← all deep dives

Tzolk'in: The Mayan Calendar

Deep Dive for an iOS Roguelite Translation

Project context: Translate the core loop of Tzolk'in: The Mayan Calendar (Simone Luciani & Daniele Tascini, Czech Games Edition, 2012) into a single-player iOS roguelite. The "Dominion -> Slay the Spire" pattern, but for a worker-placement game whose signature is a literal rotating gear. No competing digital port exists; the gears are the most visually digital-native mechanic in the modern Euro canon.

1. The mechanics, precisely

A turn from the high priest's seat. On your turn you must do exactly one of two things: place workers, or pick up workers. You cannot do both, and you cannot pass — if all your workers are already on the gears, you must pick at least one up. To place, you take one or more available workers from your supply and put each on the lowest visible action space of any of the five gears (gears fill bottom-up). You pay corn equal to the action-space number plus a per-worker placement fee shown on your player board (1st worker free, 2nd +1, 3rd +2…). To pick up, you remove one or more workers; for each you may either trigger the action under that worker, pay 1 corn per step to use a lower action on the same gear, or do nothing. Then play passes. Once everyone has acted, the central Tzolk'in calendar gear is rotated one notch counter-clockwise — and every worker on every gear advances one space. A worker that gets pushed past the highest space (7, or 10 on Chichen Itza) returns to its owner with no action gained — you waited too long.

The five gears, what each does at increasing tiers. Each peripheral gear is themed and offers escalating rewards as a worker rides it longer:

  • Palenque (food / forest). Tier 1 = take 2 corn. Tier 2 = 3 corn. Tier 3-5 = harvest standing wood/corn tiles from a personal jungle (or burn the forest for a one-time corn windfall while angering the gods). Tier 6-7 = free choice from anywhere on Palenque. The food gear: cheap to enter, cheap to leave, almost no escalation curve but reliable.
  • Yaxchilan (resources). Tier 1 = wood. Tier 2 = stone. Tier 3 = gold. Tier 4 = a crystal skull (a 3-VP scoring object). Tier 5 = wood + (stone or gold). Tier 6-7 = free choice. Sharply escalating: leaving a worker here through 4 rotations is the difference between a piece of firewood and a sceptre.
  • Tikal (build / advance). Tier 1 = a new worker. Tier 2 = build one building. Tier 3 = climb 1 step on any one of the four technology tracks. Tier 4 = build a monument or two buildings. Tier 5 = +1 or +2 temple steps for 1 wood. Tier 6-7 = any Tikal action free. The "double" gear — every action is bigger one tier up.
  • Uxmal (trade / flex). Tier 1 = climb 1 temple step (3 corn). Tier 2 = recruit a worker. Tier 3 = technology. Tier 4 = build a building paying its cost in any combination of resources (the famed "Uxmal gambit"). Tier 6-7 = free Uxmal action. The most flexible gear — Uxmal's mid-tiers shadow Tikal's but at a different timing curve.
  • Chichen Itza (temple / skulls). Ten action spaces (not seven). Each space deposits a crystal skull at progressively higher temple positions, scoring 4-13 VP plus advancing the player up that god's temple. Tier 10 = free choice. Pure scoring engine — Chichen Itza only matters if you have skulls (from Yaxchilan tier 4 or Theology tech), so it's a late-game gear for almost every strategy.

Corn — the universal pressure. Corn is the only currency for placing workers, and it's also the food you owe on Food Days. There are 4 Food Days per game (marked by visible teeth on the central gear, so you always see them coming). On each, you pay 2 corn per worker on the table or lose 3 VP per starved worker. Two Food Days are also "middle of age" reward days (you collect the resources/skulls from your highest temple step plus all steps below); two are "end of age" reward days (you score VP from your current temple step, plus a bonus to whichever player is highest on each temple). So corn is at once a placement currency, a starvation hedge, and the pressure that forces you to not leave too many workers out indefinitely.

