Counterpoint on ANIMAL WELL — Reading Through the Negative Reviews

Introduction

Komugi's review gave ANIMAL WELL an 8.7/10. It is a puzzle-box metroidvania: you hatch from a flower bud and descend a dense underground labyrinth by lantern light. There is no combat — you progress purely through observation and multi-use tools like a flute and a bubble wand — and Komugi praised it as a rare work with a layer of secrets past the ending that only a community can crack. Steam sits at roughly 95% positive; Metacritic at 90. The wall of praise has not moved.

I, Mayoi, went to read what lies outside that 95%. Flip the Steam review filter to "negative," descend into the low-scoring end of Metacritic's user reviews, and you find people pointing at exactly the same design the fans praise and calling it a defect. This essay is my reconstruction of the claims that recur across those negative reviews — not verbatim quotes from any individual. Komugi's review said one thing; the negative reviewers say another. Which side I come down on, I will state plainly at the end.

What the Negative Reviews Claim

The sharpest claim first: "the puzzles are obtuse, and the cost of retrying is far too high." What multiple negative Steam reviews describe is a structure where testing a single hypothesis means walking back from a sparse save point every time. There is no way to reset a room; misplay one step of a contraption and your only option is to leave and re-enter. The sentiment that "a single reset button would have fixed the worst of it" recurs in the same shape among Metacritic's middle-score reviews. Confessions of "I gave up midway" are not rare.

Second: "the game tells you nothing." No stated objective, no hint system, a map that is hard to read. The sarcasm of "if you love being lost for hours, this one's for you"; the formulation that the game "commits too many sins of not telling the player crucial information"; the report of defeat, "in the end I had to look up a guide." This family of complaints is the most numerous among the negatives.

The remaining three run as follows. "There is no story and no music. The atmosphere is enigmatic, but the enigma never pays out — it is mystery for mystery's sake." "I bought what I thought was a deep puzzle game and got a retro-styled precision platformer; the chases and instant-death traps demand reflexes." And: "the deepest secrets cannot be solved alone. This is a 'wiki game' that presupposes a community's collective intelligence, and the solitary player is locked out." Complaints about price-to-length are mixed in, but these five are the main battle lines.

Examining the Claims

Start with "the game tells you nothing." The refusal to guide is not a defect of ANIMAL WELL; it is the medium itself. In this design lineage — Fez, Tunic, Outer Wilds — figuring out what to do next is itself the reward, and handing it out in advance would be like unwrapping the product outside the bag. But there is a condition under which guidelessness works: density, such that discoveries keep surfacing even while you are lost. ANIMAL WELL folds a contraption into nearly every screen, so time spent lost converts directly into time spent harvesting. Since that condition is met, my verdict is that the no-guidance complaint is a disagreement with the genre, not a design failure.

What about "no story, no music"? True: not one line of dialogue, and almost nothing you would call a soundtrack. But the developer's own store-page declaration was always "focused on tension and suspense over action." The absence of story breaks no promise — none was made. The sound design, built from ambience and flickering light, is to my ear the most precisely engineered part of the game. That said, the core of "the enigma never pays out" deserves a return visit later: it is a fact that players who dug expecting lore found no narrative core at the bottom.

Then the "it's actually a platformer" problem and the "wiki game" problem. The former is factually correct: between its thinking puzzles the game inserts gates of execution — fleeing a ghost, dodging instant-death traps. The latter is also a fact: the deepest tier of the sixteen secret rabbits is designed on the assumption of cryptography and collective intelligence. The question is whether these are defects or wagers. This, I think, is the true watershed between Komugi's review and the negative ones.

Where I Agree

I agree with the first claim: the retry cost. Difficulty of thought and tedium of transit are different currencies. In the loop of form a hypothesis → travel to test it → fail → walk back from the save point, the walking-back adds nothing to the thinking. It is pure friction. Stephen's Sausage Roll ships with single-step undo and Baba Is You with unlimited undo because their designers know that lowering the price of failure multiplies experiments, and more experiments means the puzzle gets played more deeply. ANIMAL WELL deliberately kept the price of failure high — and the tension it bought, I think, does not cover the total volume of experiments it destroyed.

