Dicefolk

Every roguelite deckbuilder since Slay the Spire has sold you the same emotional contract: you draft a plan, the random number generator tests it, and you learn to love the churn. Dicefolk tears that contract up. Developed by LEAP Game Studios — the Lima-based team behind the beat-‘em-up Tunche — in collaboration with Tiny Ghoul, and published by Good Shepherd Entertainment, it takes the most notoriously chaotic object in game design, the die, and quietly strips out the chance.

You play as Alea, a Chimera Summoner in a world where the Dicefolk have been overrun. The setup is standard-issue monster-collecting comfort food: build a team, walk a branching map, punch a boss, unlock a harder version of the same. What isn’t standard is the combat. Each battle presents a shared dice pool covering both your Chimera and the enemy’s, and you spend faces from it to trigger attacks, blocks, heals and specials — on either side of the field. You also set the turn order for your own monsters. The enemy doesn’t roll against you. You allocate the enemy’s actions, which means you also allocate their mistakes.

That single inversion moves Dicefolk out of the deckbuilder aisle and into a shelf closer to chess problems or Into the Breach. Outcomes are near-deterministic. A fight you lose is a fight you misread, not a fight the dice stole from you.

The market has noticed. The iOS release sits at 4.2 out of 5. It ships on Steam, Nintendo Switch, iOS and Android — the mobile versions free to try with a single unlock purchase, no ads, no virtual currency, which counts as a small act of defiance. It’s a confident, unusually principled release from a mid-size studio that clearly had one strong idea and refused to dilute it.

• Two-Sided Dice Control: You spend faces from a shared pool to drive both your Chimera and your opponents. Randomness becomes an input you manage, not a force you endure.
• Turn-Order Manipulation: Beyond the dice, you dictate the sequence your Chimera act in — the difference between a wasted heal and a perfect one.
• 100+ Hand-Drawn Chimera: A genuinely large recruitment pool, each monster with its own die faces and role: tank, damage dealer, status applier.
• Die-Face Modification: Items and relics let you rewrite the faces themselves, which is where real build identity emerges — stacking poison triggers, converting blocks to counters, engineering a team that wins in three turns flat.
• Short, Escalating Runs: Branching maps, boss encounters, trials and unlockable challenge tiers, sized for 30-to-45-minute sessions.
• Consumer-Friendly Packaging: 49 achievements, Steam Cloud, ten-language localization on PC; a one-time unlock on mobile.

The Gameplay Loop

Dicefolk’s combat is a puzzle box that resets every encounter. You see the enemy’s dice, your dice, everyone’s health and status. You have complete information. What you don’t have is enough resources, and that gap is the entire game.

The moment-to-moment texture is closer to a logic problem than a battle. Say the enemy pool holds a big attack face and a heal face. You can’t stop them from acting — but you choose which face they burn. Feed the boss its heal while it’s already at full health and you’ve deleted a turn from its clock. Sequence your tank to soak before your glass cannon swings and you keep a fragile build alive one round longer. Wins are earned in a specific, mathematical way that most roguelites can’t offer, because most roguelites reserve the right to roll a critical hit and end your run for reasons that have nothing to do with your play.

The cost of that clarity is real, and Dicefolk pays it. Determinism removes the drama. There is no heart-in-throat top-deck, no miracle survival at 1 HP. Once a fight is solved in your head, executing it is clerical. Encounters that would be tense in Monster Train are, here, closer to filling out a form correctly.

Build Identity and the Roster

The 100+ Chimera roster is the game’s other headline number, and it mostly delivers on breadth. Recruiting is where each run’s personality forms: a poison-stacking backline, a counterattack wall, a bruiser core that just wants more attack faces. Die-face modification is the deepest system on offer — rewriting a monster’s faces changes what that creature is, not just how hard it hits.

