Incremental War: Idle Cards
The card game War as an idle game: every hand you win pays points and every hand you lose costs you half of one, and you spend those points on a chance to level up the card you play — until every card you turn over becomes an ace and you stop losing hands entirely. No ads, no subscription, no energy timers.
What it does
- Solo War against a CPU using a complete, seeded, deterministic War rules engine: one standard 52-card deck split 26/26, simultaneous reveal, war on a tie with a face-down stake, recursive wars when the war cards also tie, and correct handling of a player running out of cards mid-war. Cards change hands under the real rules; the match ends when one side holds all 52
- Scoring: the winner of a hand takes both cards and is paid per card captured - 1 point for a card below jack, 2 points for a jack, queen, king or ace. Losing a hand always costs exactly half what winning it would have paid. That fraction is FIXED and is never upgradeable: the only way to lose less is to win more often, which funnels a new player straight to Level-Up Chance as their first purchase. Points floor at zero and can never go negative. Winning a match pays a separate, larger bonus
- The core mechanic is a proc, not a deck: when the player turns a card over, there is a Level-Up Chance that it jumps up by Level-Up Amount ranks before it is compared, capped at ace, shown as a green glow and a visible rank change on the card itself. Both sides always draw from the same standard deck, so cards changing hands never shifts the odds - which is exactly why this replaced an earlier deck-composition design, where winning diluted the player's own deck and flattened all progress
- Nine upgrades bought with points, all reset on prestige: Level-Up Chance (ten levels of 10%, to 100%), Level-Up Amount (+1 with the proc, then a rank a level to +12, which turns any card into an ace), Card Value, Face Bonus (multiplies what a jack and above pays, so it scales with Level-Up Amount), Match Bonus (taking all 52 cards pays ten hands' worth of the average pot, times this, so it stays worth winning at the endgame), War Multiplier (applied to the summed value of everything in the war pot), Deal Speed (1.2s to 0.05s per hand; the ceremonial deal between matches is four hands long and speeds up with it), Tables (1 to 8), and Auto-Play, which is bought per table: each level hires a dealer for one more table, priced at about half of its table, and a dealer deals at exactly Deal Speed - never slower than a thumb. The four income upgrades multiply (20-40% a level against a price that roughly triples); the two proc upgrades, Tables and Auto-Play are individually priced gates spread across the whole save. Every tuned number lives in Sources/Economy and is mirrored in pacing.py, which is the authority and asserts the pacing
- A single prestige layer, meaning one kind of reset - not one reset. The player prestiges as many times as they like, forever; there is no second button that later wipes Prestige Points, and prestige is never automated because deciding when to reset is the only real decision in the game. Prestiging wipes Points, all nine run upgrades, every dealer and tables back to 1, keeps every Prestige Point ever earned and everything bought with them, and pays out on a cube-root curve of the run's LIFETIME points (so spending during the run does not reduce the payout). Because a run twice as long pays only about 1.26x, short runs beat long ones - that asymmetry is the engine of the whole game. Prestige Points are a currency and nothing else: earning or holding them multiplies nothing, and everything a reset is worth comes from what they are spent on in the shop. The first reset is about seventeen minutes in and pays for a level or two. The shop has eight entries: five that make runs start further along and therefore cap out (Head Start, Dealers On Staff, Starting Points, Better Base Chance, Better Base Amount) and three uncapped (Point Multiplier, which is half as much again per level and is the engine of the loop; Faster Base Speed; Prestige Payout). The prestige screen shows the payout on offer, the player's current total beside it, one sentence (prestiging resets the run and grants Prestige Points to spend on permanent upgrades) and the shop - no rates, no averages, no lists of what is kept and lost, no preview of the next run. The first time prestige is offered the same sentence is shown once; players know what a prestige is and it must not be explained further
- The endgame is a finite, visible win condition and it is the END of the game, eight to twelve hours in. Reaching it takes the whole screen once - four aces dealt in, gold confetti, a heavy run of haptics and a line saying the game is won - and the upgrades screen states the goal and shows the hand win rate as its one headline figure: 100% Level-Up Chance at +12 Amount means every card the player turns over becomes an ace. Measured over 4,000 simulated matches that is 91.8% of hands won outright, 8.2% wars, and 0% of hands lost - the point of the game is reaching the state where losing is impossible. The hand win rate must climb visibly the whole way there (about 60% at half an hour, 70% at two hours, 85% at four) rather than sitting at 51% for hours. Match length falls from 176 hands to 20 across the same span, and match win rate climbs 50% to 100%
- One to eight tables running simultaneous independent matches, ALL of them on screen at once. One table gets the full stage with the war ceremony beat by beat; two to eight share the board as a grid of cells sized so every table is visible without scrolling, each showing its own two cards, the level-up glow, the result and what it earned, the deck split and whether it has a dealer. The board carries no per-minute rate, no run clock, no win streak and no "coming up" line: those were removed on purpose as clutter. Every cell is driven by a value the store publishes about ten times a second and holds one sampled hand long enough to read, so animation cost does not grow with deal speed. A table without a dealer is dealt by the one Deal button, which deals every such table at once
- A ceremonial dealing animation between matches lasting exactly as long as the shuffle interval (four hands' worth of Deal Speed), plus per-hand flip, resolve and collapse animations. Tables keep drawing cards at every deal speed: past the point where the eye can follow it, each samples one hand and holds it long enough to read, preferring a war when one landed in the window. Reduce Motion renders every table as a bar. Bar and cell rendering ticks at roughly ten frames per second rather than at display refresh rate, because this is a game people leave running for hours and pinning a 120Hz display would cook the battery
- There is deliberately no offline progress and no accrual while the app is closed: this is an active idle game, every run is played, and the prestige loop is tuned to resolve inside a single sitting
- Free players are capped, not walled. Level-Up Chance stops at 50%, Level-Up Amount at +6, Tables at 2, and Point Multiplier at level 12 and Faster Base Speed at level 10 - enough that a free player sees the level-up mechanic clearly working (about a 68% hand win rate) but never reaches every-card-an-ace. Everything else is uncapped and the game never stops - a free player keeps playing, prestiging and earning forever, they simply plateau. Every capped upgrade displays its free ceiling from the first time it is seen, never as a surprise on reaching it
- Four optional deck skins sold separately at $1.99 each as their own non-consumables - com.ryanallen.incrementalwar.skin.brass, .cobalt, .rose and .slate - each a complete colour system across all 52 cards plus the card back, re-authored from War Nearby's renderer for this app's palette. They are COSMETIC ONLY: a skin changes no number, no odds, no rate and no cap, and none of them is included in Pro. This is five non-consumables in total, which AppFactory's concept schema cannot express - it models exactly one - so the four skin products must be created in App Store Connect alongside the Pro product rather than by setup-iap alone
Available on the App Store
Support and legal