Card-en-Ciel

There’s a particular kind of game that only exists because a studio has been around long enough to accumulate ghosts. Inti Creates has spent two decades building other people’s franchises — they co-developed Mega Man 9 and 10, they built Mega Man Zero and ZX — and then spent another decade building their own: Azure Striker Gunvolt, Blaster Master Zero, Gal*Gun, Grim Guardians. Card-en-Ciel is what happens when a studio decides to raid its own attic and turn the contents into a card game.

The premise is deliberately silly and quietly clever. You play Neon, self-appointed “Gaming Chair Detective,” who dives into Rust Tactics — an unfinished full-dive VR game — after its developer, Ancie, reports that characters from completely unrelated games have started bleeding into the world. That framing device does real structural work. The framing device justifies a roguelite’s fundamental weirdness, where incompatible systems collide in a single run and the world resets when you die. Unfinished game. Corrupted data. Of course your deck is a chaotic pile of characters who shouldn’t be in the same room.

Commercially, this thing landed in an awkward spot. It shipped into a market where “roguelite deckbuilder” is closer to a warning label than a pitch — a genre so thoroughly farmed since Slay the Spire that most releases get maybe a weekend of attention. Card-en-Ciel avoided that fate by not really being one. The grid combat is the differentiator, and it’s not cosmetic. It’s the thing the entire design hangs on.

• 6x3 Grid Combat: Battles play out on a Battle Network-style field. Your position and your enemy’s position determine whether a card is devastating or completely wasted. Cards have shapes — columns, rows, splashes, single tiles — and reading the board is as important as reading your hand.
• 300+ Voiced Character Cards: Every card is a character, fully voiced in English and Japanese, who talks back mid-fight. Cards can be leveled, duplicated, and fused, which turns deckbuilding into a persistent project rather than a per-run scramble.
• Ten-Plus Themed Dungeons with Branching Routes: Each dungeon apes a different game world, with 50+ random events, route choices, and permanent unlocks that carry between attempts.
• Ten Escalating Difficulty Tiers: Clearing a dungeon unlocks ascension-style modifiers that change the rules — and, critically, can be exploited rather than merely endured.
• 50 Original Vocal Tracks: Not stingers. Full songs, written to match the era and genre of each character’s home game. A 2014 synth-rock character gets 2014 synth-rock.
• Live Competitive Layer: Real-time PvP, daily dungeons, and online leaderboards.

The Gameplay Loop

Most deckbuilders are spreadsheets with art. You optimize a probability curve, you resolve it in a straight line, and the interesting decisions all happen at the card-reward screen. Combat is arithmetic.

Card-en-Ciel breaks that by adding a second axis — literally. The 6x3 grid means a card isn’t a value, it’s a value plus a shape plus a position requirement. A card that hits the front column is a dead draw when the enemy retreats to the back. A card that clears a row is a war crime when three enemies line up and a shrug when they don’t. Movement costs you tempo, so every turn is a negotiation between “play the best card in my hand” and “get to the tile where a card becomes the best card.”

This is the design’s central achievement, and it’s why positioning, not draw luck, is the skill the game is built to reward. Bad hands in a conventional deckbuilder are dead turns. Bad hands here are geometry problems. You reposition, you set up, you eat a small hit to make next turn devastating. The randomness gets converted into a puzzle instead of a tax.

The Card Economy

The card-as-character conceit could have been pure fan service. It isn’t, because Inti Creates attached real mechanical weight to it. Leveling, duplicating, and fusing turn your collection into something you cultivate across dozens of runs, not something you assemble and discard. That’s a meaningful departure from the genre’s usual amnesia, where every run starts from a blank slate and the only persistent progress is unlock-gated variety.

There’s a cost. A pool of 300+ cards is too large to hold in your head, and the game’s onboarding does not adequately prepare you for that. Early runs involve a lot of squinting at unfamiliar characters trying to parse their attack shapes under time pressure. The information architecture is dense — arguably cluttered — and the learning curve is front-loaded in a way that will lose impatient players early. That’s a real design failure, not a matter of taste. A game with this much depth owes you a better ramp.