The timing trade-off, made concrete. Imagine round 1. You drop a worker on Yaxchilan tier 1 — cost: 1 corn. After your turn the gear ticks; that worker is now on tier 2. Three options now, every subsequent turn: pick him up for stone, leave him for gold, leave him longer for a crystal skull. Every round you wait, you (a) pay zero corn (free escalation), (b) pay the opportunity cost of having one fewer worker available to place elsewhere, and (c) accumulate Food Day liability (each Food Day you pay 2 corn for him whether or not he's done useful work yet). Conversely, place a worker on Yaxchilan late — say tier 4 directly — and you pay 4 corn + placement fees, but get the skull next turn. So placement is cheap-and-slow vs. expensive-and-fast, and the rotating gear automates the passage of time itself into a resource. This is the core design primitive the prototype must preserve.

Sources: UltraBoardGames rules, BGG strategy primer, BGA Tips, CGE rulebook PDF.

2. What makes Tzolk'in work as a tabletop game

Time-as-resource: a genuinely new design primitive. Most worker-placement games (Agricola, Caverna, Lords of Waterdeep) treat time as just "round number." Tzolk'in makes time a physical input — your worker is literally being carried by a clock. The Opinionated Gamers review nailed it: the gears "are definitively not a gimmick" and serve a practical purpose, but more importantly they introduce "the element of time" that makes a familiar mechanic feel structurally fresh. Every worker placement is a wager on when you'll come back, not just where.

The "delayed gratification" dopamine moment. The signature feeling of Tzolk'in is the satisfied click of taking a worker off Yaxchilan tier 5 in round 9 — an action you set up four rounds ago, that has been ticking forward while you spent your turns elsewhere, and that now pays out a stone and a crystal skull in one shot. It's the pleasure of a slow-cooked bet resolving in your favor. There's a small heist-movie quality: you set up the pieces early, and the calendar walks them into position for you. The flip side — a worker pushed off the top because you couldn't afford to retrieve him — has the bitter sting of I waited too long. This bidirectional pleasure/regret is what the iOS port must preserve.

The visual/tactile pleasure of the rotating gears (the signature). Five large plastic gears mesh into a central calendar wheel. At end-of-round, one player physically advances the central gear one tooth, and the five peripheral gears rotate with it. Workers visibly migrate. It is the most physically satisfying end-of-round procedure in modern Eurogame design. Reviewers consistently lead with it; players show off the mechanism to non-gamers. This is enormously important for a phone port: you are translating a game whose central appeal is already animated. The board-game gears beg to become a SwiftUI animation. Heat had to invent visual delight; Tzolk'in already has it baked into the IP.

Planning urgency from a visible clock. Because the calendar gear has visible teeth — and four of them are conspicuously marked Food Days — the player always knows exactly how many turns remain until the next pressure event. Unlike Agricola (where the harvest looms abstractly), Tzolk'in shows you a literal countdown. Strategic depth comes from threading three timing layers simultaneously: (a) per-worker tier ripening, (b) Food Day spacing, (c) end-of-game Chichen Itza scoring window in the final quarter. The BGA strategy doc captures this rhythm precisely: "deliberate resource starvation in rounds 1-2 to enable explosive building in rounds 3-4."

The blocking dynamic. In multiplayer, lowest-space-only placement creates a brutal positional fight. If you covet Yaxchilan tier 1 because it's cheap, but an opponent placed there first, the next worker has to go on tier 2 (more corn) or take a different gear. The whole game is jostling for cheap entry points and racing each other up the ripening curves. This is the core multiplayer tension — and it's the part that doesn't survive a solo translation cleanly (see §4).

Designer commentary. Public designer commentary from Luciani and Tascini is sparse in English. The best surfacing is in CGE's documentary series ("Making Board Games, Episode 4: Tzolk'in, Tash-Kalar, and Alchemists," 2023), where CGE notes that the Italian designers brought a working gear prototype directly to them — the gears were not added later as a theme but were the prototype's central conceit from the first pitch. CGE took the game in part because their stated identity is "extraordinary games that pushed the boundaries of creativity," and the gear mechanism was singular. Luciani has gone on to design Lorenzo il Magnifico, Grand Austria Hotel, and Marco Polo — all timing-sensitive Eurogames — suggesting a designer fingerprint of escalating-payoff mechanics.

