Chapter 35: Genre Analysis — How Design Principles Manifest Across Game Types

There is a question I ask every junior designer who pitches me a game: "What genre is it?" The answer tells me almost nothing about whether the game will be good. It tells me almost everything about whether the designer knows what they are making.

If they say "it's a 2D action-platformer with metroidvania exploration and souls-like combat in a Hollow Knight-shaped world," I know they have looked at the shelf and chosen a place. I might disagree with the place. I might think the genre is too crowded, the combat too punishing for the platforming, the exploration mismatched with the difficulty. But the designer can have the conversation. The designer knows the language.

If they say "it's hard to explain — it's kind of its own thing — there's no game like it," I know we are in trouble. Not because the game is bad. Maybe it is brilliant. But the designer has not done the work of locating it. They have not asked who plays games like this, what those players expect, what conventions they will bring to the controller. They are about to ship a game into a market that does not know how to receive it, and they will be confused when nobody buys it, and they will blame the marketing budget.

This chapter is about that locating work. Genre is the map of the medium. Designers who refuse to read the map get lost; designers who memorize the map but never leave the marked roads make derivative games; designers who know the map well enough to break it on purpose make the games we remember. We will walk every major genre, look at the design DNA each one carries, and study the cross-genre fusions and genre-breaking subversions that move the medium forward. By the end you will be able to answer the question. You will know what your game is. You will know who it is for. And — this matters more than students think it does — you will know how to talk about it.

Why Genre Matters for Design

Genre is a contract with the player. Every genre label triggers a set of expectations that the player carries to the controller before they have pressed a single button. Soulslike means: stamina-based combat, punishing enemies, bonfire-equivalent checkpoints, currency-on-death, obscure storytelling. Roguelite means: short runs, permadeath, persistent meta-progression, randomized content. Cozy farming sim means: slow tempo, low stakes, seasonal cycles, social bonds with NPCs. Platform fighter means: percentage damage, ring-out victory, four-player chaos, easy controls, hard mastery.

The contract is implicit but real. When players load Hades they expect to die a lot, get a little stronger between runs, and watch the story unfurl across dozens of attempts. They are not surprised by these things. They are not annoyed by them. They came for them. The genre advertised the contract; the player accepted it; the designer's job is to honor it while delivering specific delights.

This is what genre gives you for free: it pre-loads the player's expectations and their tolerance for friction. Dark Souls fans tolerate boss-room fog walls because they are part of the deal. They tolerate cryptic NPC questlines because they came for cryptic. A Stardew Valley player who picked up Dark Souls by accident would be furious at the same features. Same features, different audience, different verdict. The genre does the audience-filtering for you.

Violating the contract is a design choice. Sometimes the violation is the entire point — Undertale offers the JRPG framework and then asks the player to refuse to fight, exploding the central convention of its genre. Spec Ops: The Line dresses itself as a generic third-person military shooter and then weaponizes the genre's clichés against the player. Inscryption opens as a card game and then breaks the fourth wall hard enough that "card game" stops being an accurate description by the second act. These violations work because the designers understood the contract before they tore it up. They earned the betrayal.

But violating the contract by accident is a different thing. A platformer where the controls are mushy because the designer did not study Mario. A shooter where the time-to-kill makes no sense because the designer did not play CS:GO or Halo and learn what each TTK buys you. An RPG where character builds do not actually let you build characters because the designer reached for the genre label without internalizing the genre's promises. These are not subversions. These are mistakes dressed in genre clothing.

💡 Intuition: Genre is the default contract. You can rewrite the contract, but you have to know what you are rewriting. The penalty for ignorance is not "innovation"; it is a confused player. Confused players quit.

Conventions also give designers and players a shared language. When I say "this is a dash mechanic, like Celeste's," every reader who has played Celeste nods. When I say "this boss is a Soulsborne-style endurance test with a punish-window after a five-hit combo," the audience knows what I mean without thirty paragraphs of explanation. Conventions are vocabulary. Vocabulary lets you say things compactly. The designer who insists every mechanic be invented from scratch is the writer who refuses to use any word coined after 1900: theoretically purer, practically useless.

The Problem With Genre Labels

Now the heretical part. Genre labels are mostly garbage.

They are imprecise, retroactive, marketing-driven, and frequently invented by the press long after the games they describe were made. "MOBA" — Multiplayer Online Battle Arena — was coined by Riot Games' marketing team to describe League of Legends in 2009, after DoTA (a Warcraft III mod) had already existed for half a decade and after Demigod and Heroes of Newerth were already shipping. The genre label arrived after the genre had been established by a community that did not need it.

"Metroidvania" is two words taped together: Metroid (1986) and Castlevania: Symphony of the Night (1997). Two games released eleven years apart, in different countries, by different studios, with different control schemes and different aesthetics. The label binds them by one shared trait — interconnected map with backtracking gated by abilities — and then is applied to every 2D exploration game that even approximately fits, retroactively swallowing Hollow Knight, Ori, Axiom Verge, Bloodstained, and arguably half the indie scene of the 2010s.

"Souls-like" is the worst offender. The label originated to describe games that explicitly imitated Demon's Souls and Dark Souls — The Surge, Lords of the Fallen, Nioh, Mortal Shell, Lies of P. But it metastasized to describe any 3D action game with a stamina bar and a checkpoint system. Star Wars Jedi: Fallen Order is "souls-like" because it has a rest spot that respawns enemies. God of War (2018) gets called "souls-influenced" because its combat is methodical. By this standard, The Legend of Zelda: A Link to the Past is souls-like because you can die and respawn. The label has eaten itself.

