The one domain where the reps are their own portfolio. Games hiring weights a shipped thing more heavily than most software hiring — a jam entry with twenty ratings outperforms a perfect interview answer.

Difficulty: 🟢 starter (hours–days) · 🟡 intermediate (a week or two) · 🔴 ambitious / portfolio-grade. ⭐ = highest signal.

The ladder

  • 🟢 ⭐ Do a game jam. Ludum Dare, GMTK, or any of the hundreds on itch.io. A theme, 48 hours, ship something. The single highest-value item in this section — it forces scope discipline and an actual finish line, and produces a public artefact with feedback from strangers → 08.
  • 🟢 Ship Pong, then Breakout. Two weekends, publicly on itch.io. Menus, sound, a build someone else can run. The last 20% is where the learning is → the ladder.
  • 🟢 Write a shader that colours a surface by its normal, then by a light direction. Ten lines on Shadertoy. It demystifies the whole of graphics → graphics exercise 1.
  • 🟢 Implement Euler integration and watch a pendulum gain energy, then change two lines to semi-implicit and watch it stop. Ten minutes, and note 04’s central claim becomes a memory → 04.
  • 🟡 A platformer with good feel — coyote time, jump buffering, variable jump height, all hand-authored rather than physics-driven. Feel is a real, teachable skill and it’s what interviews ask about → 04.
  • 🟡 Profile a frame with RenderDoc. Capture, find where 16 ms goes, fix one thing, re-measure. Seeing the budget is worth more than reading about it → 03.
  • 🟡 A top-down roguelike — enemy state machines, A* pathfinding on a navmesh or grid, procedural levels → 05.
  • 🔴 ⭐ Build a 2D engine from scratch and ship one small game with it. Window, fixed-timestep loop, sprite renderer, collision, audio. The game is what forces you to build the parts engine projects skip → from-scratch.
  • 🔴 Networked multiplayer for a simple game — client prediction, server reconciliation, entity interpolation. The hardest common problem in the field; do it after something simpler ships → 06.

If you do one: the game jam. It tests scope, finishing, and whether you actually enjoy this — for the price of a weekend, before you commit years.

If you only do one

Ship a game jam entry. 48 hours, a hard deadline, and a finished thing — which is the exact skill the industry screens for, and the one this vault has least of.

↑ Back to Game Development