You play against an opponent written from scratch using the minimax algorithm with alpha-beta pruning — the same family of algorithms behind chess engines. It memorizes no moves: each turn it explores the game tree a few thousand positions ahead and picks the best one. You are red, you move first.
Positive = the AI is ahead, negative = you are.
🌳 How it works: the AI builds the tree of every possible continuation down to the chosen depth and assumes you both play optimally — minimax.
Alpha-beta pruning cuts the branches that cannot change the outcome, reducing the positions to evaluate hundreds of times over.
Positions with no clear win are scored with a heuristic: aligned windows of 4 cells are worth more, and the center column is valuable. Ordering the moves from the center outward makes pruning even more effective.