This matters for two reasons. First, when you are a designer thinking about your own game's genre, you cannot trust the labels to mean what you think they mean. "Roguelike" famously broke into "roguelike" (real, classical, NetHack-shaped) versus "roguelite" (modern, run-based, often action-oriented), and the schism was so heated that the community wrote the Berlin Interpretation in 2008 to nail down what "roguelike" actually meant — a manifesto that has since been ignored by the marketing of every game claiming the label.

Second, when you are positioning your game commercially, the label is doing real work whether or not it is precise. The Steam tag "metroidvania" sells games. The Steam tag "souls-like" sells games. The Steam tag "deck builder" sells games. These are not genres in any rigorous sense. They are search terms — heuristics players use to find games shaped like the games they already love. You cannot ignore these tags, because your customers will not find your game without them. You also cannot let them dictate design, because copying a genre to chase its tag is how you end up the seventeenth Metroidvania this month with no audience left.

🔵 Practitioner Note: Treat genre labels as descriptive, not prescriptive. They describe what exists; they do not tell you what to make. Use them as tools for finding your audience and for reasoning about player expectations, not as commandments about how your game must work.

Major Genres and Their Design DNA

The remainder of this chapter walks the major genres of the medium. For each, I will sketch the core loop (what the player does moment to moment), the progression model (how the experience grows over time), the feedback cadence (how often the game tells the player they are doing well), the accessibility demands (what skills the genre asks of new players), the typical session length, the audience expectations, and at least one design innovation that reshaped the genre. This is the map. Internalize it.

Platformer

Core loop: Move, jump, time the jump, land. Repeat with increasing precision and complexity.

The platformer is the medium's foundational genre. Donkey Kong (1981) established it; Super Mario Bros. (1985) perfected it; Sonic the Hedgehog (1991) accelerated it; Super Mario 64 (1996) translated it to 3D; Celeste (2018) made it about feeling. Every platformer is a study of one verb — jump — and what you can build around it.

Platformers split into three rough variants. Precision platformers (Celeste, Super Meat Boy, The End is Nigh, Hollow Knight's Path of Pain) test the player's mastery of input timing under penalty of death. The session is short loops of "die, retry, die, retry, succeed." The progression is internal — your fingers learn. Flow platformers (Sonic Mania, Rayman Legends, Pizza Tower) reward momentum and rhythm. The pleasure is the run, the perfect linkage of jumps, the chain of speed. Death exists but is incidental. Exploration platformers (Super Mario 64, Banjo-Kazooie, Astro Bot, A Hat in Time) build sandboxes you traverse at your own pace, collecting things, finding secrets. Movement is a verb you spend leisurely.

Celeste is in our anchor set because it is a clinic in precision platformer design. The dash mechanic is one button, regenerates on landing, and is the entire game — every chapter explores a new ramification of that single verb. The B-Side and C-Side variants of each level are progression for the same verb: harder levels, same vocabulary. The accessibility demands of precision platforming are real — twitch reflexes, pattern recognition, resilience to repeated failure — and Celeste's assist mode lets players reduce those demands without the game judging them. That is craft.

The 3D platformer survives in 2026 mostly through Nintendo (Mario Odyssey, Mario Wonder — though Wonder is 2D) and the Sony PlayStation Studios mascot revival (Astro Bot won Game of the Year 2024, an event nobody predicted in 2010). The genre is small but vital. When Astro Bot succeeded, half the industry started asking whether the 3D platformer could be commercially viable again; the other half realized it had never stopped being viable, just under-served.

Innovation that reshaped the genre: Celeste's accessibility-without-condescension. The game treats the player as an adult who can choose how hard their experience should be, removing the gatekeeping that precision platformers had unconsciously practiced for decades.

Action / Action-Adventure

Core loop: Explore, fight, solve light puzzles, gain items that unlock new areas, fight bigger.

The action-adventure genre descends from The Legend of Zelda (1986) and has dominated AAA development for two console generations. Zelda gave us the template — overworld with hidden things, dungeons with themed obstacles, a tool gained per dungeon that recontextualizes earlier areas. Ocarina of Time (1998) ported the template to 3D and codified the modern action-adventure: a navigable open space, a combat system with multiple verbs (attack, block, item use), a progression of items that gate areas.

The lineage runs through Tomb Raider, Devil May Cry, God of War, Uncharted, The Witcher 3, Assassin's Creed, Horizon Zero Dawn, and culminates in The Legend of Zelda: Breath of the Wild (2017), which broke the lineage open by removing the dungeon-progression structure and replacing it with systemic emergence. Tears of the Kingdom (2023) extended the experiment further, layering an absurd construction system on top.

Breath of the Wild is in our anchor set because it represents a genre redefining its own DNA in real time. The traditional action-adventure model said: hand-author the experience, gate it carefully, dispense progression in measured doses. BotW said: build systems that interact, populate the world with affordances, trust the player to find their own path. Both approaches still work. God of War: Ragnarök shipped a year before Tears of the Kingdom using the traditional approach and was rapturously received. Genre evolution does not mean genre replacement.

Innovation that reshaped the genre: BotW's replacement of authored progression with systemic emergence. The genre has been arguing about it ever since.

RPG / JRPG / WRPG

Core loop: Make a character, fight enemies, gain experience, level up, customize, fight bigger enemies, advance the story.

