This is the repository for a genetic algorithm controlled AI player in a basketball game. The goal of the project is to compare various generation algorithms, fitness functions, selection and mutation methods in genetic algorithms in a basketball game setting, and hopefully merge different approaches to achieve a faster learning curve that avoids local optimals more effectively.
The basketball game is built with Unity2D, and it incorporates a large proportion of basketball game rules into it to better simulate an environment for learning basketball.
@ Cliff Yang & Wayne Wang
- Fixed a bug where the player will no longer be able to touch the ball.
- Fixed a bug that allows the player to ready a shot when the ball is not in hand.
- Playable version released
- Changed editor version to avoid animator controller freezing bug.
- Allowed the ball to change color back after scoring.
- Added player animations to the game.
- Shifted logic so that arm-ball collision is disbaled when holding ball.
- Fixed ball reset issues.
- Fixed a bug where the ball gets extra downwards velocity whne shooting.
- Added shooting of the ball towards the mouse direction.
- Allowed releasing of the ball.
- Added functionality of ball change color on score.
- Added material for arm to allow stronger hits.
- First version of game is published.