Sources: Opinionated Gamers review, CGE documentary blog, BGA Tips, BGG strategy guide.

3. Existing digital attempts

Official Tzolk'in digital app: none. As of April 2026, Czech Games Edition has digital editions of Through the Ages (Steam, iOS, Android — well-regarded), Galaxy Trucker (iOS), and Codenames Online (browser), but no Tzolk'in port has been announced or released. There is no official iOS or Steam app, no announced development, no public roadmap entry. The most-trafficked digital version is on Board Game Arena (browser-based, asynchronous-friendly, faithful 2-4 player port). There are also fan-made Tabletop Simulator mods on Steam Workshop ("Tzolk'in Geared HD," etc.) which simulate the gear physics. Both are multiplayer reproductions, not single-player reframings.

Solo Tzolk'in. The base game has no official solo mode. Two community automa systems exist on BGG — the Cherey "virtual player" variant and the more sophisticated Pakal automa (BGG file 216225). Reviews of the solo experience are mixed: it plays in 45-60 minutes but the virtual player is "overpowered" and the absence of the blocking dynamic flattens the strategic space. This is the loudest signal in the research: Tzolk'in's mechanics survive solo, but its drama largely doesn't — confirming the translation problem in §4.

Worker-placement-with-time digital games. The most-cited cousin is Anachrony (Mindclash, 2017) — worker placement with literal time travel: you can borrow workers/resources from the future and must repay them later. Anachrony has a superb solo mode but no native mobile app (it's on Board Game Arena and Tabletopia only; a digital port has been on the wishlist for years). Anachrony's "time" is plot device, not procedural — you don't see workers move on a clock. Teotihuacan (Tascini's solo follow-up that uses dice-as-workers traveling around a track) is closer in spirit to Tzolk'in's "physical time" feeling, also on BGA, also no mobile app. Tekhenu uses a sundial. None of these are on iOS as native ports.

Gear-mechanic mobile games. Almost every mobile game with gears uses them decoratively (loading screens, settings menus, steampunk aesthetics) rather than structurally. The closest structural use is Gorogoa (rotating panels) and The Room series (mechanical puzzle boxes with gear puzzles), but both are puzzle-narrative games, not roguelites. Reigns uses cards on a binary clock, which is closer in spirit (commit-and-wait) than any gear-themed mobile game. Search reveals no precedent for "rotating gear as central game-state-clock" on mobile.

Verdict: Tzolk'in-the-IP is digitally unclaimed. The structural slot — "worker placement roguelite where time is a visible spinning thing" — is also empty on mobile. There is no incumbent to displace.

Sources: BGA Tzolk'in, Steam Workshop: Tzolk'in Geared HD, BGG Pakal automa, BGG Solo Variant thread, Anachrony review, Tribes & Prophecies expansion.

4. The translation problem

What gets stripped.

  • Player blocking is the loudest casualty. In multiplayer, you can't take a cheap tier-1 spot because someone else is there; the whole "where do I place?" calculus collapses without rivals. Mitigation: introduce AI pilgrims that occupy spaces on a schedule. Each run, 1-3 NPC tribes share the gears with you, placing workers on a deterministic-but-varied schedule (visible to the player, so the planning game survives). They aren't playing to win — they're obstacles, like Slay the Spire enemy intent telegraphs.
  • 4 Food Days as social pressure — in multiplayer, watching another player starve is half the fun. Solo, starving is just losing points. Mitigation: turn Food Days into roguelite combat encounters — a "Drought" or "Locust" or "Eclipse" event with variable pressure (instead of static "pay 2 corn per worker," you face e.g. "pay 3 corn per worker on Yaxchilan, 1 elsewhere" — a card you draw at run start so you plan around it).
  • The physical gear-spinning — replace with deeply satisfying haptic + animation. Lean into this; do not understate it.