The RPG is the largest and most internally fractured genre in games. It branches roughly into Western-style (WRPG) — Baldur's Gate, The Elder Scrolls, Fallout, The Witcher, Pillars of Eternity — and Japanese-style (JRPG) — Final Fantasy, Dragon Quest, Persona, Octopath Traveler. The line between them is not as clean as it used to be (Square Enix's Final Fantasy XVI is essentially an action game; Larian's Baldur's Gate 3 uses turn-based mechanics that feel more JRPG than WRPG), but the cultural divide remains.

Turn-based RPGs (Persona 5, Baldur's Gate 3, Octopath Traveler, Pokémon) move at the speed of thought. You queue actions, the game resolves them, you observe the result. The session is medium-length (90 minutes is comfortable) and the experience asks for tactical patience. Baldur's Gate 3 won Game of the Year 2023 and proved that mainstream audiences will absolutely play turn-based combat in 2023, contradicting a decade of industry wisdom.

Action-RPGs (Dark Souls, Elden Ring, Diablo, Path of Exile, Monster Hunter) put the RPG progression layer underneath real-time combat. Your stats matter; your reflexes matter equally. Dark Souls — our anchor — is the genre's purest expression: the stats are real, the build matters, and yet the moment-to-moment skill of dodging and timing is the game.

Stat-less narrative RPGs (Disco Elysium, Citizen Sleeper, Pentiment) keep the RPG's character-creation and dialogue-tree DNA but throw away or radically reshape combat. Disco Elysium is technically a CRPG that runs on dice rolls and skill checks, but it has no combat in the traditional sense — its "combat" is reading a thermometer and trying not to have a nervous breakdown.

The cross-pollination is heavy. Hades is an action-roguelite with RPG character relationships; Stardew Valley has RPG combat in the mines; Dark Souls is an action-RPG with roguelite-flavored death loops; Mass Effect is a third-person shooter with WRPG dialogue trees. The RPG genre is less a discrete shape than a suite of mechanics other genres can borrow.

Innovation that reshaped the genre: Dark Souls and Elden Ring's "difficulty as core feature" — the assertion that the RPG's promise of growth means more when growth is hard-won. Inverted by Disco Elysium's "RPG without combat is still an RPG."

Shooter

Core loop: Aim, shoot, kill, reposition. Vary the targets, the weapons, the cover.

The shooter is the medium's most commercially dominant genre — the AAA cash machine of the last twenty years. It is also one of the most internally fractured.

Military FPS (Call of Duty, Battlefield, Six Days in Fallujah) sells the fantasy of being a modern soldier. The verbs are aim-down-sight, sprint, take cover, reload. The progression is unlock-based — kill to earn weapons and attachments. Session length is short (10-minute Team Deathmatch), the cadence is fast, and the audience is huge. Call of Duty: Modern Warfare (2019) defined the contemporary shape of the genre.

Arena FPS (Quake, Unreal Tournament, DOOM 2016, DOOM Eternal) is the older, faster, more punk-rock branch. Movement is the game. DOOM 2016 and Eternal revived a genre many believed dead by re-centering the loop on aggressive movement, rapid weapon-swapping, and "glory kills" as a resource system. The genre's audience is small but rabid.

Hero shooters (Overwatch, Apex Legends, Marvel Rivals) put a roster of asymmetric characters into team-based competitive play. The genre exploded in the late 2010s, contracted brutally in the early 2020s (RIP Concord, Battleborn, LawBreakers), and stabilized around a handful of dominant titles. The barrier to entry is high — you have to launch with a polished, tested roster of 20+ heroes — and the LiveOps cost is enormous.

Extraction shooters (Escape from Tarkov, Hunt: Showdown, Marathon (2025)) introduced a brutal twist: bring in your loadout, find loot, escape with it, or die and lose everything. The session is medium-length (30-60 minutes), the tension is constant, and the audience is small but obsessed. The genre is hard to design for because the punishment for losing is severe — a casual player will bounce immediately.

Tactical shooters (Counter-Strike, Valorant, Rainbow Six Siege) trade the speed of arena shooters for slow, deliberate, high-information play. Time-to-kill is brutal, economy is real (CS:GO's buy menu is half the game), and rounds reset state. Precision matters more than reflex.

Battle royales (PUBG, Fortnite, Apex Legends, Warzone) compress 100 players into a shrinking map and let one survive. The genre, invented effectively by the H1Z1 mod community in 2017, was the dominant force in shooter design for 5+ years and is now in mature decline as the audience consolidates.

Innovation that reshaped the genre: PUBG's 100-player elimination loop, which created an entire sub-genre and reshaped multiplayer monetization for half a decade.

Strategy

Core loop: Plan, build, harvest, deploy, react, win the long game.

Strategy games ask the player to think in time horizons longer than any other genre.

RTS (StarCraft, Age of Empires, Command & Conquer, Age of Mythology: Retold) compresses strategic planning into real-time pressure. APM (actions per minute) is a real skill ceiling; StarCraft II pros hit 300+ APM. The genre commercially declined in the 2010s as MOBAs ate its competitive niche, but the 2020s have brought a resurgence — Age of Empires II: Definitive Edition and the Microsoft revival of the franchise drew real audiences.

4X (Civilization, Stellaris, Endless Legend, Old World) — eXplore, eXpand, eXploit, eXterminate — gives you turn-based macro strategy across decades or millennia of in-game time. Civilization VI has sold over 12 million copies; Civilization VII shipped in 2025. Sessions are brutal — "one more turn" is the genre's signature drug, and 8-hour sessions are common.

