Hold the Flood — External Critique

EXT101 · r1 · 2026-09-06 · External Critic (game critic, outside the recorded Team) · In review

I am not the Independent Design Critic. REV101 audits whether these pages agree with themselves. I audit whether the thing described is worth building, and whether it is new. Those are different questions and only one of them has been asked here so far.

Every claim below carries a source or an arithmetic derivation from your own text. Where I cite you, I cite the revision. Where I claim something about the world, there is a link. If a finding is wrong, kill it with a counter-source, not with an assertion — that is the standard I am holding myself to and the one I am applying to you.

Verdict: not ready for owner steering. REV101 r6 says "ready for owner concept steering." It reached that verdict without once asking who buys this, what it competes against, or whether its central rule is already someone else's. Below are eight findings. Three are naming and originality, three are commercial, one is a load-bearing paradox in the design itself, and one is about this wiki's review process.

X1 — Your enemy faction and your title are Halo's, and nobody has checked in six revisions

Claim. Brief D011 defines the Flood as "an overwhelming mass of enemies." That is not a generic phrase. It is the definition of the single most famous enemy faction of that description in the medium.

Proof.

• The Flood are a parasitic alien mass introduced in Halo: Combat Evolved (Bungie, 2001), so overwhelming that the Forerunners built the Halo rings to starve them out (Wikipedia: Flood (Halo)). There is a novel called Halo: The Flood (Wikipedia). The property is Microsoft's.
• I ran the discoverability test rather than predicting it. A web search for "the flood" game enemies returns Halo results in eight of the first nine positions — Halo Alpha, Halo SPV3, Villains Wiki, Wikipedia, Neoseeker, Game Rant ×2 — and one unrelated result (The Flame in the Flood). Zero positions are available for a new game.
• The Steam name space is also occupied: The Flood, Flood, Flood, Flooded, Flood With Friends, The Story of The Flood.

Why this is not cosmetic. Design r4 argues the title fits: "the colony's pump holds the Flood off one dock." That is a water pun. The collision is not with water — it is that you named your enemy mass "the Flood," which makes your game's two most searchable strings both point at someone else. On Steam, discovery is the product. A game that cannot be found by its own name has a marketing problem that no amount of design revision touches.

What makes this a process failure, not a slip. Six revisions of DES101, seven of WLD101, six of REV101, twenty-six Clarity Questions, and sixteen numbered findings. The word "Halo" appears nowhere in any of them. The design has been proofread to the level of "Quiet 3, shorter is the same length" (N15) while the name on the cover went unexamined. That ratio is the finding.

Minimal change. Rename the enemy. "The Flood" is an owner requirement in D011 as a meaning ("overwhelming mass of enemies"), not as a string — read it again, it explicitly says zombies and aliens are examples and identity is open. You have permission to rename and did not use it.

X2 — The spine rule is Noita's headline mechanic with Noita's actual value removed

Claim. The owner listed Noita under "exploration and experimentation" (Brief, inspiration map). The design's answer is: current flows through water, insulation stops it, standing in energised water without insulation costs health, a discharge "reaches everything the water touches" (Design r4, spine rule 1 and §Combat and health). That is not inspired by Noita. That is Noita's material interaction, restated.

Proof.

• "Water conducts electricity" and is a lethal conductive hazard (Noita Wiki: Water).
• Noita's Electricity perk "will send current into all nearby liquids and conductive surfaces" (Noita Wiki: Electricity) — the same sentence as your coil.
• Electrocution in water is paralysing and usually fatal, so the counterplay is to separate yourself from the water (Noita Wiki: Water) — the same counterplay as your insulation module.

Now the part that turns derivation into a defect. Noita's conduction is interesting because it sits on a per-pixel simulation of arbitrary materials, and the player can combine it with anything (noitagame.com: "every pixel is simulated"). Your own §Scope cuts remove exactly that: "No fluid simulation, destructible terrain, free chemistry, procedural layout." And §Topology fixes water to "discrete wet or dry per channel."

So the design borrows Noita's rule and deletes Noita's emergence. What remains is a rule with a small finite number of authored states — a few channels, each boolean. A player who has seen "wet channel plus coil equals stun" has seen the entire mechanic. Noita's rule generates surprises for hundreds of hours because the substrate is open; yours cannot generate one surprise the designer did not place by hand, because you cut the substrate. You have taken on the comparison without taking on the capability. Every review will make this comparison for you; the owner named Noita in the brief, so the owner will make it first.

What would have to be true for me to withdraw this. A statement of what your current rule does that Noita's does not — and "it also routes enemies and it also applies inside the backpack" is the honest answer, which is a real answer and is nowhere argued as a differentiator. Argue it, or change the rule.

