Game design fundamentals

In short

Game design is the craft of deciding what the player does, what the system does back, and why that exchange stays interesting over hours of play — it is about rules, goals, feedback and pacing, not about art or code. It sits above the engine: the same core loop can be built in Unity, Unreal or Godot, and the designer's deliverables are documents, spreadsheets, paper prototypes and tuned numbers. Indian universities examine it in B.Sc Animation/Gaming and B.Des papers as "Principles of Game Design", "Game Mechanics and Level Design" and "Game Production", usually with a GDD plus a playable prototype as the practical. It is the first-round filter for game designer, level designer and technical designer roles at Indian studios (Nazara, Dream11, Games24x7, Zynga India, Ubisoft Pune/Mumbai, Rockstar India, SuperGaming, Lila Games, Moonfrog, Octro, Mech Mocha alumni teams) and for live-ops and game-economy analyst jobs.

Explained

What a game is: formal elements and MDA

• Working definition: a system in which players engage in an artificial conflict, defined by rules, that results in a quantifiable outcome (Salen and Zimmerman). Exam answers usually also quote Sid Meier's "a series of interesting choices".
• Formal elements (Fullerton's framework): players, objectives, procedures (rules of action), rules (constraints), resources, conflict, boundaries, outcome. A question like "identify the formal elements of Ludo" expects all eight.
• Dramatic elements: challenge, play, premise, character, story — the layer that gives the formal system meaning.
• MDA framework (Hunicke, LeBlanc, Zubek): Mechanics (rules and data the designer writes) → Dynamics (run-time behaviour when players interact) → Aesthetics (the emotional response). Designer reads left to right; player experiences right to left.
• LeBlanc's 8 kinds of fun: sensation, fantasy, narrative, challenge, fellowship, discovery, expression, submission. Used to answer "what fun are you designing for?" instead of the vague word "fun".
• Elemental tetrad (Schell): mechanics, story, aesthetics, technology — all four must support one experience.
• Mechanic vs dynamic: "a grenade has 3 s fuse and 5 m radius" is a mechanic; "players cook grenades and flush campers" is a dynamic. Designers cannot author dynamics directly, only the mechanics that produce them.

The core loop, verbs and goals

• Core mechanic / verb: the one action the player repeats most — jump (Mario), shoot (Valorant), match-3 swap (Candy Crush), swipe-to-dodge (Subway Surfers). Every other system exists to make that verb interesting.
• Core loop: the shortest repeating cycle — *act → feedback → reward → upgrade → act again*. Measured in seconds (a shot), minutes (a match), and sessions (a daily quest).
• Nested loops: micro loop (one input, <1 s), mid loop (one level/match, 1–10 min), meta loop (progression, days or weeks). A design fails when any one loop is weak — strong combat with no meta loop gives poor retention; strong meta with weak combat gives poor review scores.
• Goals: long-term (finish the campaign), short-term (clear this wave), and self-set (speed-run, fashion). Clear goal + visible progress + meaningful feedback is the minimum for a readable loop.
• Interesting choice: has at least two viable options, non-obvious trade-offs, and consequences the player can perceive. Opposites are *dominant strategy* (one option always best) and *false choice* (no outcome difference).
• Game feel / "juice": input responsiveness, animation anticipation, screen shake, hit-stop, particles, sound and haptics. Cheap to add, large effect on perceived quality; examined as "polish" or "feedback layer".

Mechanics, systems and randomness

• Resource systems: sources (how resources enter), sinks (how they leave), converters, and traders. A healthy economy has sinks matched to sources, or inflation ruins progression.
• Feedback loops: a positive (reinforcing) loop amplifies a lead — snowballing, as in Monopoly or farm-based RTS; a negative (balancing) loop closes gaps — rubber-banding in Mario Kart, comeback gold in MOBAs. Design answer: positive loops shorten games, negative loops keep them tense.
• Catch-up mechanics: rubber-banding, bonus for losing side, draft order reversal. Risk: punishing skill ("participation reward").
• Information structure: perfect information (chess), imperfect (poker, fog of war), hidden roles (Among Us). Hidden information creates bluffing dynamics.
• Randomness types: input randomness happens before decisions (board layout, card draw, procedural map) and creates variety; output randomness happens after the decision (hit chance, crit, loot drop) and creates tension but can feel unfair.
• Taming randomness: pity timers, pseudo-random distribution (PRD, as in Dota 2 crits), shuffle bags, bad-luck protection, deterministic damage with random flavour.
• Skill vs chance: legally important in India — formats of rummy, fantasy sports and poker have been litigated on the "game of skill vs game of chance" distinction, and the Promotion and Regulation of Online Gaming Act, 2025 bans online money games while promoting e-sports and social games; implementation and rules have been under challenge, so state the position as "real-money formats are restricted under the 2025 central law" rather than quoting a settled nationwide rulebook.
• Determinism and netcode: lockstep deterministic simulation for RTS; server-authoritative for competitive shooters — a designer must know which, because it limits whether randomness can be client-side.