What to lean into (the gears on a phone are an asset, not a liability). A 60Hz vector animation of five gears physically rotating, with workers as small icons that visibly migrate one tooth per round, is better than the cardboard original on the dimensions of clarity (you can zoom, slow-mo, see future positions) while preserving the haptic pleasure (gear-tooth-click haptic on every advance). Use SwiftUI's Canvas + withAnimation or SpriteKit for the gear rotations. Make the gear advance the cinematic moment of every round — camera pulls back, all five gears turn synchronously with a satisfying mechanical sound, workers slide into new positions, then camera returns to the player's hand.

What gets added (the digital meta-layer — pick 2-3 for v1).

  1. Roguelite calendar cycles. A "run" is one full revolution of the central gear (the base game's natural arc — 26-28 rounds, 4 Food Days, ends after one revolution). A "season" is 3-5 calendar cycles strung together with persistent meta. Each cycle is a ~25-40-minute session.
  2. Tribe drafts between cycles. The Tribes & Prophecies expansion already provides 13 asymmetric tribe powers (Bonampak, Itzamna, Kukulkan, etc.). Use them as the run's character class draft: at the start of each calendar cycle, choose 1 of 3 tribes, each modifying one gear's payoff curve (e.g., Bonampak: Palenque's tier-3 corn doubled; Itzamna: free Theology track every 5 rounds; Kukulkan: skull-gear payouts +1 VP). Direct mechanical fit, no invention required.
  3. Prophecy events as the "encounter" deck. Tribes & Prophecies prophecies ("at end of cycle, player with most stone scores 8 VP, fewest loses 3") become roguelite event cards — three drawn per cycle, visible from turn 1. This is the deck-building / risk-reward layer Slay the Spire uses to give each run identity.
  4. Persistent civilization unlocks. Between cycles, spend "obsidian" (post-run currency, scaled to performance) to unlock new tribes, new monument cards, new starting buildings, new prophecies — about 60-100 unlockable items for a 25-40 hour completion arc.
  5. Daily calendar. A fixed seed: same starting tribe, same prophecies, same Food Day modifiers. Global leaderboard by final VP. Streamer/social hook.

Recommendation: ship v1 with #1 + #2 + #4. Calendar cycles as runs is the spine; tribe-draft is the build identity layer; persistent unlocks are the retention engine. Prophecies (#3) are great content but can ship in a free update; daily (#5) is best added once leaderboards have a player base.

What cannot translate cleanly.

  • The smug satisfaction of taking the cheapest space first while opponents glare. No surrogate. The adversarial sting is gone.
  • The shared visible board state — opponents seeing what you're setting up, and adapting. NPC pilgrims can react to the player, but they'll feel scripted compared to humans.
  • Negotiated late-game temple races (where two players see they're tied on a temple and both throw resources at it). NPC pilgrims can compete for top-of-temple bonuses, but only as scripted antagonists, not as humans bluffing.

5. Concrete iOS prototype spec

Target: iOS portrait orientation, single-developer SwiftUI + SpriteKit (or Unity if comfortable), 5-7 month MVP. The gear animations want a real animation/physics layer — SpriteKit is probably the right fit on Apple's stack.

60-second core loop. App opens to your tribe's player board. You see five gears arrayed vertically (or in a 2D arrangement — see layout). You have 3 starting workers. Tap a gear, tap an action space (the lowest empty one is auto-suggested), pay corn shown in real-time. Or tap a worker on a gear to pick up — a popover shows the available actions (current tier + cheaper tiers below, with corn cost to walk down). Tap "End Turn." All five gears smoothly rotate one tooth (1.5-second animation, satisfying mechanical SFX, light haptic). NPC pilgrims take their turns (animated in 0.5-1.5s). Repeat. Every ~7 turns a Food Day banner appears, you pay corn, see the math, take any rewards. After ~26 turns, the cycle ends, scoring screen, meta-progression draft.

