Game Design

Iteration

Why do many games change radically during development? Because design is discovery. You don't know what works until you build and test. Iteration is HOW games get good.

**Celeste** - jump feel iterated hundreds of times. Tight loop = tight controls
**Portal 2** - F-Stop mechanic killed after months. Returned to portals. Better game
**Fortnite** - pivoted from survival to battle royale. Original vision abandoned → billions

Iteration Cycles

**Iteration** - repeated cycles of building, testing, improving. Not 'fixing bugs' - reshaping design at the core based on what you learn. Each cycle: hypothesis → test → learn → change. Games are DISCOVERED, not perfectly designed upfront.

**Celeste's iteration.** The jump feel was iterated hundreds of times. Each tweak tested immediately. Tight cycle: change number → test jump → feels wrong → change again. Minutes per iteration = tight controls.

**Iteration principles:** - Small changes, frequent tests - Measure against hypothesis, not opinion - Failed iterations = successful learning - Speed of iteration > quality of first attempt

A team spent 3 months building a feature without testing. Now players hate it. The problem?

Kill Your Darlings

**Kill your darlings** - cutting features/ideas you love but that don't work. Hardest skill: emotional attachment blinds judgment. If it doesn't serve the game, it must go - no matter how much work was invested.

**Portal 2's cut content.** F-Stop (a photography mechanic) was prototyped extensively. The team loved it. It didn't fit Portal. Killed entirely, returned to portals. Darling killed = better game.

**Darling detection:** - Ask: 'Would I fight for this if I joined today?' - If defending with sunk cost → red flag - If players don't miss it when removed → it wasn't needed - Pain of cutting < pain of shipping broken

A team spent 6 months on a crafting system. Playtests show players ignore it. The lead says 'We can't cut it, too much work invested.' This is...

Scope Management

**Scope management** - controlling what's in and out of the game. Scope creep kills projects. Every 'cool addition' costs time, adds bugs, dilutes focus. Great games are tight, not sprawling. Less = more.

**Undertale's scope.** One developer, tight scope. No open world, limited combat options. Every element polished. Scope discipline → cult classic. Less = more when executed well.

**Scope discipline:** - Define the core experience early - Every addition must justify its cost - 'No' is a design decision - Better to ship tight than sprawling

Mid-development, the team wants to add multiplayer. You're the lead. Your response?

Pivot Decisions

**Pivot** - a fundamental change in game direction. Not iteration (small changes) - a pivot is 'this isn't working, try a completely different approach.' Scary, expensive, sometimes necessary. Better than shipping failure.

**Fortnite's pivot.** Originally a co-op survival builder. Years of development. Battle royale mode added late, exploded. Pivot from the core game = biggest game in the world. Original vision abandoned → billions.

**Pivot wisdom:** - Pivot early if needed (cheaper) - Pivot is not failure (it's learning) - Some team members may leave (natural) - Better to pivot than ship a disaster

Good designers get it right the first time

Good designers iterate fast and kill bad ideas early. Nobody gets it right the first time. The skill is: rapid experimentation, honest evaluation, willingness to change. Process, not innate talent

Even masters iterate. Miyamoto plays paper prototypes endlessly. Kojima rewrites constantly. Design is a search, not the execution of a perfect vision

A game has been in development for 2 years. The core loop is just unfun despite iterations. Options?

Summary

**Iteration cycles** - build → test → learn → change → repeat. Faster cycles = more learning = better game
**Kill your darlings** - cut what you love if it doesn't work. Sunk cost ≠ value. Attachment blinds judgment
**Scope management** - less is more. **Pivot** when the core is broken - painful but sometimes necessary

Вопросы для размышления

Have you ever held onto an idea too long? What made you finally let go?
How do you distinguish 'needs more iteration' from 'broken at the core'?
If you had to cut half your game's features, which ones would stay?

Связанные уроки

st-01-feedback-loops

Game Design

Iteration

Why do many games change radically during development? Because design is discovery. You don't know what works until you build and test. Iteration is HOW games get good.

**Celeste** - jump feel iterated hundreds of times. Tight loop = tight controls
**Portal 2** - F-Stop mechanic killed after months. Returned to portals. Better game
**Fortnite** - pivoted from survival to battle royale. Original vision abandoned → billions

Iteration Cycles

**Iteration principles:** - Small changes, frequent tests - Measure against hypothesis, not opinion - Failed iterations = successful learning - Speed of iteration > quality of first attempt

A team spent 3 months building a feature without testing. Now players hate it. The problem?

Kill Your Darlings

**Portal 2's cut content.** F-Stop (a photography mechanic) was prototyped extensively. The team loved it. It didn't fit Portal. Killed entirely, returned to portals. Darling killed = better game.

A team spent 6 months on a crafting system. Playtests show players ignore it. The lead says 'We can't cut it, too much work invested.' This is...

Scope Management

**Undertale's scope.** One developer, tight scope. No open world, limited combat options. Every element polished. Scope discipline → cult classic. Less = more when executed well.

**Scope discipline:** - Define the core experience early - Every addition must justify its cost - 'No' is a design decision - Better to ship tight than sprawling

Mid-development, the team wants to add multiplayer. You're the lead. Your response?

Pivot Decisions

**Pivot wisdom:** - Pivot early if needed (cheaper) - Pivot is not failure (it's learning) - Some team members may leave (natural) - Better to pivot than ship a disaster

Good designers get it right the first time

Good designers iterate fast and kill bad ideas early. Nobody gets it right the first time. The skill is: rapid experimentation, honest evaluation, willingness to change. Process, not innate talent

Even masters iterate. Miyamoto plays paper prototypes endlessly. Kojima rewrites constantly. Design is a search, not the execution of a perfect vision

A game has been in development for 2 years. The core loop is just unfun despite iterations. Options?

Summary

**Iteration cycles** - build → test → learn → change → repeat. Faster cycles = more learning = better game
**Kill your darlings** - cut what you love if it doesn't work. Sunk cost ≠ value. Attachment blinds judgment
**Scope management** - less is more. **Pivot** when the core is broken - painful but sometimes necessary

Вопросы для размышления

Have you ever held onto an idea too long? What made you finally let go?
How do you distinguish 'needs more iteration' from 'broken at the core'?
If you had to cut half your game's features, which ones would stay?

Связанные уроки

st-01-feedback-loops

Iteration

Iteration Cycles

Kill Your Darlings

Scope Management

Pivot Decisions

Summary

Related Topics

Вопросы для размышления

Связанные уроки

Iteration

Iteration Cycles

Kill Your Darlings

Scope Management

Pivot Decisions

Summary

Related Topics

Вопросы для размышления

Связанные уроки