Balance, difficulty and progression

• Types of balance: fairness (symmetry or compensated asymmetry), challenge-vs-skill, cost-vs-benefit, strategy diversity (no dominant build), payout balance (reward matches effort).
• Transitive vs intransitive balance: transitive = a strict power ladder priced by cost (a ₹-cost tank beats a cheaper scout); intransitive = rock-paper-scissors counters (pike > cavalry > archer > pike).
• Flow channel (Csikszentmihalyi): keep challenge slightly above current skill; too high → anxiety, too low → boredom. Difficulty should rise as a sawtooth — spike, then relief — not a straight line.
• Difficulty levers: enemy count, HP and damage, time pressure, resource scarcity, precision required, information withheld, enemy variety and composition.
• Dynamic difficulty adjustment (DDA): silent tuning based on player performance; must be invisible or players feel cheated.
• Progression curves: linear, exponential and logarithmic. XP thresholds usually grow geometrically; power grows more slowly than cost so each level feels earned but not trivialising.
• Tutorialisation: teach one mechanic at a time in a safe space, then combine (Nintendo's "introduce, develop, twist, conclude"). Avoid walls of text; prefer guided discovery.
• Onboarding metrics: FTUE (first-time user experience) completion rate; a drop at a specific step is a design bug, not a player failure.

Level design and pacing

• Level design process: concept and beat sheet → paper/graph layout → greybox (blockout with primitives) → gameplay tuning → art pass → optimisation pass.
• Guiding the player: composition and leading lines, light and colour contrast, landmarks, breadcrumbs, affordances (ledges painted yellow), and critical path vs optional branches.
• Layout archetypes: linear, hub-and-spoke, open world, arena, maze/metroidvania (gated by abilities), three-lane (MOBA).
• Pacing: alternate intensity — combat, exploration, puzzle, cut-scene — to make peaks register. Plot the intensity curve per level.
• Encounter design: enemy roles (fodder, ranged pressure, shielded, boss), spawn points, cover density, sight lines, choke points, flank routes, and reward placement.
• Metrics in level design: blockout must use the character's measurements — jump height and distance, crouch height, sprint speed, weapon effective range. These are locked early, because changing them mid-project forces every built level to be re-tuned.
• Procedural generation: rules plus seeds for replayability (roguelikes); needs validation passes so every generated level is completable.
• Accessibility: remappable controls, subtitles with size options, colour-blind-safe palettes, difficulty options, no essential information carried by colour alone.

Players, genres and documentation

• Bartle's player types (MUD-derived): achievers, explorers, socialisers, killers; useful for multiplayer feature planning. Quantic Foundry's motivation model (action, mastery, social, immersion, achievement, creativity) is the modern commercial version.
• Genres and their design pressure: action/platformer (controls and metrics), shooter (weapon balance, map symmetry), RPG (stat systems, quests), strategy (economy, counters), puzzle (rule teaching), simulation, sports, racing, idle/clicker (number curves), hyper-casual (one-tap, instant loop), card/board.
• Game Design Document (GDD): overview and pillars, target audience and platform, core loop, mechanics and controls, progression, economy, level list, UI/UX flows, art and audio direction, technical requirements, monetisation, risks, milestones. Modern practice: a short living GDD plus a wiki plus balance spreadsheets, not a 200-page book.
• One-page design and pitch: elevator pitch, USP, reference games, target platform, scope in team-months.
• Iterative loop: prototype → playtest → analyse → revise. Paper prototyping tests rules in hours; vertical slice proves one polished minute of the real game.
• Playtesting rules: do not explain, do not defend, watch hands and face, ask questions after; use a fresh tester for onboarding tests, a returning tester for depth.
• Monetisation models: premium (one-time price), free-to-play with IAP (cosmetics, battle pass, soft/hard currency), ads (rewarded video is the least intrusive), subscription. Dark patterns — pay-to-win, countdown pressure, confusing currencies — cost trust and may run into consumer-protection scrutiny.
• Ratings and compliance in India: Google Play and other IARC-participating stores use the IARC questionnaire for age ratings, while Apple's App Store runs its own age-rating questionnaire; the Information Technology Act, 2000 and the IT Rules, 2021 govern intermediary and online-content obligations; for personal data, the Digital Personal Data Protection Act, 2023 is enacted but the Data Fiduciary duties under the DPDP Rules, 2025 (notice, consent, breach intimation, erasure) apply only from 13 May 2027 — until then IT Act s.43A and the SPDI Rules apply, which matters for games collecting minors' data. GST on online gaming services is at the standard rate structure (5%/18%/40% slabs since 22 September 2025) — the applicable slab depends on the supply classification, so confirm before quoting a rate.