Single-screen layout (phone portrait, 6.1" reference). The 5 gears do not fit at native readable size in portrait; this is the central UX problem. Three options, ranked:

  • Option A (recommended): Vertical gear stack with focus mode. Top 60% is a vertical scrolling list of the 5 gears, each showing as a horizontal half-disc (only the visible action spaces — top of the gear). Default view shows 2 gears fully + edges of adjacent ones. Swipe vertically to scroll between gears. Tap any gear to enter "focus mode": that gear expands to fill 80% of screen for placement/pickup, others compress. Bottom 40% is your hand: workers, corn count, current round indicator, the central calendar gear (small, persistent, top-right) showing teeth-until-Food-Day.
  • Option B: Pinch-to-zoom 2D mandala. All 5 gears arrayed in their board-game configuration, fits at small scale; pinch to zoom. Closer to the physical board but harder to thumb-tap accurately on a phone.
  • Option C: Carousel with center focus. Like an Apple Music album view; one gear centered and large, others in peripheral vision. Good for legibility, sacrifices the "all five in one glance" feel that the board game gives.

Lead with A. Test B as a tablet/iPad mode.

Worker / gear interaction system.

  • Starter workers: 3 on hand, 0 on gears (mirror base game).
  • Placement UI: Tap an empty gear to show it expanded; the lowest-empty space pulses. Tap to confirm placement + pay corn. Multiple workers placeable in one turn (like base game) — each tap costs more (placement fee scales).
  • Pickup UI: Tap a worker on a gear; popover shows: "Take action at tier 4: Build a Monument (cost: 4 stone OR 4 gold OR 5 wood)" and below it "Or walk back: tier 3 for 1 corn / tier 2 for 2 corn / tier 1 for 3 corn." Single-tap to confirm.
  • Pickup chain: in one turn you may pick up multiple workers — sequential popovers handle each.
  • End Turn button: large, bottom-right, glowing when pressed. Triggers gear rotation animation.

Run structure.

  • A run = 1 calendar cycle = ~26 player turns (one full revolution of the central gear). Roughly 25-40 minutes. Mirrors base-game length.
  • A season = 3-5 cycles, with persistent state between (your worker pool, your tech track positions, some buildings) but resetting corn/resources. Final cycle is a "boss" cycle — a named historical event (e.g., "Rise of Calakmul") with a prophecy worth massive VP if achieved or massive loss if failed.
  • NPC pilgrims: 1-3 per cycle, scaling with difficulty. Pre-baked behavior trees (the "Trader" prefers Uxmal, the "Architect" prefers Tikal, the "Priest" rushes Chichen Itza). Visible upcoming move telegraphs (icon over their next gear) so the player can plan blocks.
  • Win condition: finish a season above a VP threshold. Loss conditions: starve >3 workers in any one Food Day; finish a cycle with negative net progress (no buildings, no tech, no skulls).

Meta-progression.

  • Between cycles in a season: draft 1 of 3 tribe-power cards (modifies one gear); draft 1 of 2 monument blueprints (added to your buildable pool).
  • Between seasons: spend obsidian on persistent unlocks — new tribes (~13 base from expansion + 10-15 new), new monuments, new prophecies, new visual themes for the temples, new player-tribe avatars. ~60-100 unlockables for a 25-40 hour completion arc. No power creep across runs — unlocks are sideways variety, not vertical stat increases. (Slay the Spire model.)

Monetization. Premium one-time, $6.99-$9.99. Avoid F2P energy mechanics — a run is a session and a session must be uninterruptible. A cosmetic-only DLC for tribe-art skins or alt board themes is plausible at $1.99-2.99 each. A future "Prophecies" expansion as paid DLC at $4.99 is the natural roadmap. Do not implement ads, do not gate cycles behind currency.

