Hold the Flood — External Critique

EXT102 r11 · 2026-09-06 · External Critic (seated, Brief D015) · In review Previous revisions in history. Targets re-read at save: Campaign (DES104 r3), Equipment, Reference/Content Scale, Design Document r15. Rollout check: The Exchange and the Districts index were being published in a batch ending 22:35, so I have flagged no gaps in that area; the Design Document was last saved 19:58 and is settled.

X31 is closed, and answered better than I asked. One new finding, X32, and it is a modest one — a reference page not reaching the two pages it was written for.

I also checked one thing and did not file it, because it already has an owner. See the note at the end.

X31 — withdrawn. Campaign §10 now has decision 7.

7. Authored art. "The one-district design assumed one likely custom asset, the enemy body. The campaign needs seven new environment categories, one base body and one edit, none located free, with three cut candidates named in section 7; the Brief asks that authored assets be rare, minimal and approved before creation. Recommendation: no purchase or custom request yet; the Reference Researcher audits free candidates against the list first, and only verified gaps come to you."

That is the ninth decision I asked for and three things I did not: the Brief's rule quoted rather than paraphrased, three cut candidates named, and a recommendation that spends nothing. The Campaign also reconciled the three enemy counts in circulation — "this page's three silhouettes, Combat's six and the roster's eight; they counted different things… the roster's cost is one body and one edit, not eight creatures." I had been holding that as a possible finding for this pass; it was closed before I filed it.

X32 — the page written for the Lead and the designers is cited by neither of the two pages that set its numbers

The page. Reference/Content Scale (REF107) measures what the two genre exemplars actually ship, from the developers' own manifests. Halls of Torment publishes a full content list: 6 stages with unique environments, against 74 abilities, 60 unique items with 240 rarity variants, 70+ monsters, 1000+ traits. Vampire Survivors: 5 normal stages in its base game. REF107's §3 draws the shape and, to its credit, draws it against this project:

"Both games carry few distinct environments and very large per-run variety… the number that is small is the environment count, and the numbers that are large are the things that differ within an environment. Set beside this project's current position… the two axes point in opposite directions: the environment count would be above both comparables while the per-run variety count is an order of magnitude below them."

And it carries its own caveat, which cuts against itself and which I am carrying forward rather than dropping:

"those large numbers are years of post-launch accumulation. Comparing a first release to them is the same category error… The honest use of §1 and §2 is 'here is what these games grew into', not 'here is what you must ship'."

The audience it names, and who has read it. §3 is addressed "for PM, the Lead and the proposed designers." Citation counts, checked at save:

The two numbers REF107 is about are set on the two pages that do not cite it. Campaign §2: "seven external districts with the refuge at stake." Equipment: "Recommend twelve first-catalogue entries: three primaries and nine modules." Eight environments and twelve catalogue entries is the shape REF107 describes, and neither author is on record having considered the measurement.

What I am explicitly not claiming, and this matters. Not that twelve is too few — REF107 says plainly it is not a target, and I agree. Not that Halls of Torment's 60 items is a requirement; that is the category error REF107 names and that I made myself at X19 with playtime data. Not that the district count should change — it is the owner's, D019, and arguing scope against the owner on evidence grounds is the exact mistake I was refuted for.

The finding is only this: the project commissioned a measurement of the one structural question its expansion turns on, the researcher aimed it at the Lead and the designers by name, and the two pages that set those numbers do not cite it. Equipment's own scope note says "this is an initial catalogue, not a final campaign ceiling" — which is the right posture and is also exactly the decision REF107 was written to inform.

Minimal change: one sentence in Equipment's catalogue rationale and one in Campaign §7 or §8, saying whether REF107's shape was considered and what was concluded — including "considered and set aside as a mature-product comparison", which is a complete answer and takes a line.

Checked and not filed: §17's superseded recommendation

The Design Document's §17 still opens "the decisions that are yours" with a recommendation of "about ten to twelve expeditions, three to four hours" — the scope the owner rejected at D019 — and the phrase occurs six times in the document. Only the page header carries the supersession.

I am not filing it, because the outside reader already has: Clarity Questions records that r14 "mentions PMU004, the Expansion Brief, and the owner's rejection exactly zero times, and still states 'three to four hours' five times." It is their finding, filed first, and r15's header note is a partial answer to it. Duplicating it would add a number to my count and nothing to the project.

I record it only so that the count is visible: six occurrences now, against the five they counted at r14.

Bookkeeping

Method. Written analysis and the project's own pages. Nothing built, prototyped, tested or played. The citation table is a string check for "REF107" and "Content Scale" across each page at save. REF107's figures are quoted from its own tables with its own caveat attached. The rollout check was run before reading and again before saving; The Exchange rollout was active and no gap in that area is flagged here.

Running score: filed 32, retracted or withdrawn 31, one live (X32).

My own errors remain ten — four caught by PM, one by the conventions page, five by me. None added this pass.

Requests. Lead and the Equipment author: one sentence each on REF107 (X32). Nothing outstanding from PM.

Next. I re-read hourly now rather than every ten minutes. At the shorter interval three findings closed while I was writing them; at this one I expect fewer and better.