Key points

• Core mechanic - the single verb the player repeats; everything else serves it.
• Core loop - act → feedback → reward → upgrade; nested as micro, mid and meta loops.
• MDA - mechanics are authored, dynamics emerge, aesthetics are felt.
• Formal elements - players, objectives, procedures, rules, resources, conflict, boundaries, outcome.
• Interesting choice - two or more viable options with perceivable, non-obvious trade-offs.
• Dominant strategy - an option always best; a balance failure that collapses choice.
• Input vs output randomness - before the decision (variety) vs after it (tension, possible unfairness).
• Positive feedback loop - amplifies leads and ends games faster; negative loop keeps them close.
• Flow channel - challenge kept just above skill; sawtooth difficulty, not a ramp.
• Transitive balance - power priced by cost; intransitive - rock-paper-scissors counters.
• Source and sink - every currency needs both, or the economy inflates.
• Greybox - playable blockout tuned before any art is made.
• Character metrics - jump height, reach, speed; fixed early, levels built to them.
• Vertical slice - one fully polished minute proving the whole game.
• Paper prototype - rules tested with cards and dice in hours, not weeks.
• Game feel / juice - responsiveness and feedback layer; cheapest large quality gain.
• Retention D1/D7/D30 - the live-game scoreboard for whether the loop works.
• Pity timer / PRD - mechanisms that cap bad luck in random rewards.

Formulas and examples

Example 1 — Weapon balance by TTK. Rifle A: 25 damage per shot, 6 shots/s, 20% crit chance, 2× crit multiplier. Target: 150 HP, 25% damage reduction.

• DPS = 25 × 6 × (1 + 0.20 × (2 − 1)) = 150 × 1.2 = 180 HP/s.
• EHP = 150 / (1 − 0.25) = 200 HP.
• TTK = 200 / 180 = 1.11 s.
• Design check: if the design target is 1.5 s, reduce fire rate to 6 × (1.11/1.5) = 4.44 shots/s, i.e. 4.4 shots/s (DPS 133.3 → TTK = 200/133.3 = 1.50 s).

Example 2 — Progression curve. Geometric XP with B = 100 and r = 1.5.

• XP(1) = 100, XP(2) = 150, XP(3) = 225, XP(4) = 337.5 → 338, XP(5) = 506.
• Cumulative to reach level 5 = 100 + 150 + 225 + 338 + 506 = 1,319 XP.
• A mission gives 60 XP and lasts 4 min → 1,319/60 = 21.98 → 22 missions ≈ 88 min to hit level 5. If the design wants level 5 inside 60 min, raise mission XP to 1,319/15 ≈ 88 XP (15 missions × 4 min = 60 min).

Example 3 — Gacha drop rate and pity. Per-pull rate for a rare hero p = 0.015 (1.5%).

• P(at least one in 50 pulls) = 1 − (1 − 0.015)⁵⁰ = 1 − 0.985⁵⁰ = 1 − 0.4697 = 0.5303 ≈ 53.0%.
• Expected pulls to first success = 1/p = 66.7 pulls.
• With a hard pity at 80 pulls, the worst case is bounded at 80. At ₹80 per pull, worst-case spend = ₹6,400; expected spend without pity = 66.7 × 80 ≈ ₹5,336.

Example 4 — Economy source/sink balance. A player earns 1,200 soft coins per day (sources: 800 from matches, 300 daily quest, 100 ad reward). Sinks: upgrades cost 900/day at the intended pace, repairs 150/day.

• Net = 1,200 − 1,050 = +150 coins/day, i.e. 4,500 surplus per month → inflation; late-game players stop valuing coins.
• Fix options: add a cosmetic sink of 150/day, or remove the 100/day ad reward and add a 50/day sink. Choosing the sink keeps the reward loop intact and absorbs the whole surplus.

Example 5 — Is the loop worth building? (live-ops maths). 100,000 installs, D1 = 38%, D7 = 14%, D30 = 5%; ARPDAU ₹2.40; average lifetime 9 days; CPI ₹18.

• LTV ≈ 2.40 × 9 = ₹21.60 > CPI ₹18 → marginally profitable, margin ₹3.60 per install (20% over CPI).
• Total contribution over the cohort = 100,000 × 3.60 = ₹3.6 lakh.
• Sensitivity: if D1 falls to 30% and lifetime to 7 days, LTV = ₹16.80 < CPI → the cohort loses ₹1.20 per install. D1 is therefore the metric onboarding design must protect.