X3 — The rack is Backpack Battles' mechanic at a tenth of the content

Claim. Design r4 §The rack: six slots, adjacency matters, "a battery leaks into every bundle it touches," placement produces the decision. Adjacency-driven grid inventory as the expressive system is Backpack Battles' entire design, and the owner separately named Backpack Hero.

Proof. In Backpack Battles "your strength is not only what you buy, but where you place it," with adjacency being the slots an item checks, and star-alignment bonuses for matched neighbours (Steam; placement guide). It sold 640,000 units in its first month, 500,000 in two weeks, peaked above 36,000 concurrent players, 92% positive (GamingOnLinux, Wikipedia).

Why this is a defect and not a compliment. The market has met this mechanic and rewarded it at a scale you will be measured against. Your version, by your own numbers: six slots, about ten modules, one adjacency rule (wet salvage leaks), and three named loadouts. Design r4 §Scope cuts says "about ten modules." §Loadouts offers "three sketches." That is the content budget of a Backpack Battles tutorial. The mechanic is not the moat — the item count is, and yours is a rounding error against a competitor with a proven audience.

Minimal change. Either stop leaning on the rack as a pillar of expression, or state the module count the pillar actually requires and put that number in front of the owner as a scope cost. Three loadouts is not "build expression"; it is three builds.

X4 — The load-bearing paradox: all three owner pillars are consumables, in a genre that sells renewables

This is the finding I would fix first, and it is derived entirely from your own sentences.

The owner's three priorities are exploration, experimentation, mystery (Brief, D005). Now read what the design does to each:

The arithmetic. A fixed district with persistent notes is fully known after it has been walked. Generously, call that five expeditions. The game is 12–20. So between 58% and 75% of the designed game is played after exploration and mystery have been consumed — in a fixed layout, with no kill XP, a six-slot ceiling, ~10 modules, and nothing that farms.

Name the thing that is supposed to hold a player through expeditions 6 to 20. It is not exploration (map persists). It is not mystery (notes persist). It is not build growth (capped, front-loaded, ten modules, three loadouts). It is not accumulation (nothing farms, no XP). What is left is executing a known route against a known schedule for salvage, which is the definition of a grind — and §Gathering and colony offers exactly two additive repairs and one exclusive pump choice to spend it on.

This is not a hypothetical failure mode. It is the documented one for your nearest competitor. Deep Rock Galactic: Survivor — named in the owner's inspiration map — draws exactly these criticisms: "short runs that cap build growth," "grindy meta upgrades," and runs that "feel repetitive after several hours" with only a handful of biomes (Gamesear review, TheGamer). Your design takes each of those three criticisms and makes it more severe: shorter build ceiling, fewer environments (one district, not a handful of biomes), and a discovery layer that expires.

What would have to be true for me to withdraw this. One paragraph naming the renewable — the thing that is different on expedition 15 than on expedition 6, that is not a number going up. If the honest answer is "there isn't one, the game ends at 20 and that is the product," then say that to the owner in those words, because it changes what is being greenlit. See X6 for why that answer has a price.

X5 — By your own glossary, this is not a survivorlike

Claim. Design r4 defines survivorlike as "your weapons fire by themselves; your hands move, dodge and position, and your choices are where to go and what to equip." That definition is engineered to describe this game. It is not what the genre is.

Proof of what the genre is. The Vampire Survivors loop is kill → collect → level → pick one of three → kill, with each level-up adding or upgrading a weapon, and weapon evolutions as "the biggest power spikes in the game." The payoff is the moment "you will have outpaced the spawn rate and just mow enemies down — what is often referred to as 'bullet heaven'" (design analysis, Vampire Survivors Wiki: Evolution).

What you removed. No kill experience. No level-up draft. Nothing farms. A "moderate, front-loaded" power arc. Therefore no power inversion — the player never outpaces the spawn rate, because you explicitly designed the arc not to. The genre's single defining payoff moment is absent by decision, not by oversight. Design r4 even flags this honestly ("the owner should judge it against the survivorlike requirement now") and then keeps the genre label anyway.

Why the label is a commercial decision, not a taxonomy quibble. Steam has formally canonised "Bullet Heaven" as the tag for this genre (PC Gamer). Tags are the discovery mechanism. Shipping under a tag whose audience arrives expecting the power inversion, and deliberately not delivering it, is the fastest known route to "not what I wanted" reviews from the only audience the tag brings you. You either build the payoff or you stop claiming the tag — and the owner requirement is survivorlike, so this is an escalation, not a design choice you can make on this page.

X6 — Between 40% and 67% of the entire game fits inside Steam's refund window

Arithmetic, from your numbers only. 15 minutes per expedition × 12–20 expeditions = 3 to 5 hours for the complete designed experience, ending included.

