← All learning paths
Learning path

Your first verifiable prototype

One scene: from a question and input to feedback, measurement and reviewing agent work.

Before you start

Start with a prototype question. No existing game project is required.

    Lesson sequence

    1. Scope and a testable prototype

      Choose one question, limit the work, and decide what to do after evaluating a prototype.

      Beginner · 6 minRead Markdown
    2. The game loop and time

      Separate simulation speed from frame rate and verify pauses, frame drops, and focus changes.

      Beginner · 6 minRead Markdown
    3. Input and feedback

      Turn button presses into gameplay intent and show whether an action was accepted, rejected, or completed.

      Beginner · 6 minRead Markdown
    4. Collision fundamentals

      Separate overlap detection, physical response, and gameplay events; check boundaries, fast objects, and repeated damage.

      Intermediate · 6 minRead Markdown
    5. States and transitions

      Define legal character states, event priority, and transition checks without a growing tangle of flags.

      Beginner · 6 minRead Markdown
    6. Game feel: making actions readable

      Tune response, emphasis, and effects so players understand an action while staying in control.

      Beginner · 6 minRead Markdown
    7. A performance budget

      Set a measurable target, locate the bottleneck with a profiler, and compare changes under the same conditions.

      Intermediate · 6 minRead Markdown
    8. Developing a game with an AI agent

      Write testable tasks, preserve project context, and accept changes based on observable gameplay behavior.

      Beginner · 6 minRead Markdown
    9. Verifying gameplay behavior

      Turn “the mechanic is done” into reproducible scenarios with initial conditions, an action and an observable result.

      Beginner · 6 minRead Markdown

    Put it into practice

    Start with a paper test of the idea. Then use the lesson tasks to build one scene and check it against the conditions above.

    Review through lenses

    Lessons contain original teaching examples and primary-source links. Numbers and scenarios illustrate a principle; verify the result in your own game.

    Outcome and verification

    One playable scene with a single action, quick restart and a reproducible check.

    • The prototype answers a recorded question and has a stopping criterion.
    • The action, pause, restart and feedback are checked in the target environment.
    • Agent changes are explained and supported by observable results.