Learn game development by shipping one small game in four months

Four months of regular practice — about 45 minutes a day on the free Godot engine, guided by the best free tutorials there are — gets a complete beginner from nothing to one small, finished, released game. Roughly 80 hours total. You will not have built the next indie hit, and you will not be employable at a AAA studio. You will have done the one thing that almost everyone who "wants to make games" never does: finished one.

4 months · ~80 hours · one small finished game, exported and released

Weeks 1–4 · 45 min/day

1.Godot, and its official "Step by step" guide

Download Godot. It is free, open source under the MIT licence, has no royalties or sign-up, and the whole editor is a download of well under a hundred megabytes. Then work straight through the official "Step by step" getting-started series, which teaches nodes, scenes, signals, and your first GDScript, and finishes by walking you through building a complete little 2D game. Build it for real in the editor — do not just read. This is the canonical, always-current introduction, maintained by the people who make the engine.

Free

Godot · Step by step →
Weeks 4–8 · 45 min/day

2.Brackeys' Godot beginner tutorials

The docs teach you the engine; Brackeys teaches you how a real game comes together. After years as the most-watched Unity educator, Brackeys returned specifically to make free Godot tutorials, and his "How to make a video game" beginner series is the single best video on-ramp there is — player movement, level building, enemies, scoring, death and respawn, sound, and exporting, in about an hour and a quarter. Follow along building it yourself, then change something: different art, an extra mechanic. The moment you modify the tutorial game is the moment you start actually learning.

Free

Brackeys on YouTube →
Weeks 9–16 · 45 min/day

3.Ship one small game — and actually finish it

This is the entire point of the page. Pick a scope so small it feels embarrassing — one screen, one mechanic, no menus you don't need: a single-screen platformer, a falling-blocks puzzle, a tiny top-down dodge game. Build it, then do the unglamorous last 20% nobody does: a title screen, a win and lose state, sound, and an exported build you can hand to a friend. Upload it to itch.io. A finished bad game teaches you more than ten abandoned ambitious ones. Resist scope creep with everything you have.

Free; itch.io is free to publish on

itch.io →

If this doesn't fit you

If your specific goal is a job at a games studio rather than making your own games, lean toward Unity instead — it still dominates studio job listings, and its C# is a transferable, marketable language. Use the free, official Unity Learn platform, starting with the "Create with Code" course, and shore up the language itself through C#. The "finish one small game" rule applies identically; only the engine changes.

Why this path

The thing that stops aspiring game developers is never the engine or the language — it is that they start ten projects and finish none, each one bigger and vaguer than the last. The bottleneck is scope and follow-through, not skill. Godot is the right first engine because it is free, light, and has no friction; the official docs and Brackeys get you building fast; and the final phase exists solely to force you across the finish line on something tiny. Shipping one finished game, however small, teaches you the whole loop — design, build, polish, release — that no tutorial can. Skip it and you will have watched game development without ever having done it.