Mahjuro

Mahjuro wants to do to mahjong what Balatro did to poker

The Balatro-likes keep coming. Most are bad. A handful are interesting. Mahjuro, a solo-dev project from Zelda Hessler heading to Steam in mid-2026, is one of the rare ones that picked a genuinely smart base game to bolt the roguelite scoring engine onto.

Mahjong already has yaku. It already has scoring multipliers tied to hand purity, and a whole culture of breaking your own rules to chase bigger numbers. The concept isn't the hard part. The hard part is keeping the math from collapsing the second you stack a hundred relics on top of it.

The pitch, briefly: you draw tiles, form pairs, sequences, triplets, kongs, cash them in for escalating scores, and push through seven wings and twenty-one chambers of a gambling house that openly cheats. Bosses lock out tiles. Scoring gets twisted. Between rounds you buy relics, tile packs, talismans, zodiacs to keep up. Fifteen yaku to unlock and level. More than a hundred relics, some of which the description openly admits are designed to break the game on purpose.

Why mahjong actually fits, and why that's the trap

Real mahjong, riichi specifically, is one of the deepest scoring systems any tabletop game has ever produced. Yaku stack. Dora multiply. A clean hand can quintuple in value because of one quiet little condition you set up four turns earlier. If you've ever watched a serious riichi player walk through a 24,000-point hand, you already see why this maps to deckbuilder logic almost too cleanly.

The trap is the audience.

Balatro got away with poker because everyone on Earth knows a pair beats a high card. Mahjong's hand structure is a brick wall for anyone who hasn't put twenty hours into the base game. Hessler's design choices in the v0.5.0 beta footage suggest some awareness of this. The UI seems to want to teach yaku as you encounter them rather than dumping the rulebook on day one. Whether that holds up under the weight of a hundred-plus relics that rewrite scoring is the open question. Roguelite deckbuilders live or die on whether a new player can lose their first run and still understand what killed them.

What the early footage shows

The two gameplay videos out so far, an alpha v0.1.1 and a beta v0.5.0, are doing the thing solo devs do when they're still finding the loop: posting raw runs without commentary and letting the systems talk. The jump between builds is significant. Tile interactions are denser, the shop layer looks more populated, and the visual language has tightened around the "shuttered gambling house" framing instead of the placeholder UI from the alpha.

It's still recognizably an in-development build. Animations are stiff. Some of the relic art feels like draft work. But the scoring loop is clearly there, and clearly already capable of the Balatro thing where your number suddenly has too many digits.

Community interest, as far as it exists this far out from release, is concentrated where you'd expect: the small but very loud overlap of Balatro grinders, mahjong heads, and the roguelite-deckbuilder Discord ecosystem that obsessively dissects every new entry in the genre. Nobody is writing thinkpieces about Mahjuro yet. That's fine. Nobody should be.

The competition problem

Here's the part that's harder to ignore. Mahjuro is not the only mahjong-flavored deckbuilder in development. It's not even the only one trying the Balatro pivot. The 2024-2025 deckbuilder gold rush produced a lot of half-finished prototypes, some of which will ship in the same window as Mahjuro.

The ones that survive will be the ones with a genuine mechanical wrinkle the others lack, or a clear visual and tonal identity that's recognizable in a thumbnail. Preferably both.

The "house that does not play fair" framing is doing real work. It's a hook. Bosses that actively break rules, restrict tile pools, force adaptive play, that's closer to Slay the Spire's Act bosses than to Balatro's blinds, and it's an interesting place to land tonally. Whether the relic design supports that promise across twenty-plus boss ordeals is the thing nobody can answer until a wider beta or launch.

The marketing reality

A game like Mahjuro cannot be marketed the way a publisher markets a normal Steam release. The TAM for "mahjong roguelite deckbuilder fans" is not the kind of number that survives a $50k ad spend. Mass-market Steam discovery campaigns would torch the budget chasing impressions from people who bounce the second they see tile racks instead of playing cards.

The realistic path is the inverse. Find the maybe four thousand people on Earth who would unironically pre-order this on title alone, and the maybe forty thousand who'd buy on a trusted recommendation, and reach them through the creators they already watch. Riichi mahjong YouTubers and streamers. The Balatro-adjacent roguelite analysts who do hour-long deckbuilder breakdowns. Small but devoted Japanese tabletop channels. The indie-roguelite curators on Twitch who built audiences around scoring-engine games specifically. This is the kind of problem CreatorFetch positions itself for at this scale, a way to actually find and contact hyper-specific creators by niche instead of cold-emailing the top 500 gaming channels and praying. For a solo dev like Hessler, that targeted-outreach layer is the difference between launching to a pre-warmed audience that already knows what a kong is, and launching into the Steam new-release silence.

Worth watching, with caveats

Mahjuro is the kind of project that could legitimately end up as a small cult hit, or could get buried under the launch-week pileup of June 2026.

The fundamentals are promising. The base game is rich enough to support the scoring-engine treatment. The dev is shipping visible build progress publicly, which is the right signal. The risk is the onboarding problem and the genre saturation, both of which are real, and neither of which can be hand-waved away with clever relic design alone.

If the next public build, whenever it lands, shows a tutorial layer that can teach yaku without insulting players who already know them, this jumps from "interesting" to "actually track this." For now it's on the watchlist, which is already more than most roguelite deckbuilders get.