Grand strategy (Crusader Kings III, Europa Universalis IV, Hearts of Iron IV, Victoria 3) is Paradox Interactive's house specialty: real-time-with-pause simulation of historical political and military systems at extreme depth. The audience is small, hyper-engaged, and obsessed. The accessibility demands are absurd.

Turn-based tactical (XCOM 2, Into the Breach, Fire Emblem: Three Houses, Tactical Breach Wizards) compresses strategy onto a small grid and one squad. Sessions are short (a single mission is 20-40 minutes) but the campaign is long. The genre is one of the strongest indie spaces — Into the Breach is a 4-person team's masterpiece.

Tower defense (Plants vs. Zombies, Kingdom Rush, Bloons TD 6, Mindustry) is strategy distilled to its essence: place units, watch them work, optimize the layout. Mobile-friendly, indie-friendly, and weirdly stable as a genre.

Innovation that reshaped the genre: Into the Breach's "perfect information" twist — the game shows you exactly what enemies will do next turn, eliminating randomness and turning every encounter into a chess puzzle.

Simulation

Core loop: Maintain a system, optimize a system, watch a system thrive.

Simulation games turn the player into the operator of a virtual world.

Life sim (The Sims, Animal Crossing, inZOI (2025)) lets you run a household or a town. The progression is open-ended — there is no "win." The audience skews older and broader than most genres; The Sims 4 is one of EA's most reliably profitable franchises a decade after launch. Animal Crossing: New Horizons sold 47 million copies because it released into a global lockdown and met an emotional need nobody had named.

Farming sim (Stardew Valley, Harvest Moon / Story of Seasons, Coral Island, Fields of Mistria) is a sub-genre that should not have survived past 2000 and instead has become a juggernaut. Stardew Valley sold over 30 million copies, was made by one person (Eric Barone), and revived the genre singlehandedly. The cozy aesthetic, the seasonal rhythm, the slow accretion of progress — this is comfort food in software form.

Factory sim (Factorio, Satisfactory, Dyson Sphere Program, Shapez 2) gives you the engineer's fantasy of building production lines. The session length is long; the genre attracts the kind of player who builds spreadsheets to track their factories. Highly compatible with the autistic-coded engineering brain. Factorio has been in active development since 2016 and never goes out of style.

Vehicle sim (Microsoft Flight Simulator, Euro Truck Simulator 2, iRacing, Train Sim World) is a niche the mainstream forgets exists, but those niches sustain decades-long franchises. Microsoft Flight Simulator (2020) recreated the entire planet at photorealistic fidelity and has sold across millions of seats.

Social sim (Tomodachi Life, Persona 5 school-life systems, Eternights) is hard to monetize but has cult appeal.

Innovation that reshaped the genre: Factorio's "automate everything" loop, which spawned a genre of game-as-spreadsheet that AAA still does not understand.

Puzzle

Core loop: Encounter a problem, manipulate the system, find the solution, see the next harder problem.

Puzzle games are the medium's most direct expression of the joy of figuring things out.

Pure puzzle (Tetris, The Witness, Baba Is You, Stephen's Sausage Roll) presents systems and lets you reason your way through them. The Witness is famously a 100+ hour game with no text tutorial, just escalating panels that teach you their own rules. Baba Is You is a puzzle game where the rules of the puzzle are themselves manipulable objects in the puzzle. Tetris — released in 1984 — remains in 2026 the most-played video game in history.

Puzzle-platformer (Portal, Braid, Limbo, Inside, FEZ) marries puzzle logic to traversal. Portal is the canonical example: a single mechanic (place two portals, things travel between them) extrapolated into 19 puzzles of escalating depth. Once you understand portal physics, you understand the entire game's grammar.

Match-3 (Candy Crush, Bejeweled, Puzzle & Dragons) is the most commercially dominant puzzle subgenre and the entry-level mobile genre. Almost nobody outside of mobile-game studios takes them seriously, and yet they make billions. The design is deceptively sophisticated — Candy Crush's level-design pipeline has produced 14,000+ levels and counting.

Innovation that reshaped the genre: Baba Is You's "rules are objects" insight — the realization that a puzzle game can let you manipulate its own grammar.

Roguelike / Roguelite

Core loop: Start a run, gain power, die, lose most progress, start again with permanent meta-upgrades, push further.

The roguelike was once a tiny genre of ASCII dungeon crawlers; in the 2010s it ate the indie scene.

Classical roguelikes (NetHack, Dungeon Crawl Stone Soup, Caves of Qud, Cogmind) honor the Berlin Interpretation: turn-based, grid-based, permadeath, randomized everything, ASCII or near-ASCII graphics. The audience is small, the depth is bottomless. Caves of Qud exited Early Access in 2024 after 17 years of development and is a masterpiece almost no one has finished.

Modern roguelites (Hades, Dead Cells, Slay the Spire, Balatro, Returnal, Hades II) loosened the rules — real-time combat became fine, persistent progression became fine, run-based replay became the loop. Hades (2020) showed that you could marry roguelite structure to character-driven narrative; Slay the Spire (2019) crystallized the "deck-builder roguelite" subgenre that has since spawned Monster Train, Inscryption, Wildfrost, Astrea, and dozens more; Balatro (2024) became the surprise hit of its year by combining poker with deck-building roguelite structure.

The "deck-builder" sub-genre explosion is one of the most important indie phenomena of the last decade. Slay the Spire was a single template — three acts, build a deck, fight bosses — and it spawned an entire commercial category. The lesson: when a genre's core loop is procedurally regenerative, every iteration of the loop is fresh content for the player. You build the system once and the system generates the levels forever.