Difficulty and the RNG Problem

The sharpest structural criticism of the design is RNG swing in the later dungeons. The ten-tier difficulty system stacks rule modifiers, and the interaction between aggressive modifiers and unlucky event rolls can produce runs that are effectively decided before you’ve made a real decision. The design’s defense is that modifiers are meant to be turned to your advantage, and clever builds can weaponize them. But that defense assumes system mastery the game doesn’t reliably teach, and it doesn’t fully account for the compounding variance at the top tiers.

The other honest knock: repetition. Ten-plus dungeons sounds like a lot until the event pool starts recycling deep into a campaign. The 50+ random events are well-written, but a roguelite lives or dies on whether the late game feels different from the early game, and the middle stretch sags.

The Story Problem

Pacing between battles is slow. The Gaming Chair Detective framing is charming and the dialogue is better than it has any right to be, but there is a lot of it, and it arrives in the gaps where a roguelite normally wants you clicking “next room.” The rhythm of a run keeps getting interrupted by the rhythm of a visual novel. Both are fine. Together they fight.

And the crossover cast — this is the uncomfortable part — lands significantly harder if you already know Inti Creates’ back catalogue. For Gunvolt and Blaster Master Zero veterans, the roster is a long-overdue reunion. For everyone else, they’re 300 anime strangers with good attack patterns. The mechanics carry it either way. The emotional payload does not.

The art direction is sharp, high-contrast, and unapologetically anime — pixel-adjacent sprite work with modern effects layered on, which is Inti Creates’ house style refined over twenty years. It reads cleanly, which matters enormously on a grid where you need to parse three enemy positions and a telegraphed attack shape in under a second. The visual clarity is doing serious functional work.

The UI is the weak link. It’s dense to the point of intimidation, with card information competing for attention against the board, the hand, and a character portrait that’s talking at you. Once internalized, it’s efficient. Getting there is friction the game inflicts on itself.

The audio is the standout, and it isn’t close. Fifty original vocal tracks, each written in the era-appropriate genre of its character’s source game, is an absurd amount of production for a mid-budget release. It’s the kind of excess that only happens when a team is having fun. And as a 2D grid game with modest technical demands, this is the kind of release that should be light on hardware across the board.

Steam is the better version for the competitive layer. Mouse input suits card selection and grid targeting, ultrawide and high-refresh support are there, and the real-time PvP and leaderboard chase benefit from a desktop’s input precision and connection stability. If you intend to push the ten difficulty tiers or take PvP seriously, this is where you should be. It’s also a natural fit for handheld PC play — turn-based, low demands, well-suited to short sessions.

Nintendo Switch is the better version for how most people will actually play it. Runs chunk naturally into commute-sized pieces, and the pick-up-and-drop rhythm of a roguelite is exactly what portable hardware is for. The tradeoff is real: docked play means navigating a dense UI with a stick, which adds a beat to every card selection. Over a long run, that adds up. Handheld touch input smooths most of it over.

Pick Steam if you want the ceiling. Pick Switch if you want the habit.

Card-en-Ciel is a game made by people who stopped asking permission. The grid-plus-cards fusion isn’t a gimmick layered on a proven formula — it’s a structural change that addresses the genre’s oldest complaint, that a bad draw is a wasted turn. Here, a bad draw is a repositioning problem, and that single shift makes hundreds of small moments interesting that would otherwise be dead air.

It is not a tidy game. The UI is overloaded, the late-game balance is loose, the story keeps stepping on the combat’s toes, and the emotional core is priced in fan knowledge that most buyers won’t have. Those are legitimate marks against it, and they’re why this lands as a solid-but-flawed recommendation rather than an easy one.

Still, the value proposition is hard to argue with. This is a mid-budget release carrying a fully voiced 300-card roster and fifty original songs at $24.99 — an ambition-to-price ratio most publishers wouldn’t dare. The design flaws are real, but they’re the kind that matter most to newcomers and completionists; the core loop is built to reward the audience it was made for. We recommend it to that audience, and to them specifically.