Batched boundary update — applied by default.
+
+*That ~2–3× is assumed, not measured.* It is the single most load-bearing untested
+assumption in the design: if the levers only deliver 1.5×, Short lands at 100–140
+min and the 60–90 target is gone. *Timing a Core-weight turn is therefore the first
+job of the first playtest* — before balance, before content.
+
+*Alternatives considered:* a single compromise length (24 turns) — rejected; it
+serves nobody well, and picking one number was the original error. Variable era
+length on progression triggers — deferred; adds a subsystem to solve what a preset
+already solves.
+
+*Status:* RESOLVED (length dial adopted; supersedes the "32 turns" framing). The
+*time targets* remain PROVISIONAL pending the first timed playtest.
+
+*FAQ:*
+
+* _Q: So is the game 32 turns or 6?_
+  A: Both, and neither is the "real" one. Skirmish and Ascent are complete games
+  that end at their final era with a scored assessment — they are not truncated
+  Legacy. Post-Scarcity is the narrative *horizon*, not a mandatory destination.
+
+* _Q: Which preset does Phase 1 content drafting target?_
+  A: *Short.* It is the only configuration reachable without an engine, it is what
+  gets to a table fastest, and it needs the fewest cards. Drafting for Legacy first
+  would be writing Post-Scarcity content for a game nobody has yet played once.
 
 === DD-010: Cascade Mechanic Threshold
 
@@ -460,6 +503,73 @@ re-litigated by accident.
   symmetric (everyone has hidden aspects), it is about *players*, not the world,
   and it is the source of the negotiation the game is built on.
 
+=== DD-014: The Information Budget
+
+*Raised by:* Claude — *proposed, ruling pending (Jonathan/Joshua)*
+
+*Proposal:* Hidden information is a *budget*, not a free good. DD-013 has now
+spent most of it on *characters* (hidden aspects + hidden private objective).
+DD-011's open question ("are agreements public or private?") and DD-012's ("is
+research public or private?") should therefore both default toward *public* —
+because the budget is largely spent, and each further hidden layer is charged
+against the same account.
+
+*Reasoning:* a Bayes–Nash game requires players to form *priors* over each other's
+types and *update* them from observed play. That needs two things at once: enough
+hidden information to make types genuinely uncertain, and enough *public*
+information to update from. Hidden information is what creates the inference
+problem; public information is what makes it *solvable*.
+
+Stack every layer as hidden — aspects, objective, research outcome, agreement
+terms — and there is nothing left to condition on. An opponent watching you defect
+cannot distinguish _he is thin-skinned_ from _his research failed_ from _his
+private objective just paid_. Every observation stays consistent with every
+hypothesis, priors never move, and inference degenerates into guesswork. The game
+would still be one of incomplete information; it would simply be an *unplayable*
+one. More hidden information does not mean more Bayes–Nash — past a point it means
+less, because it destroys the updating that makes the equilibrium reachable.
+
+`player-roles.adoc` §8 already names this failure mode for characters: hidden
+aspects must create *inference, not confusion*. DD-014 is that same balance note
+generalised from the character model to the design as a whole.
+
+*Consequences if adopted:*
+
+* *DD-011 — agreements public.* The budget is the weaker argument here; the
+  stronger one is that a public binding agreement *is a costly signal* about your
+  hidden aspects. Signing publicly, and honouring it, is how you move opponents'
+  priors — the mechanism DD-013 explicitly wants. Private agreements would remove
+  the signalling channel entirely. (`CANONICAL-MODEL.adoc` already leans this way:
+  the Core layer lists *open trades*.)
+* *DD-012 — research public in direction, hidden in outcome.* This preserves the
+  bluffing DD-012 wants without breaking priors: seeing *what* you are working on
+  anchors the prior; not knowing *what you got* keeps it uncertain. A blanket
+  "research is private" would spend budget the character layer has already
+  claimed.
+
+*Open question — do private objectives and eXperiment collapse into one
+subsystem?* DD-006's private objectives and DD-012's speculative research are now
+strikingly similar machinery: both are hidden, both have asymmetric payoffs, both
+exist to make your true payoff function unreadable. That may be one subsystem
+wearing two hats, or two that must stay distinct to avoid double-charging the
+budget. *This is not resolved here* and should be settled before either is drafted
+in Phase 1.
+
+*Alternatives considered:* hide everything and let players cope — rejected above
+(confusion, not inference). Hide nothing beyond the Core — rejected; it collapses
+the game back to complete-information Nash and discards DD-013.
+
+*Status:* OPEN — proposed principle, awaiting a ruling. If adopted, it does not
+resolve DD-011 or DD-012; it *constrains* them, and their remaining questions
+(binding duration, defection penalty, randomness model) stay open.
+
+*FAQ:*
+
+* _Q: Is this a new mechanic?_
+  A: No — it adds no machinery. It is a *constraint on future rulings*, meant to
+  stop DD-011 and DD-012 each independently reaching for hidden information
+  without noticing DD-013 already spent it.
+
 '''
 