Innovation that reshaped the genre: Slay the Spire's deck-builder fusion, which proved that "roguelite" was less a genre than a structural template that could be applied to almost any other genre.

Sandbox / Immersive Sim

Core loop: Explore a systemic space, find your own goals, use systems against each other, make stories.

The immersive sim is one of the medium's noblest genres and one of its hardest to commercialize.

Immersive sims (Deus Ex, Thief, System Shock 2, Dishonored, Prey (2017)) build environments where every problem has multiple solutions and every system interacts with every other system. The Looking Glass / Ion Storm lineage that birthed the genre is mostly dead; Arkane Studios kept the flame alive with Dishonored and Prey and was then absorbed into the AAA grinder. Prey (2017) sold poorly. Dishonored 2 sold poorly. Deathloop sold okay. The genre is critically beloved, commercially marginal, and indie revivalists like Cruelty Squad and Peripeteia are keeping the form alive on small budgets.

Sandbox (Minecraft, Terraria, Valheim, Core Keeper) is the immersive sim's mass-market cousin — open systems, player-set goals, but without the immersive sim's pretensions to authored narrative. Minecraft is the best-selling video game of all time at 300+ million copies. Terraria has sold over 60 million. Valheim did 5 million in its first month of Early Access in 2021.

Colony sims (Dwarf Fortress, RimWorld, Songs of Syx, Going Medieval) put you in charge of a tiny civilization and let the systems generate stories. Dwarf Fortress is the genre's mythological progenitor — Tarn Adams has been writing it since 2002, and the Steam release in 2022 finally gave it a graphical interface. The genre's signature pleasure is the emergent disaster — your fortress falls because of an unanticipated chain of system interactions, and the failure is interesting in a way no scripted failure could be.

Innovation that reshaped the genre: Minecraft's "creative mode + survival mode" duality, which let one game serve both expressive and challenge-driven players.

Narrative / Adventure

Core loop: Walk, look, talk, choose, watch the story unfold.

The narrative-adventure genre is the medium's oldest and one of its most diverse.

Point-and-click adventures (The Secret of Monkey Island, Grim Fandango, Thimbleweed Park, Return to Monkey Island) had a 1990s heyday, a 2000s death, and a 2010s indie revival. The genre is now a small-but-stable corner of the market, with developers like Ron Gilbert, Tim Schafer, and the Daedalic team keeping the lights on.

Walking sims (Dear Esther, Gone Home, Firewatch, What Remains of Edith Finch, The Vanishing of Ethan Carter) are minimal-mechanic narrative experiences where exploration and observation are the main verbs. The genre was mocked when it emerged ("they're not even games") and is now respected as an essential narrative form. What Remains of Edith Finch won every narrative award in 2017.

Visual novels (Ace Attorney, Steins;Gate, Doki Doki Literature Club, 13 Sentinels: Aegis Rim) are dominant in Japan and growing globally. The genre's interactivity is minimal — read text, occasionally choose — but the depth of writing is often higher than other genres allow. Doki Doki Literature Club used the visual-novel framework to commit one of the medium's most famous genre subversions.

Innovation that reshaped the genre: What Remains of Edith Finch's "every chapter is a different mechanic" approach — using gameplay shifts as a storytelling tool rather than a consistency.

Fighting

Core loop: Pick a character, learn their toolkit, condition the opponent, win the round.

Fighting games are the most mechanically dense genre in the medium.

Traditional 2D fighters (Street Fighter 6, Tekken 8, Mortal Kombat 1, Killer Instinct) operate on inputs measured in frames (1/60th of a second). The skill ceiling is monstrous. The community is global, competitive, and stable. Street Fighter 6 (2023) introduced "Modern Controls" that lowered the input barrier and brought new players into the genre for the first time in years.

Platform fighters (Super Smash Bros., Multiversus, Rivals of Aether 2, Nickelodeon All-Star Brawl) use percentage-and-knockback instead of health bars, support 4+ players in chaotic matches, and lower the input barrier (no quarter-circle motions). Smash is by far the largest fighting-game community on Earth.

Anime fighters (Dragon Ball FighterZ, Guilty Gear Strive, Granblue Fantasy Versus Rising, Under Night In-Birth II) emphasize aerial combat, long combo chains, and visual flash. Guilty Gear Strive (2021) was Arc System Works' attempt to lower the genre's barrier without diluting it; it largely succeeded.

Innovation that reshaped the genre: Street Fighter 6's Modern Controls, which decoupled "knowing the moves" from "executing the inputs" — a heresy that worked.

Rhythm / Music

Core loop: Hit the beat, miss the beat, feel the music.

Rhythm games make music itself the gameplay.

The genre's commercial peak was Guitar Hero's 2007-2009 explosion; the post-bust years scattered the genre across mobile (BTS World, Cytus, Arcaea), indie (Crypt of the NecroDancer, Patapon revivals), VR (Beat Saber, the best-selling VR game of all time), and bizarre experimental works (Thumper, Sayonara Wild Hearts). Hi-Fi Rush (2023) revived the action-rhythm hybrid and was then immediately killed when Microsoft shut down Tango Gameworks — a cautionary tale about how even successful niche genres are commercially fragile.

Innovation that reshaped the genre: Crypt of the NecroDancer's "rhythm + roguelike" hybrid, which proved the rhythm mechanic could be a layer rather than the whole game.

Sports / Racing

Core loop: Compete, lose, learn, compete again.

