Alpha   •   11 months ago

Using AI Heuristics to Build a Smarter Block Blast Puzzle Solver

I'm working on a project that involves solving the Block Blast puzzle game like https://blockblastsolverlive.com/ a logic-based grid-filling game where players must place Tetris-like blocks optimally to clear rows or columns.

To take it a step further, I’ve been exploring the use of AI techniques like heuristic search (greedy algorithms, A* variants) and reinforcement learning to create a smarter solver that mimics or outperforms human strategies. The idea is to evaluate multiple possible board states in real time and choose placements that maximize long-term score, not just immediate gains.

I’m curious how others have used AI in similar grid-based games or decision-heavy environments.

Have you tried applying AI to games that involve spatial reasoning or piece placement?

Any thoughts on integrating ML models versus rule-based solvers for this kind of task?

Would love to exchange ideas and hear about other “AI in Action” applications in game logic, puzzle solving, or strategy planning.