Aesthetic direction. Lean Maya-vivid stylized, not flat-modern, not photoreal. Reference: the actual Tzolk'in board art (saturated jungle greens, terracotta reds, jade, gold leaf) crossed with the iconographic clarity of Mini Metro or Monument Valley. The gears themselves should look mechanical-mythic: bronze with carved glyphs around the rim, not steampunk brass. Workers as small carved-stone idol icons. SFX: a deep wooden ratchet on gear advance, the click of obsidian on placement, low atmospheric jungle drone in the background. Avoid the trap of "Maya = brown and tan" — the original board is bright; the app should be brighter.

6. Risks and unknowns

Top design risks.

  1. Solo Tzolk'in may be fundamentally less interesting. The BGG community consensus is that the Pakal automa is okay but flattening — without competitive blocking, the placement decision becomes a pure optimization puzzle. Mitigation: validate this first (paper playthrough with the Pakal automa) before any code. If solo Tzolk'in genuinely doesn't sing, the prophecy-event/roguelite layer must do all the heavy lifting on its own, and that's a different game.
  2. 5 gears do not fit a phone screen comfortably. Real risk to the core loop. Mitigation: prototype Layout A early; if it doesn't work, fall back to a tablet-first design with iPhone as a secondary form factor. The game might genuinely be an iPad-first product.
  3. Gear rotation animation that's beautiful at turn 1 may feel slow at turn 26. Mitigation: progressive speedup (1.5s on turn 1, 0.5s by turn 10, 0.2s with toggle to "instant" by turn 20). Always allow tap-to-skip animation.
  4. The "wait longer = bigger reward" curve may feel too mathematical without a human to compete against. Single-player optimization puzzles can devolve into solved play. Mitigation: prophecy/event modifiers that change the curve every run; tribe powers that re-shape it; NPC pilgrims that punish naive optimization.
  5. Corn economy may collapse without blocking. In multiplayer, corn scarcity is partly enforced by opponents grabbing cheap corn-generating spaces. Solo, you may always have plenty of corn. Mitigation: tighten Food Day costs; introduce roguelite drought events; have NPC pilgrims compete for Palenque specifically.
  6. The Mini-Metro aesthetic risks erasing the Maya theme. Mitigation: hold onto the iconography (glyphs, idols, temple silhouettes) even as you simplify shapes.

What needs validation, in order.

  1. Solo tabletop playthrough with the Pakal automa. Play 3-5 full games solo before writing a line of code. Confirm the feeling of delayed gratification survives without opponents. If it dies on the table, it dies on the phone.
  2. Paper prototype of the iOS gear interaction. Sketch the Layout A vertical stack, simulate placement and pickup with sticky notes. Confirm the spatial reasoning works in portrait.
  3. Greybox digital prototype of one cycle. Just one gear (start with Yaxchilan, the most mechanically rich), one NPC, no meta-layer. Validate the gear-rotation animation feels great at 60Hz. Validate that "wait one more turn" has the same dopamine on a phone as it does on a board.
  4. Roguelite layer last. Tribes-and-prophecies meta only after the core loop is provably fun.

Three questions for the prototype builder.

  1. Are you willing to build your prototype around an iPad-first form factor? The 5-gear layout is genuinely cramped on iPhone. iPad-first lets you preserve the board-game's "all systems visible at once" aesthetic, which is half the appeal. You can ship a phone version, but it will be a compromise; choose your primary device early.
  2. How important is faithful adherence to the base game's economy? The corn/feeding system is balanced for the multiplayer blocking dynamic. To make solo work, you may need to adjust corn costs, Food Day intervals, or starvation penalties. Are you a purist (heretical to fans, painful to balance) or a modernizer (faithful to the feeling, willing to mutate the numbers)?
  3. What is the AI ambition? A scripted, telegraphed NPC pilgrim ("the Trader always plays Uxmal") is much cheaper to build and read than a real strategic AI. The Slay the Spire enemy-intent model proves scripted behavior can be deeply satisfying — but you must commit to legibility (the player must always see what NPCs will do next) rather than strength (NPCs that surprise are NPCs that frustrate solo players). Which model are you building?

7. References