Where it is used

• Mobile free-to-play studios (Nazara, Moonfrog, Octro, SuperGaming, Gameberry): core-loop design, level packs, economy tuning and A/B-tested onboarding for Indian-language audiences.
• Real-money and fantasy platforms (Dream11, Games24x7, MPL, Zupee): contest design, matchmaking and skill-vs-chance structure under the Promotion and Regulation of Online Gaming Act, 2025 and state-level restrictions.
• AAA development and support studios (Ubisoft Pune and Mumbai, Rockstar India, Sumo/Keywords India, Lakshya Digital): level design, encounter and systems design, co-development and QA design support.
• E-sports and competitive titles: map balance, patch cycles, rating systems (Elo/Glicko-style) and tournament formats — BGMI, Valorant and EA Sports FC circuits run by Indian organisers.
• Serious games, ed-tech and training: gamified learning (BYJU'S-style reward loops), skill simulators, defence and medical training sims, government awareness games — same loop and progression theory.
• Job roles: game designer, level designer, technical designer, systems/economy designer, live-ops manager, game analyst, game producer, UX designer; portfolio expectations are a GDD, one playable prototype and a balance spreadsheet.

Interview questions

• Q: What is a core loop and how is it different from a core mechanic? A: The core mechanic is the single repeated verb (shoot, match, jump); the core loop is the repeating cycle of action, feedback, reward and upgrade built around that verb.
• Q: Explain MDA with an example. A: Mechanics are the authored rules (a 3 s grenade fuse), dynamics are the behaviour that emerges (players cooking grenades to flush cover), aesthetics are the resulting feeling (tension and mastery).
• Q: What makes a choice "interesting"? A: At least two viable options with real trade-offs the player can perceive, and no option that is always correct.
• Q: Difference between input and output randomness? A: Input randomness occurs before the decision and creates varied puzzles; output randomness resolves after the decision and creates tension but risks feeling unfair.
• Q: How do positive and negative feedback loops change a game? A: Positive loops amplify leads and shorten matches; negative loops close gaps and keep matches tense — most commercial games use a negative loop to protect engagement.
• Q: How would you balance an overpowered character without nerfing its fun? A: Raise its cost or cooldown, narrow its effective range, give it a clear counter, or shift power from reliability to high-risk burst — change the conditions of success rather than deleting the ability.
• Q: What is a vertical slice and why do producers insist on one? A: One short, fully polished segment with final-quality mechanics, art and audio; it proves the experience and the pipeline before the team scales up content.
• Q: Your game has strong D1 retention but D7 collapses. What do you investigate? A: The meta loop — progression pacing, daily goals, social hooks and whether rewards still feel meaningful after the tutorial — because onboarding clearly works.
• Q: How do you decide difficulty for a mobile puzzle game? A: Set a target win-rate band per level from playtest and analytics (for example 60–75% on early levels, lower at designed spikes), then tune move counts and board layout to hit it.
• Q: Why greybox before art? A: Layouts change constantly during tuning; blocking out with primitives lets you fix pacing, sight lines and metrics at near-zero cost, and protects the art team from rework.
• Q: What is a pity timer and why use it? A: A guaranteed reward after a set number of unsuccessful pulls; it caps the worst-case experience of output randomness and keeps spending predictable and fair.
• Q: Is a game of skill legally different from a game of chance in India? A: Yes — the distinction has long driven Indian litigation over rummy, poker and fantasy sports, and the Promotion and Regulation of Online Gaming Act, 2025 now restricts online money games while supporting e-sports and social games; parts of the framework have been challenged, so the operative rules should be checked for the specific format.

Common mistakes

• Describing art, story or engine features and calling it game design — examiners want rules, loops, goals and feedback.
• Writing a feature list instead of a loop: no stated core verb, no reward cycle, no reason to replay.
• Designing systems with sources but no sinks, so currency and power inflate and progression stops mattering.
• Confusing a mechanic with a dynamic in MDA answers, and claiming you can author the player's emotion directly.
• Ramping difficulty as a straight line with no relief beats, then concluding "the players are bad" when testing shows a wall.
• Explaining the game to playtesters, then reporting that onboarding works — the explanation is the thing being tested.
• Scoping a student project at open-world size; a tuned 10-minute vertical slice scores and interviews far better than an unfinished epic.

Related topics

Game engines: Unity and Unreal · UX research and usability · VFX: compositing, rotoscoping and matchmoving · 3D modelling and texturing pipelines · Game narrative and level art · Game analytics and live operations