Sports and racing split into arcade and simulation halves.

Arcade racers / sports (Forza Horizon, Rocket League, Mario Kart 8, NBA Street) prioritize fun and accessibility. Rocket League — soccer with rocket-powered cars — is one of the most successful original genre fusions of the last fifteen years.

Simulation racers / sports (F1 24, iRacing, Gran Turismo 7, FIFA / EA Sports FC, NBA 2K) reward technique and licensing. The annual sports-game release schedule (FIFA / NBA 2K / Madden / NHL) is its own quasi-genre — same engine, new roster, same buyers, year after year. The genre's monetization model relies heavily on Ultimate Team modes that Chapter 33 covered as ethically dubious.

Innovation that reshaped the genre: Rocket League's genre fusion (soccer + driving + acrobatics) opened a vein of "sports + something weird" experiments that has yet to be fully explored.

MOBA

Core loop: Pick a hero, lane, farm, fight, push, win the team game.

The MOBA — Dota 2, League of Legends, Smite, Heroes of the Storm (RIP) — has the strictest genre conventions of any modern major genre. Three lanes. A jungle. Towers. A nexus. Five-vs-five. Heroes with four abilities. Items bought with gold from minion farming. The conventions are nearly inviolable; Heroes of the Storm tried to break some of them and was punished commercially.

The genre is dominated by two titans (LoL and Dota 2) and a long tail of survivors. Mobile Legends: Bang Bang alone outpaces both Western titans by player count, primarily in Southeast Asia. The MOBA is the world's largest competitive esport.

Innovation that reshaped the genre: Riot Games' League of Legends business model — free-to-play with cosmetic monetization — was invented for the MOBA before it spread to every other multiplayer genre.

Horror

Core loop: Explore, fear, manage scarce resources, survive.

Horror games split by what kind of fear they evoke.

Survival horror (Resident Evil 4 Remake, Silent Hill 2 Remake (2024), Alien: Isolation, The Evil Within, Signalis) puts you in a dangerous space with limited resources. The genre had a near-death period in the 2000s and has come roaring back with the Capcom Resident Evil remake program (RE2, RE3, RE4) — each of which has sold huge numbers and proved that classical survival horror sells when made well.

Psychological horror (Alan Wake 2, Layers of Fear, SOMA, Visage) prioritizes atmosphere and unease over jump scares. The genre is small but produces some of the medium's most respected work.

Indie horror shorts (Iron Lung, Mouthwashing, World of Horror, the entire Puppet Combo catalog, the Amanita Design adventures that lean horror) — typically 1-3 hour experiences from solo developers or tiny teams, distributed on Steam and itch.io for $5-$15. This subgenre is one of the healthiest niches in indie development. Iron Lung — a single-developer game about piloting a tiny submarine in a sea of blood — became a viral hit and a YouTuber-favorite genre staple.

Innovation that reshaped the genre: Alien: Isolation's adaptive Xenomorph AI, which created emergent horror that scripted scares could not match.

Card Game / TCG / Deck-Builder

Core loop: Build a deck, play cards, react to the opponent (or the procedurally generated dungeon), win.

Card games split into two main subgenres.

Trading card games (Magic: The Gathering Arena, Hearthstone, Marvel Snap, Yu-Gi-Oh! Master Duel, Pokémon TCG Live) are competitive multiplayer with deck construction and card-rarity economies. Hearthstone defined the modern digital TCG; Marvel Snap (2022) reinvented the format with three-minute matches and zone-based mechanics; Magic Arena digitized a 30-year-old paper game with surprising fidelity.

Single-player deck-builders (Slay the Spire, Monster Train, Inscryption, Wildfrost, Balatro) are roguelite cousins covered above. The two genres share DNA but have different audience expectations — the TCG audience cares about competitive ladder; the deck-builder audience cares about run variety and unlockables.

Innovation that reshaped the genre: Marvel Snap's three-minute match length, which finally cracked the question of how to make a TCG that fits in a phone-screen attention span without feeling shallow.

Cross-Genre Design

The most interesting games of the last decade have been cross-genre fusions.

Stardew Valley is a farming sim + RPG + social sim + light combat game + fishing minigame + brewing minigame. Strip any one of those layers and it would be a worse game. The fusion is the design.

Dark Souls is an action-RPG + roguelike feedback loop (death-as-learning) + immersive sim (interconnected world) + horror (the dread of returning to lost souls). FromSoft did not invent any of these elements; they invented the combination.

Hades is a roguelite + character-driven narrative RPG + action game with three distinct combat systems. The narrative was the real innovation — most roguelites had treated story as an afterthought because run-based structures fight against narrative arcs. Hades found that if every run advances the story by a small amount, the structure becomes a feature: every death is a story beat.

Vampire Survivors (we will study it as a case study) is Castlevania aesthetic + bullet-hell mechanics + roguelite structure + idle-game progression. None of those ingredients were new. The combination was a genre.

Deep Rock Galactic is a co-op shooter + dungeon crawler + procedurally-generated mining sim + class-based RPG. The dwarves yell "Rock and Stone!" and the genre fusion is so distinctive that no clear competitor exists.

💡 Intuition: Fusion is one of the cheapest forms of innovation. You do not have to invent a new mechanic; you have to combine existing mechanics in a way nobody has tried. The risk is lower than pure invention, and the upside — a genuinely new design space — can be enormous.

