Key art for Random Chaos Dungeon

Game review

Random Chaos Dungeon

6.5 / 10

Developer
denhonator
Publisher
denhonator
Released
6 March 2026
Platforms
Windows
Genre
Action
Monetisation
Free-to-play, no in-app purchases found
Players
Local co-op (up to 4 players)

The short version

Random Chaos Dungeon is a free co-op roguelike from solo developer denhonator, released 6 March 2026, built around a genuinely distinctive hook: not just the loot but the enemies, bosses, and their attacks are all randomly generated on every run. This review's full 22-review sample sits at Positive overall (20 positive to 2 negative), and the developer's own six-month patch history shows sustained, hands-on solo support, including a tutorial overhaul that directly answered a specific complaint raised in the sample.

Biggest strengths

  • This review's sample runs 20 positive to 2 negative, and repeatedly credits the game's central hook directly, with reviewers describing runs built around specific randomized combinations, one writing 'black hole with blood magic carried my run, rare no shield run, 5/5' and another describing 'endless weapon variety making every run feel unique'
  • This review checked the developer's public partner-events feed and found 10 posts spanning six months, from a 17 March balance pass through a 5 September controller and hotfix update, consistently written in first person by what reads as a single solo developer working through community-reported bugs in real time
  • This review's sample includes a detailed Polish-language reviewer who wrote out a full unofficial control scheme because none was provided at launch, scoring the game 8/10 anyway and calling the gameplay itself 'engaging' despite that gap; this review confirmed via the partner-events feed that an August 5 update specifically added a control list and expanded tutorial guidance

Biggest weaknesses

  • This review's sample includes a negative reviewer who wrote that the game reads as 'clearly someone's Gamemaker project that they thought the premise alone would carry them on,' and this review confirmed the developer's patch notes are written in first-person, solo-developer language ('I ran into while testing,' 'I don't know why yet'), consistent with a small, still-maturing production
  • This review checked the game's original launch state via its earliest partner-events post and found no dedicated tutorial content until an August 5 update landed five months after release, meaning the game's full 6-8 March launch window relied on player-written guides like the Polish-language reviewer's control breakdown
  • This review checked the Steam global achievement API directly and found even the most common non-starter achievement sits at only 16.2% completion, with most tracked achievements clustering between 9.6% and 10.6% and the game's secret-content achievement at just 2.0%, consistent with reviewer descriptions of a high, only partly adjustable difficulty ceiling

Who it's for: Roguelike and co-op fans looking for a free, chaotic, high-difficulty pick with real replay value; less suited to players wanting a polished, guided introduction, since this review found the game shipped without a proper tutorial and leaned on its community to fill the gap for months before one arrived.

Our recommendation: A rough-edged but genuinely inventive free roguelike whose randomization hook holds up across this review's sample, made better by a developer who has kept fixing and expanding it for half a year without asking for a cent in return.

6.5 / 10

Even the bosses don't know what they're going to do next. Random Chaos Dungeon is a free co-op roguelike from solo developer denhonator, released 6 March 2026. This review draws on the game's full 22-review sample, its global achievement data, and six months of the developer's public patch history.

A roguelike where nothing is fixed, not even the enemies

This review checked the game's own store description, which frames the hook plainly: "even enemies, bosses, and their attacks are randomly generated," on top of the genre's usual randomized loot and layouts. This review's sample treats that as a genuine differentiator rather than marketing language. A Spanish-language reviewer called it "an entire game randomized with RNG, the enemies the layout and weapons," praising boss patterns specifically for staying "surprising" even across repeat runs, while a Polish-language reviewer's detailed writeup, despite docking points elsewhere, still called the underlying design "Unique."

Chaos as a feature, praised across languages

This review's sample repeatedly credits specific, memorable moments produced by that randomization. One reviewer described a run carried by a "black hole with blood magic" combination, calling it a "rare no shield run" and rating it "5/5. Simple game, best with friends!" Another described a healing staff that accidentally launched a friend "at Mach speed down a ravine," also rating it 5/5. A longer-tenured reviewer with over 1,800 minutes played summarized the appeal more plainly: "endless weapon variety making every run feel unique."

A missing tutorial, fixed five months later

Set against that praise, this review's sample flags a real early gap. A detailed Polish-language reviewer, scoring the game 8/10 overall and calling the gameplay "engaging," nonetheless wrote out an entire unofficial control scheme in their review because "I didn't find tutorial while playing the game," listing movement, attack, parry, and dash bindings by hand for other players to find. This review checked the developer's public partner-events feed directly and confirmed that gap wasn't addressed until an August 5 "Tutorial Update," five months after launch, which added a control list to the help section, expanded in-game tutorial wording, and let players skip the controls explanation once they knew it.

A solo developer's visible hand

This review checked the partner-events feed in full and found 10 posts spanning six months, from a 17 March balance pass through a 5 September controller and hotfix update, consistently written in first-person, hands-on language rather than studio press-release copy. Beta-branch patch notes describe the developer testing risky changes personally before promoting them to the main build, writing that a given patch is held back "until I've confirmed the changes aren't game breaking in any way," and one bug fix note simply admits "I don't know why yet." That voice runs through balance changes, boss-specific fixes, and a September "Secret Content Update" that took previously-beta content fully live.