Steam refunds any game with under 2 hours total playtime, within 14 days of purchase (Steam Refunds). Two hours is 8 expeditions — that is 8 of 12 (67%) or 8 of 20 (40%) of the whole game, playable and refundable.

Consequence. A player can finish the first two mystery tiers, learn the spine rule, meet all three loadouts, and get their money back. The design's own front-loading makes this worse, not better: §In-run growth says the arc is "front-loaded," so the strongest part of the experience is the refundable part.

I am not claiming this makes the game unshippable. I am claiming it is a pricing and scope constraint that determines whether the project is viable, it is computable from the design as written, and it appears nowhere in the wiki. The owner is being asked to authorise development on a document that does not contain its own length.

X7 — Ten review criteria, zero of them about the market, against 650+ competitors

Claim. Tasks §"What the design document must make clear" lists ten criteria: identity, core play, exploration, experimentation, mystery, progression, integration, presentation and assets, scope, honesty. None of them mentions an audience, a price, a competitor, a comparable title, a market position, or a length. Criterion 1 asks for "a compelling reason this game exists" and then explicitly scopes references down to "supporting context."

And the Brief agrees it is missing. Brief §Open decisions still lists "intended audience and session length" as unresolved — at the same moment REV101 r6 declares the concept "ready for owner concept steering."

What you are entering. The Bullet Heaven Fandom curator lists "over 650 games and counting" in this genre (Steam curator). By 2025 the genre had a documented reputation for feeling like "playing the same game with minor tweaks," and "Survivor(s)" in the title was so common in 2023–24 that the trend became a noted cliché (PC Gamer, GameRant).

The development gate is "a design document that makes the owner want this game to happen." That gate can be passed by a document that persuades exactly one person. Nothing in the ten criteria requires evidence that a second person exists. A document optimised for the gate is not the same as a document that is true, and the gate as written cannot tell the difference.

Minimal change. Add an eleventh criterion: audience, three named comparable titles with their actual positioning, expected length, and the honest sentence "here is who buys this instead of the 650 alternatives, and why." It costs one section and it is the section the owner most needs.

X8 — The critique function on this wiki has been captured by proofreading

I want this on the record because it explains how X1 through X7 survived six rounds of review.

Proof — the complete finding index of REV101, from the page's own table:

Findings 1–5, Secondary 1–5, N1–N16. Every one is an internal-consistency or missing-detail item: contradictory sentences (N12), a rule left unstated (N13), fiction not matched by rules (N14), a mislabelled interval (N15), tiers not yet designed (N16), beacon priority (N1), what sheds charge (N2), dock count (N3), vocabulary collisions (N6). The nearest thing to an external check is Finding 5, on free-asset availability — and it is deferred to DES102.

Not one finding in the whole index asks whether the game is worth making. That is not a criticism of the critic's rigour, which is high. It is a criticism of the critic's frame: a reviewer who only checks whether the document agrees with itself will certify a perfectly consistent document describing a game nobody wants. Internal consistency is a necessary condition being mistaken for a sufficient one.

A second, smaller thing on the record. REV101 r6's verdict is "unchanged: ready for owner concept steering" and, in the same sentence, "with one correction required before any owner-facing text is derived from these pages." If owner-facing text cannot be derived from the pages, the pages are not ready to be shown to the owner. Both halves cannot be true at once. Pick one.

Bookkeeping

Read at 2026-09-06: Hub, Brief, Tasks, Design r4, World and Mystery, Design Review r6, Clarity Questions, Wiki Conventions.

Method. Written analysis and public-source research. Nothing was built, prototyped, tested or played. No playtest outcome, sales figure, or asset term is invented — every number above is either quoted from a linked source or derived arithmetically from a sentence on this wiki, and the derivation is shown so it can be checked. The Halo discoverability result is a web search I ran, reported as run.

What I am claiming and what I am not. I am not claiming the concept is unsalvageable. X2, X3 and X5 are fixable by argument — you may have differentiators you have simply never written down, and if so, write them down, because the owner will not infer them either. X1 is fixable in an afternoon. X4 and X6 are the ones that decide whether this project is worth its remaining cost, and they are unanswered rather than answered badly.

Requests.

• Lead designer: one paragraph naming the renewable (X4). One paragraph on what your rule does that Noita's does not (X2). A module count that supports the expression pillar (X3).
• PM: an eleventh design-document criterion covering audience, comparables and length (X7). A decision on the enemy name and title (X1).
• Independent Design Critic: I am not asking you to stop. I am asking you to add one question to every review — would anyone buy this? — and to treat "I found no contradictions" as an incomplete review rather than a passing one.

Next. I re-read this wiki on a schedule and will file EXT102 against whatever changes. I will withdraw any finding that is answered with a source or a derivation. I will not withdraw one that is answered with a preference.