The trick with fusion is coherence. Two genres glued together do not automatically make a good game; they often make a confused one. Stardew Valley works because every system serves the same emotional tone (cozy progression). Dark Souls works because every system serves the same emotional tone (earned mastery in a hostile world). When the systems' emotional registers conflict, fusion fails.

Genre Expectations and Breaking Them

Subverting genre conventions is one of the most powerful tools in design. It is also one of the most dangerous.

Undertale (2015) presented itself as a JRPG. Random battles, bullet-hell-style action commands, monsters with HP bars, an XP/LV progression system. And then it asked: what if you do not have to fight? Every monster could be spared. Killing them was a choice, and the choice changed the world. The "Pacifist" run subverted the genre's fundamental loop. The "Genocide" run weaponized the player's expectations of completion. Undertale is one of the most influential RPGs of the last decade because it took the genre seriously enough to know what to break.

Spec Ops: The Line (2012) presented itself as a generic third-person military shooter. Cover-based combat, regenerating health, two-weapon loadout, Middle Eastern setting. And then it spent its entire runtime deconstructing the war-shooter genre — making the player commit increasingly horrific acts and refusing to let them feel heroic about it. The genre's clichés became the game's themes. The white phosphorus scene is one of the most cited moments in game writing.

Inscryption (2021) presented itself as a creepy single-player deck-building roguelite — until it suddenly was not. The act break that transforms the entire genre of the game is one of the most audacious design choices of recent years. (We will not spoil it more than that.) The subversion works because Act 1 is genuinely a great deck-builder roguelite — Daniel Mullins earned the right to break the genre by mastering it first.

The pattern is the same across all three: the subversion only works if the genre is honored first. Undertale is a real RPG. Spec Ops is a real cover shooter. Inscryption Act 1 is a real deck-builder. Then, having delivered the genre's expected pleasures, the designer pivots — and the pivot lands because the audience trusted the contract before it was broken.

🔴 Common Pitfall: Subverting a genre you do not understand produces a confused mess, not a subversion. If you want to break a convention, master it first. The first hour of Undertale could be the first hour of any earnest JRPG, and that is not an accident.

Genre as Market Signal

We have to talk about commerce, because Steam tags pay rent.

When you list a game on Steam, you choose tags. Players browse by tags. Wishlist conversion correlates with how well your game's tag profile matches the searches of players who would buy it. "Roguelite," "Metroidvania," "Deck Builder," "Cozy," "Soulslike," "Survival," "City Builder" — these tags carry audiences. A game tagged "Metroidvania" enters a search funnel of users actively looking for metroidvanias. A game with no clear genre identity enters no funnel at all.

This creates a tension. The honest designer wants to make the game that is in their head. The commercial designer wants to make a game whose tags will find buyers. The successful designer holds both tensions: they pick a primary genre that will find an audience, then they make the specific game they wanted to make within that genre.

The "Soulslike" SEO problem is the cautionary tale. Once "Soulslike" became a high-converting Steam tag, every action game with a stamina bar applied for the label. The market got flooded. Buyers became cynical. Pure imitations failed; only games that brought something genuinely new to the form (e.g., Lies of P's pastiche of Pinocchio, Wo Long's Chinese mythology, Black Myth: Wukong's production values) cut through. The lesson: genre tags work when they are honest signals, not when they are SEO bait.

Choosing your primary genre is a business decision as much as a creative one. Before you commit, ask:

• Who plays games in this genre?
• How crowded is the genre right now?
• What recent successes are stealing oxygen?
• What recent failures suggest audience fatigue?
• Is my game distinct enough to compete inside this category?

These questions are not creative betrayal. They are the difference between a game that finds an audience and a game that disappears.

Indies and Genre Innovation

Indies invent new genres faster than AAA. AAA does not invent genres at all anymore; AAA refines and exploits genres that indies and modders proved out.

Vampire Survivors (2022) birthed the "bullet heaven" subgenre. Solo developer Luca Galante built a game where you stand in a field while monsters swarm and your weapons fire automatically. Two-hour runs. Dollar-store pixel art. It looked like nothing. It became a genre — Soulstone Survivors, Halls of Torment, Brotato, Death Must Die, 20 Minutes Till Dawn — within eighteen months.

Balatro (2024) crossed poker with deck-building roguelites and made $20M+ in its first months. LocalThunk, the solo developer, did not invent poker, did not invent deck-builders, did not invent roguelites. He combined them and produced a structure that nobody had quite seen before. The genre that followed is now called "poker roguelite" or "Balatro-like."

Deep Rock Galactic is hybrid — first-person shooter, dungeon crawler, mining sim, dwarf comedy. The hybrid is the brand. No AAA studio could have shipped DRG; the design is too weird, the audience is too narrow, the marketing pitch is too long. Indies can ship games that AAA would never greenlight, and a small fraction of those weird ships become genres.

The lesson for the working designer: if you want to invent a genre, you need to be on a budget that lets you fail cheaply. Indies invent genres because indies can afford to be wrong. AAA cannot. You will not invent a genre with a sixty-million-dollar budget; you will invent it on twelve thousand dollars and a year of evenings.

Progressive Project Update — Ch 35: Position Your Game

Open your action-adventure project. We are going to locate it on the map.

Step 1: Pick the primary genre tag. Your game is an action-adventure. Within action-adventure, you have a choice: 2D action-adventure (broad), Metroidvania (narrower, more specific audience), action-RPG (different audience expectations entirely), souls-like 2D (very specific niche). Which one does your game actually fit? Be honest. If it has interconnected exploration with ability-gated areas, it is a Metroidvania. If it has stat-based progression and difficulty as a feature, it is an action-RPG. If it has both, you have a fusion to position carefully.