But roster size and roster impact are different things. The most common complaint in user reviews — “every run feels too similar” — isn’t about the monsters. It’s about the frame around them. The branching map, the node types, the pacing of shops and elites: those repeat with little variation from run to run. When the surrounding structure is this consistent and the combat is this solvable, novelty has to come entirely from the draft, and the draft can only carry so much weight over a long campaign.

The Difficulty Cliff

This is the flaw worth taking seriously. Dicefolk reads as gentle early — the first map is a tutorial wearing a costume. Then, from the second map onward, the curve doesn’t rise so much as step. Players report hitting a wall regardless of which Chimera or items they drew, and that phrasing matters. In a game where you control the randomness, “I got a bad run” is supposed to be off the table. When players say the build didn’t matter, they’re describing an encounter-tuning problem, not variance.

The generous reading: this is a puzzle game demanding you get better at reading fights, and the wall is the lesson. The honest reading: the onboarding ramp is too flat for too long, and then the game asks for expert-level play with almost no intermediate rung. A tighter mid-game difficulty band would fix more of the “repetitive” criticism than any content patch, because much of that repetition is players re-running the easy opening act to reach the part that actually tests them.

Friction and Flow

Turn resolution is quick by design, the UI keeps the shared pool legible, and the structure respects the 30-minute session it’s built around. Where it stumbles is in giving you enough forward information: with determinism as the core promise, anything that obscures an upcoming encounter’s shape feels like a design contradiction rather than mystery.

The hand-drawn Chimera are the strongest part of the presentation — expressive, chunky, closer to a well-made card game’s illustration set than to the flat vector look that dominates indie roguelites. Each creature reads instantly at a glance, which is functionally important when you’re scanning a five-monster board for the one with the counter face.

The UI is competent and clear rather than beautiful. The dice pool, the turn order and the status icons are all where you need them, and information density stays manageable on small screens. That’s harder than it sounds and deserves credit. What’s missing is flourish: attack animations are perfunctory, and a game this deterministic could use more visual payoff for a well-solved fight. When there’s no luck to celebrate, the presentation has to sell the satisfaction, and here it merely reports it.

Technically, a turn-based 2D game of this scope is light on hardware, and the art style should scale from a Switch handheld to a modern PC without compromise.

Steam is the fullest version. Mouse input suits a game about precise selection from a crowded pool, and the PC build carries the complete feature set: 49 achievements, Steam Cloud saves, ten-language localization. If you want the completionist scaffolding — trials, challenge unlocks, achievement chasing — this is the build.

Nintendo Switch is arguably the best fit. Short runs, turn-based pacing, no reflex demands: this is handheld-native design. Cursor-based navigation on a stick is inherently slower than a mouse, but nothing in Dicefolk is time-pressured, so the friction never bites.

iOS and Android are the interesting case. Touch is a natural match for dragging dice faces onto targets, and the mobile business model is the cleanest in the lineup: free to try, one unlock purchase, no ads, no currencies, no energy timers. In a storefront where “tactical roguelite” usually means a gacha in a trench coat, that’s worth paying for on principle. The tradeoff is screen real estate — a board with a shared dice pool and five monsters wants more space than a phone gives it. A tablet is the natural home; on a phone, the density invites mis-taps.

Dicefolk earns its shelf space on the strength of one idea executed with discipline. Handing the player both sides’ dice is the kind of design inversion that most studios would soften into a gimmick — a once-per-fight ability, a special mode. LEAP Game Studios and Tiny Ghoul built the whole game on it, and the result is a tactical roguelite where skill and randomness finally stop arguing.

It’s not a complete triumph. The difficulty tuning after the opening act is the sort of problem that turns curious players into churned players, and the map structure doesn’t work nearly as hard as the combat system does. The repetition sets in well before the roster runs out of surprises.

But the ceiling here is high, $14.99 is a fair ask for a paid game of this depth, and the mobile release is a rebuke to an app store full of predatory clones. We recommend it to puzzle-strategy players who value solvable fights over spectacle — buy it for the mechanic, stay for the roster, and expect the wall, even if it arrives with less warning than it should.