⌘K

Change language Switch ThemeSign In

Narrow Mode

Technical Details: Negamax Alpha-Beta Search Implementation

Technical Details: Negamax Alpha-Beta Search Implementation

@deedydas

This approach fundamentally uses a negamax alpha-beta tree search with pruning and iterative deepening. I tested everything with a 500ms per move limit. The main way to improve it would be to get rid of the static evaluation at the nodes and replace it with efficiently updatable neural nets (NNUEs). Also uses standard opening books and a transposition table to cache moves.

There's no offline computation or training element, so each run is like the last.

Lichess bot link: lichess.org/@/deedybot

Github repo: github.com/deedy/chess

Chess AI ranking (CCRL): computerchess.org.uk/ccrl/4040/

Bayesian ELO: remi-coulom.fr

Stash measurement: dannyhammer.github.io/engine-testing…Show More

Mar 22, 2026, 2:56 AM View on X

0 Replies

3 Retweets

36 Likes

3,674 Views  Deedy @deedydas

One Sentence Summary

Technical follow-up explaining the chess engine's architecture: negamax alpha-beta tree search with pruning and iterative deepening, with NNUE neural nets as the main improvement path.

Summary

This is a technical follow-up tweet providing implementation details of the chess engine discussed in the previous tweet. It explains that the engine uses negamax alpha-beta tree search with pruning and iterative deepening, with a 500ms per move time limit. The author notes that the main improvement path would be replacing static evaluation with efficiently updatable neural nets (NNUEs). The engine also uses standard opening books and transposition tables for move caching. Multiple resource links are provided including the Lichess bot, GitHub repository, CCRL rankings, and measurement tools.

AI Score

77

Influence Score 8

Published At Today

Language

English

Tags

Chess Engine

Alpha-Beta Search

NNUE

Rust

Algorithm HomeArticlesPodcastsVideosTweets

Technical Details: Negamax Alpha-Beta Search Implementati...