Alfredo De Donno, known as sirAlfry, ported a collection of classic arcade games from Python/Pygame to Phaser 4.1. The result is nine fully playable games accessible from a single retro CRT-style launcher, complete with scanline effects, neon colours and a vignette. The whole collection runs in the browser with no install, no bundler and no external assets. Everything is rendered procedurally through Phaser's Graphics API.

The Nine Games in the Retro Arcade Collection

  • Snake — particle effects, colour gradient on the body, timed bonus fruit
  • Tetris — hold piece, ghost piece, next piece preview, level system and line clear animation
  • Pong — three modes (1P, 2P, VS AI), trail effect, AI with variable difficulty
  • Breakout — multi-resistance bricks, lives and progressive acceleration
  • Space Invaders — enemies that shoot back, progressive waves and lives system
  • Maze Runner — DFS-generated mazes, three difficulty levels and breadcrumb trail
  • Memory Game — 4x4 pairs, timed input block and move counter
  • Puzzle 15 — solvability check, tile correctness indicator and timer
  • Tic Tac Toe — Minimax AI, three modes (1P random, 2P, AI) and score tracking

How It Was Built with Phaser 4

Each game is an independent Phaser.Scene. The launcher transitions between them with fade in/out via scene.start(), and each scene manages its own input, rendering and state without shared dependencies.

There is no bundler in the stack. Phaser 4 is loaded via CDN ESM from jsDelivr, and the browser caches it automatically between scenes. All rendering goes through Phaser's Graphics API with WebGL: no sprite sheets, no external images, everything drawn in code. High scores are saved to localStorage with functions local to each scene.

The Tic Tac Toe implementation includes a full Minimax AI with three modes: single player against a random opponent, two players on the same device, and single player against the AI.

Play the Collection

All nine games, free in the browser. No install needed.

Play Now

View the Source

MIT licensed. Full source on GitHub.

View on GitHub