The Dungeon Of Horror

A dungeon crawl where you build the dungeon as you go

The Dungeon Of Horror's central hook is a genuine structural twist on the turn-based dungeon crawler formula: rather than exploring a fixed, pre-built layout, you assemble the dungeon's own corridors and intersections from tile pieces as you move through it, with certain room combinations described as producing "unexpected results." Combined with a stated design principle that "attacks aren't random" and every threat is meant to be readable if you pay attention, the game frames careful, deliberate tile placement and threat assessment as the core skill, rather than reflexes or raw character power. The setup wrapping that system is straightforward horror-fantasy: a royal squad member searching for a lost king wakes up trapped in a shifting nightmare dimension after a seemingly innocent overnight stay turns into a trap, with the way out running through defeating the pagan shaman Meir before her ritual completes.

Ten classes, and an optional path through it with friends

Beyond the tile-building hook, the game supports ten distinct playable classes — Knight, Archer, Witch, Mercenary, Monk, Wizard, Doctor, Sorcerer, Druid and Bard, drawn from the game's own achievement list rather than heavily detailed on the store page itself — alongside an optional co-op mode for tackling the dungeon with friends rather than alone. That's a reasonably ambitious amount of build variety for what appears to be a solo-developed release, and it's a structural choice several of the sample's few reviewers specifically comment on, one calling the game "cool solo, prolly cooler with friends" and another simply urging others to "play more" with them.

What a five-review sample can and can't tell you

It has to be said plainly: this is one of the thinnest review samples this review has worked with. Of five total public reviews, four are effectively one-line reactions — "very cool game," "cool solo prolly cooler with friends," a flat "Нормально" ("Fine" or "OK" in Russian), and a Korean review that reads, in translation, as something closer to faint praise than enthusiasm: "I guess there are people who like this kind of thing." Only one review offers anything to actually dig into, and it's worth quoting in full for that reason: "Very cool game with big potential!! hope in the next update we'll see maybe more floors? and some items seems not very useful..." That's simultaneously the sample's strongest endorsement and its clearest specific criticism — real enthusiasm for the core idea, alongside a direct, credible flag that item balance isn't fully there yet.

Where achievement data fills in a gap the reviews don't

Because the review text itself says so little about actual class balance, global achievement completion rates are an unusually useful independent check here. Per-class "win" achievements range from about 6.8% for the Knight down through the Archer, Witch and Mercenary in the 4% range, several casters and support classes around 3%, and two classes — the Druid-adjacent Sorcerer tier and the Bard specifically — sitting at a flat 0%. A roughly nine-times gap between the most and least successful class, with at least one class showing literally no recorded wins, is a meaningfully larger spread than pure chance or small sample size alone would typically produce, and it lines up credibly with the one detailed review's complaint that some items — and, by extension, whatever kit a given class depends on — don't feel very useful in practice.

An early sign of continued support

One additional independent data point is worth noting: the developer's Steam news feed shows an "Update 1" posted 21 August 2026, less than two weeks after the 9 August launch. There's no detail available here on what that update changed, since the post itself wasn't part of this review's research scope beyond confirming its existence and timing, but a solo or small-team developer following up quickly after launch is a reasonable, if modest, positive signal for a game whose tiniest review already asked for more content in a future update.

A tone built around dread, not just tile puzzles

It's worth stressing that the tile-building system isn't presented as an abstract puzzle layered on top of a generic fantasy skin — the store page frames it explicitly around dread, under the pillar "Fear Of The Dark," insisting that because attacks aren't randomized, "a mistake can have grave consequences" rather than being just an unlucky roll. That framing matters for how to read the achievement spread above: if the game genuinely punishes misreads as harshly as it claims to, a wide gap between a well-understood class like the Knight and a still-unsolved one like the Bard is less surprising, and arguably closer to the intended experience than a flaw to be patched away entirely. Whether that's the right balance point for a paid release at this price is a fair question the sample simply can't answer yet.

A price that raises the stakes of the uncertainty here

At NZ$22.49, The Dungeon Of Horror costs noticeably more than most of the small, early-review-sample indie games this review covers in a given month, many of which sit in the seven-to-fifteen-dollar range. That's not disqualifying on its own — genuinely distinctive mechanics can justify a premium — but it does mean the thinness of the available review sample carries more weight for a prospective buyer here than it would for a five-dollar impulse purchase. This review checked review text across every language Steam reports for the game, not just English, so the five total reviews quoted above represent the complete public record available at time of writing, not merely an English-language subset of a larger, more informative conversation happening elsewhere.

Final verdict

The Dungeon Of Horror pairs a genuinely distinctive structural idea — building your own dungeon layout from tiles as you explore it, with deliberately readable rather than random threats — with a review sample too small and too terse to say with real confidence how well that idea holds up across a full playthrough. What can be said with more confidence, thanks to achievement data rather than the reviews themselves, is that class balance currently shows a real, credible gap between a handful of consistently successful classes and at least two that nobody has yet won with. At a higher price than most of the small-sample releases this review covers, that combination argues for caution rather than enthusiasm: genre fans specifically drawn to the tile-building hook, ideally with a group of friends to split the co-op experience, have real reason to be curious, but this isn't yet a release with enough evidence behind it to recommend without reservation.