Book Note: A Theory of Fun Part 2 - Games Teach Patterns

A reading note on chapters 3-4 of A Theory of Fun for Game Design, reading games as compressed pattern-training systems.

A Theory of Fun Part 2 - Games Teach Patterns

In chapters 3 and 4, Koster treats games not as reward wrappers but as learning machines. A game is a low-risk world where a player repeats judgments: reading space, estimating probability, allocating resources, predicting an opponent, and adjusting to feedback.

This is different from shallow gamification. Points and badges do not make something a game. The real question is: what does this system make the learner practice?

L0 · Entry

  • Core idea: games compress a model of the world into repeatable problems.
  • Why this matters: a blog series, worksheet, or AI harness can also become a training space rather than a pile of content.
  • Scope: chapter 3, What Games Are, and chapter 4, What Games Teach Us.

L1 · Captures

  • Games are compressed problem spaces.
  • Fun is learning; boredom means the system has stopped teaching.
  • A good game teaches its lesson before the player leaves.
  • Games train spatial reasoning, probability, power, cooperation, and resource allocation.
  • Reward decoration is not the essence of play.

L2 · Chapter Map

Chapter My label Question
3 A concentrated lesson Is a game mainly a rule system, or a learning system?
4 What games teach What world model does play install in us?

Chapter 3 defines games as formal systems that reduce the world into repeatable challenges. Chapter 4 asks what those systems teach. Many games train ancient patterns: territory, survival, competition, and resource control. But if games teach patterns, they can also teach modern ones: negotiation, care, systems thinking, and information judgment.

L3 · Insight Cards

1. Ask what the game trains

For learning products, the first question is not “where is the scoreboard?” It is “what decision does the learner repeat?”

2. A blog can become a small game board

A single post explains. A sequence can train. Part 1 raises a question, part 2 expands the pattern, and the final part asks the reader to use it.

3. An AI harness is a feedback loop

Good prompts matter, but the loop matters more: input, test, revise, retry, record.

Comments

댓글

GitHub 계정으로 의견을 남길 수 있습니다. 댓글은 GitHub Discussions에 저장됩니다.