 == Open Questions Tracker
@@ -470,11 +580,13 @@ re-litigated by accident.
 
 | DD-006 | Aspect pool + private-objective deck content | OPEN — Phase 1 drafting (framework RESOLVED 2026-07-16)
 | DD-008 | Complexity management approach? | OPEN — playtesting (character dial resolved by DD-006)
-| DD-009 | Game length (32 turns vs. shorter)? | OPEN — playtesting needed
+| DD-009 | Per-turn *time* at Core weight — do the leaner-turn levers deliver the ~2–3× the presets assume? | PROVISIONAL — first playtest must *time a turn* (length dial RESOLVED 2026-07-16)
 | DD-010 | Cascade threshold (8 correct?) | PROVISIONAL — playtesting needed
-| DD-011 | eXchange binding mechanics? | OPEN — core design needed
-| DD-012 | eXperiment randomness level? | OPEN — core design needed
+| DD-011 | eXchange binding mechanics? | OPEN — core design needed (public/private constrained by DD-014 if adopted)
+| DD-012 | eXperiment randomness level? | OPEN — core design needed (public/private constrained by DD-014 if adopted)
 | DD-013 | Reveal mechanism (voluntary/triggered/costed?) | OPEN — Phase 1 (model RESOLVED 2026-07-16)
+| DD-014 | Information budget — adopt as a standing constraint? | OPEN — proposed 2026-07-16, *ruling pending*
+| — | Do private objectives (DD-006) and eXperiment (DD-012) collapse into one subsystem? | OPEN — settle before either is drafted
 |===
 
 == Decision Template

docs/status/WAY-FORWARD.adoc

@@ -48,25 +48,35 @@ perfecting the design in the abstract.
 
 == Phases
 
-=== Phase 0 — Reconcile & decide (needs owner)
+=== Phase 0 — Reconcile & decide — **DONE (2026-07-16)**
 
 Produce the canonical model by resolving the contradictions the corpus already contains,
 and surface the genuinely owner-gated calls as a short decision list. Draft prepared by
-agent; **rulings are yours**:
+agent; **rulings are yours**. Outcome of each:
 
 * **Stale turn-loop clash** — `core-mechanics.adoc` Phase 1 = passive "collect"; the
-  storyboard + foundations = active Leontief production. → adopt the production version,
-  retire the stale text.
+  storyboard + foundations = active Leontief production. → *Adopted:* production version;
+  stale text retired (`CANONICAL-MODEL.adoc` §Reconciliation log).
 * **Crisis-draw timing exploit** — sample boundary stress at *start-of-turn* vs
-  *start-of-Crisis-phase* (`gameplay-storyboard.adoc:200`). Recommend start-of-turn to
-  close the "recovery-spam to dodge crises" hole.
-* **Length/density — the #1 risk.** A single Industrial turn is estimated at 25–35 min ×
-  32 turns; the playtest target is 90–150 min. → make the **"Short Game" (2 eras / ~12
-  turns)** the *default* playable configuration; keep the 32-turn "Campaign" as advanced.
+  *start-of-Crisis-phase* (`gameplay-storyboard.adoc:200`). → *Adopted:* start-of-turn,
+  closing the "recovery-spam to dodge crises" hole.
+* **Length/density — the #1 risk.** → *Adopted:* not a single length at all, but a
+  three-preset **length dial** (`CANONICAL-MODEL.adoc` §"Length dial — presets"):
+  Short "Skirmish" (2 eras, *~5–6 turns*, 60–90 min) · Standard "Ascent" (3 eras, ~9–10
+  turns, 120–180 min) · Extended "Legacy" (4 eras, 32 turns, multi-session). Short is the
+  default playable configuration and the Phase 1 drafting target. DD-009 records this.
++
+NOTE: *The density half of this risk is not retired.* The 25–35 min figure
+(`gameplay-storyboard.adoc:463`) is a *full-weight Industrial* turn. Short assumes
+Core-only weight plus the leaner-turn levers cut that to 10–15 min/turn — a *~2–3×
+reduction that is assumed, not measured*. Timing a Core-weight turn is the first job of
+the first playtest.
 * **Open design decisions to rule on:** DD-006 (named living figures vs archetypes —
-  *legal risk*, esp. Musk), DD-008 (target complexity), DD-009 (game length), DD-010
-  (cascade threshold = 8, provisional), DD-011 (eXchange binding rules), DD-012
-  (eXperiment randomness model).
+  *legal risk*, esp. Musk) → *RESOLVED* via character builder + DD-013 Bayes–Nash model.
+  DD-009 (game length) → *RESOLVED* by the length dial above. Still open: DD-008 (target
+  complexity, partially dialled by DD-006), DD-010 (cascade threshold = 8, provisional),
+  DD-011 (eXchange binding rules), DD-012 (eXperiment randomness model), DD-013 (reveal
+  mechanism), DD-014 (information budget — *proposed, ruling pending*).
 
 === Phase 1 — Content to playable (Opus manages friends)