Achievement data suggests a genuinely hard, rewarding endgame

This review checked the Steam global achievement API directly and found a strikingly low, tightly clustered completion curve: even the most commonly earned tracked achievement, for defeating a boss, sits at just 16.2%, with most others clustered between 9.6% and 10.6%, and a dedicated secret-content achievement at only 2.0%. Combined with the sample's own description of the game's difficulty as "high" but adjustable, and the developer's own August 19 patch notes explicitly reworking how shield defense interacts with damage caps, this review reads that data as consistent with a genuinely demanding endgame rather than simply an unfinished one.

Co-op is the game's best-loved mode, and still under-explained

This review's sample repeatedly singles out local co-op as the game's strongest draw, with reviewers logging some of the sample's longest playtimes specifically in multiplayer sessions, including one with over 1,800 minutes calling it "super fun in Co-op" and another with nearly 1,500 minutes praising "a lot of variety in the types of weapons and enemies you can find" especially playing with others. That enthusiasm sits alongside a genuine gap: this review's sample includes an Italian-language reviewer who asked directly in their own review "how does the co-op work and do u guys have a Discord server," a basic onboarding question left unanswered in the review itself, reinforcing the same early-documentation shortfall this review found in the game's original missing tutorial.

A minority "unfinished project" complaint

Not every reaction in this review's sample is favorable. One negative reviewer wrote bluntly that the game "is clearly someone's Gamemaker project that they thought the premise alone would carry them on," adding only "at least it's free." A second negative review reads in full: "Absolute garbage..." with no further detail. This review notes both landed within the game's first six weeks, before the bulk of the developer's patch history had occurred, alongside a more measured early reviewer who wrote "it's getting there, but not ready yet."

Free, adjustable difficulty, and a fair accessibility showing

This review checked the store's accessibility feature-flag block directly and found seven of nineteen tracked flags true: camera comfort, background volume controls, stereo sound, no mandatory quicktime events, a keyboard-only control option, adjustable difficulty levels, and save-anytime support. The difficulty modifiers go further than the flag alone suggests; the developer's own patch notes describe an option that uncaps damage taken specifically to allow infinite revives on Easy, while making Hard harder still, a genuine attempt to widen who the game's brutal randomness is accessible to.

Verdict

Random Chaos Dungeon backs up its central promise: this review's sample repeatedly credits genuinely chaotic, memorable runs built from a rare combination of randomized enemies, bosses, and loot, and six months of visible, solo, community-responsive patching, culminating in a tutorial fix that directly answered a specific complaint raised in this review's own sample. It launched rough, with real gaps a couple of early reviewers understandably called out, but for a completely free game with this much sustained attention behind it, that's a trade worth making.

Pros and cons

Pros

  • This review's sample runs 20 positive to 2 negative, and repeatedly credits the game's central hook directly, with reviewers describing runs built around specific randomized combinations, one writing 'black hole with blood magic carried my run, rare no shield run, 5/5' and another describing 'endless weapon variety making every run feel unique'
  • This review checked the developer's public partner-events feed and found 10 posts spanning six months, from a 17 March balance pass through a 5 September controller and hotfix update, consistently written in first person by what reads as a single solo developer working through community-reported bugs in real time
  • This review's sample includes a detailed Polish-language reviewer who wrote out a full unofficial control scheme because none was provided at launch, scoring the game 8/10 anyway and calling the gameplay itself 'engaging' despite that gap; this review confirmed via the partner-events feed that an August 5 update specifically added a control list and expanded tutorial guidance
  • This review checked the store's accessibility feature-flag block directly and found seven of nineteen tracked flags true, including camera comfort, a keyboard-only control option, and adjustable difficulty modifiers that the developer's own patch notes describe extending as far as uncapped damage taken on Easy for players who want infinite revives

Cons

  • This review's sample includes a negative reviewer who wrote that the game reads as 'clearly someone's Gamemaker project that they thought the premise alone would carry them on,' and this review confirmed the developer's patch notes are written in first-person, solo-developer language ('I ran into while testing,' 'I don't know why yet'), consistent with a small, still-maturing production
  • This review checked the game's original launch state via its earliest partner-events post and found no dedicated tutorial content until an August 5 update landed five months after release, meaning the game's full 6-8 March launch window relied on player-written guides like the Polish-language reviewer's control breakdown
  • This review checked the Steam global achievement API directly and found even the most common non-starter achievement sits at only 16.2% completion, with most tracked achievements clustering between 9.6% and 10.6% and the game's secret-content achievement at just 2.0%, consistent with reviewer descriptions of a high, only partly adjustable difficulty ceiling
  • This review's sample includes one especially thin negative review reading only 'Absolute garbage...' with no further detail, and another more measured reviewer writing 'it's getting there, but not ready yet,' both landing in the game's first two weeks of release before the bulk of its six months of patching had occurred

Category scores

Category scores are the reviewer's read on each area. They inform the overall score but do not average into it — the number at the top of the page is a judgement, not a calculation.

  • Gameplay 6.5 / 10
  • Value 8 / 10
  • Originality 7.5 / 10
  • Accessibility 4 / 10