Step 2: Identify three direct competitors. Pick three released games that are most like yours. Not "the best games in the genre" — the games whose audience overlap is highest. If your Metroidvania is Hollow Knight-shaped, your competitors are Hollow Knight, Bloodstained: Ritual of the Night, and Blasphemous. If it is Ori-shaped, your competitors are Ori and the Will of the Wisps, Cookie Cutter, and Astalon: Tears of the Earth. The exercise is not to copy them; it is to know who you are competing with for shelf space and player attention.

Step 3: Write your design DNA in one paragraph. "My game is a 2D Metroidvania with [specific combat style] combat, [specific exploration philosophy], and [specific aesthetic / tone]. It honors the genre's expectations of [conventions you keep]. It diverges from the genre by [conventions you break or reinterpret]." If you cannot write this paragraph, you do not yet know what you are making. Find out.

Step 4: Document the conventions you honor and the ones you break. Make two lists. The "honor" list is your contract with the audience — these are the things players who like the genre will expect and that you will deliver. The "break" list is your differentiation — these are the things that make your game specifically yours. The honor list should be longer than the break list. If they are the same length or the break list is longer, you are not making a Metroidvania; you are making a strange hybrid that needs different positioning.

This document goes into your Game Design Document. It will inform your Steam page (Ch 38), your marketing copy, and the conversation you have with the first hundred players who tell you what your game is.

Common Pitfalls

Picking a genre you do not actually play. I have watched designers pitch metroidvanias having played one. I have watched designers pitch souls-likes having watched a YouTube video. You cannot design well in a genre you do not love. Play fifty hours of three games in your target genre before you commit. If you cannot stand to play fifty hours of the genre, you should not be making a game in it.

Genre-chasing a trend past its peak. The auto-battler boom died eighteen months after it started. The hero-shooter rush killed three studios. The battle-royale gold rush is now a graveyard. By the time you have shipped, the trend you chased has crested and the audience has moved on. Trend-chasing is two-year-late-to-the-party design. If you must build in a trending genre, ship within twelve months of the originating hit, or do not bother.

Claiming a new genre when you have just done a bad version of an old one. "It's a new kind of game — there's nothing else like it!" usually means "I have not researched what else exists." The cure is competitive research. Spend a week looking for games similar to yours. You will find them. Then position yourself relative to them.

Ignoring audience expectations entirely. A Metroidvania without backtracking. A roguelite without meta-progression. A horror game without scares. You can break any one expectation if the substitute is interesting. You cannot break all of them and still claim the genre tag — you have made something else, and that something else has a different audience that needs to be told it exists.

All genre and no identity. The opposite failure: the game that is so squarely in its genre that it has no specific personality. "It's a Metroidvania." Yes, but what makes it your Metroidvania? What is the hook, the aesthetic signature, the moment a player will remember and tell a friend about? Genre is the container. Identity is what fills the container. A game with all genre and no identity vanishes into the long tail.

Building fusion that does not cohere. Fusion is powerful, but two genres with conflicting tones produce confused games. A "cozy souls-like" sounds clever in a pitch and miserable to play. Make sure the genres you fuse share an emotional register, or have a deliberate plan for how you will bridge the registers.

Summary

Genre is the medium's map. It tells the player what to expect and tells the designer what conventions exist. The conventions are vocabulary — shared language that lets games communicate compactly with their audience.

Genre labels are imprecise, often retroactive, and often invented by marketing rather than designers. Treat them as descriptive heuristics, not prescriptive laws. Use them to find your audience and reason about expectations; do not let them dictate what your game must be.

Every major genre has a core loop, a progression model, a typical session length, and a set of player expectations that the designer must internalize before designing. The platformer is about jump precision and flow. The action-adventure is about exploration and combat verbs. The RPG is about character growth. The shooter is about target acquisition. The strategy game is about long time horizons. The simulation is about system stewardship. The puzzle is about figuring out. The roguelite is about run variety with permanent meta-progression. The immersive sim is about systemic emergence. The narrative game is about story-as-mechanic. The fighting game is about input mastery and conditioning. The horror game is about managed dread.

Cross-genre fusion is the cheapest form of innovation. Combining two known mechanics into a new whole carries less risk than inventing from scratch and can produce genuinely new design space (Stardew, Dark Souls, Hades, Vampire Survivors, Balatro).

Subverting a genre is powerful but only when the genre is honored first. Undertale, Spec Ops: The Line, and Inscryption all earned their subversions by mastering the genre before breaking it.

Genre is also a market signal. Your Steam tags will determine who finds your game. Choose your primary genre tag with commercial intent, then deliver the genre's promises while building the specific game you wanted to make.

Indies invent genres. AAA refines them. If you want to be on the inventing side of that line, you have to be small enough to fail cheaply and weird enough to be interesting.

Now position your own game. Pick the genre. Name three competitors. Write the design DNA paragraph. List what you honor and what you break. The act of locating your game on the map will sharpen every decision you make from this point forward — the marketing, the playtesting recruitment, the trailer cuts, the Steam capsule art, the first sentence in the press release. Genre is not a cage. It is a starting point. It tells you the room you are in. What you build inside it is up to you.

In Chapter 36 we will trace the history of genre innovation — how each of the genres in this chapter came into being, which design moves changed what games could be, and what the long arc of medium-evolution can teach designers working today. Genre is not a static category; it is a moving target that has been redrawn many times and will be redrawn again. Knowing the history makes you a better mover.