Overview
Chess has a gatekeeping problem. Not by design, but by accumulation: five centuries of opening theory, memorized lines, and a culture that treats the Sicilian Defense as table stakes. Really Bad Chess attacks that problem at the root. Instead of teaching you the canon, it burns the canon down.
The premise is simple. Every match gives you one king and 15 randomly selected pieces. You might command eight knights and a phalanx of bishops. Your opponent might open with a wall of queens. The board is the same, the movement rules are the same, checkmate is still the goal — but every scrap of opening theory ever written is instantly worthless. What’s left is pure tactical reading, which turns out to be the part of chess most people actually enjoy.
The pedigree matters here. Designer Zach Gage has built a career out of respectfully vandalizing classic games — SpellTower did it for word searches, Good Sudoku for number grids, Flipflop Solitaire for Klondike. Really Bad Chess launched on iOS on October 13, 2016 and later reached Android via publisher Noodlecake, drawing the kind of press coverage most indie games only dream about, including a feature on NPR’s Weekend Edition and TouchArcade calling it “an essential puzzle experience for anyone even remotely interested in strategy games.”
Nearly a decade on, it occupies a strange market position: widely admired by critics, sitting at a merely decent 3.8/5 from store users. That gap is worth examining — and it tells you almost everything about who this game is really for.
Key features
- Randomized Armies: One king, 15 random pieces per side. Opening theory is dead; every match demands improvisation from move one.
- Adaptive Difficulty: Win and your next random army skews slightly worse than the AI’s; lose and it improves. The game continuously calibrates itself to your skill.
- Rotating Challenge Modes: Ranked play against the AI, Daily and Weekly Challenges for competitive routine, and a pressure-free Freeplay mode.
- One Honest IAP: A single purchase kills ads, unlocks local two-player Versus, adds color palettes, and grants extra undos. No currencies, no drip-feed monetization.
Analysis
The Design Insight
The smartest thing about Really Bad Chess isn’t the randomness. It’s what the randomness reveals: chess was always two games wearing one trench coat. There’s the memorization game — openings, established lines, prep — and the tactics game — reading the board, spotting forks, calculating three moves deep. Traditional chess forces you through the first game to reach the second. Gage simply deletes the first one.
The result is that a total novice and a club player face roughly the same thing on move one: an unfamiliar board that must be read rather than recalled. That’s a radical flattening of chess’s learning curve, and it’s why the game earned comparisons to nothing, because nothing quite like it existed.
The Gameplay Loop
Matches are short and dense by design. Because armies are lopsided, positions become tactical almost immediately — no fifteen moves of careful development before anything interesting happens. Players will frequently face material situations chess literature has no name for: three queens bearing down on a knight cluster, a pawn-less midgame, a king hiding behind six bishops. Each configuration is a fresh puzzle, and the adaptive difficulty system is the engine that keeps it fresh. Win, and the next army gets a little worse relative to the AI’s; lose, and it improves. It’s an elegant ELO substitute that requires zero matchmaking infrastructure, and it means matches are nearly always contested — perpetually at the edge of a player’s ability.
The Daily and Weekly Challenges give the loop a routine — a shared seed to test yourself against — while Freeplay offers a consequence-free sandbox. It’s a lean mode structure, but each mode earns its place.
Where It Bites
Now, the catch, and it’s the one users have complained about consistently: the AI does not condescend. Player reviews repeatedly note that it “always seems to perform the best possible move,” and for a game whose entire pitch is accessibility, that’s a real tension. The adaptive system compensates by handing you better material, not a dumber opponent — so beginners win by overwhelming force rather than by outplaying a fallible rival. Some players find that empowering. Others find it demoralizing, and the 3.8/5 user rating against glowing critical reception maps almost exactly onto this divide. Critics evaluated the design; users lived with the losses.
The free tier’s ads are the other friction point. They’re the standard mobile toll, and the single IAP removes them cleanly — but the fact that local two-player Versus is paywalled stings slightly, since pass-and-play random-army chess is one of the game’s best ideas.
The Undo Economy
Extra undos, bundled into the IAP, deserve a mention. In a game this tactically volatile, an undo isn’t a cheat — it’s a teaching tool. Take back a blunder, see why it was a blunder, try again. It nudges the game toward being a tactics trainer in disguise, which may be its most durable value.
Visuals & performance
Gage’s visual signature is all over this: flat colors, clean geometry, typography doing the heavy lifting. The board is high-contrast and instantly legible, which matters enormously when parsing eight knights at a glance. The unlockable color palettes are a pleasant bonus rather than a feature.
Is it austere? Yes — bordering on plain. There’s no animation flourish, no atmosphere, nothing that would read as “production value” in a trailer. But the sparseness is load-bearing. A game about reading chaotic positions cannot afford visual noise, and this one has none. As a lightweight 2D app, it’s undemanding of hardware by design, though some users on older iOS versions have reported compatibility snags — worth checking on legacy devices.
Platform nuances
On iOS, this is the canonical version — it launched there first, and the tap-to-select, tap-to-move interaction is built the way chess on a touchscreen should be. Session length fits the platform well: a match squeezes into a commute stop or a coffee line.
The Android build, published by Noodlecake, arrived later and is functionally the same game. Same modes, same IAP structure, same clean board. Android players lose nothing in translation.
What you won’t find is this game anywhere else. There is no Steam version and no Switch version — despite this being, frankly, an obvious fit for both. If you want Really Bad Chess, your phone is the only door in.
Final verdict
Really Bad Chess is one of the sharpest design arguments shipped on mobile: proof that chess’s appeal was never the openings, but the reading of the board. Nearly ten years after launch it remains without a true imitator, which says something about how hard this trick actually is. The unforgiving AI and standard free-tier ads keep the user experience a notch below the concept, and the absence of any online or PC/console presence limits its reach. But as a free download with a single fair purchase, it delivers more genuine strategic novelty than apps with far bigger budgets. Chess purists will call it heresy. They should play three matches first.
Reviewed by the Rankeno Puzzle & Casual Desk. Scores are out of 10; the overall score is the desk’s editorial judgement, informed by the four sub-scores above. How we review