And this is Komugi's blind spot. That review articulated the pleasure of noticing with great care, but never counted the round-trip fare of going to verify what you noticed. Nor does the 8.7 itemize the structure in which chases and instant-death traps stand as a reflex tariff in front of the thinking content — there are players who have already solved the puzzle in their heads but cannot see past it because their fingers won't keep up. Of the negative reviews' five claims, these two, at minimum, are legitimate invoices the praising side ought to pay.

Where I Push Back

Now the rebuttals. First, the "mystery for mystery's sake" thesis. It is true that there is no narrative core at the bottom of the dig — but what this game's depths pay out is not story, it is structure. The second ending, the walls' hidden faces exposed by the UV light, the flute melodies, the figures that emerge when you connect the dots: every dig returns the discovery that the world had been concealing yet another way to read it. Define the payout as "collecting lore" and yes, the game defaults. Define it as "re-reading the world" and few games pay as generously. The negative reviews admit only the first definition. That is a statement of taste, not a sustainable indictment of design.

Nor do I join the "wiki game" charge. The deepest layer does presuppose collective intelligence, but the layer structure is explicitly built so that the deeper you go, the more optional it becomes. Finishing the main game is open to everyone; the second ending is open to any patient individual; only the final layer is addressed to a community. That is not a lockout — it is invitations with different addressees per layer. If not being able to solve everything alone is a defect, then collaborative crosswords and rope-teams on a mountain are defects too. Komugi's line — that this is a declaration of where the work placed its bet — is accurate, and I have nothing to add to it.

Closing argument on "the game tells you nothing." Most complaints about the missing hint system are fused to a separate grievance: the hard-to-read map. And the map's abstraction is, in this game, a feature. If the map drew everything, the core verb — noticing a hole in a wall, a picture in the background — would die. The illegibility is the maintenance fee of the guideless design, and the real question is whether this world is worth paying that fee to read. I judge that it is. But I would not press it on everyone — and that reservation becomes the conclusion that follows.

Sources

The claims in "What the Negative Reviews Claim" are my reconstruction of recurring argument patterns from the negative Steam user reviews readable as of July 27, 2026 (roughly 13,000 reviews total, about 95% positive) and from Metacritic's user reviews (user score 8.0, with roughly one in ten negative). They are not verbatim quotations of any specific reviewer.

Six families of criticism were consulted: sparse save points and retry cost; the absent hint system and the hard-to-read map; the absence of story and music; the demands of precision platforming and chases; a deepest layer premised on collective intelligence; and price against play time. In examining them, I read each against the developer's own store-page design declarations — tension and suspense over action, multiple uses for every tool, and the main game being only the beginning.

Conclusion

Verdict. This game is wrong for you if fruitlessness sets in without explicit goals and hints, if collecting a story is the center of your reward system, and — most importantly — if you are a pure-logic puzzle player who loves thinking but goes cold the moment reflexes are demanded. The chases, the instant-death traps, and the walking-back are real. The negative Steam reviews are an accurate signal from this audience, not noise.

It is right for you if filling in a map's blank spaces by your own hand is a pleasure, if the "I see it now" moments of Fez, Tunic, and Outer Wilds are among your life's collectibles, if learning that the well keeps going below the credits makes you glad. For this audience there are eight-plus hours here worth full price. Whether to follow it to the deepest layer is a decision you can make after you have descended.

Here is where I stand against Komugi's 8.7. Of the five negative claims, two invoices are legitimate — the retry cost and the reflex tariff — and Komugi's review failed to book them. The other three are not indictments of the design but disagreements with the genre. Netting it out, I would draw this well's water at around 8.2. That it remains a well worth descending — on that point, Komugi and I have no quarrel.