Info depth 5 seldepth 7 multipv 1 score cp 905 nodes 700 time 2 pv e4b4 e6e7 b4b2 c2c1 f4d3 c1d1 a3a2. Info depth 4 seldepth 6 multipv 1 score cp 813 nodes 300 time 2 pv e4b4 e6e7 b4b2 c2c1 f4d3 c1d1. Info depth 3 seldepth 4 multipv 1 score cp 813 nodes 200 time 2 pv e4b4 e6e7 b4b2 c2c1 f4d3 c1d1. Info depth 2 seldepth 2 multipv 1 score cp 807 nodes 100 time 1 pv e4b4 e6e7 b4b2 c2c1. Info depth 1 seldepth 1 multipv 1 score cp 783 nodes 0 time 1 pv e4b4. Rnbqkbnr/pppppppp/8/8/8/5N2/PPPPPPPP/RNBQKB1R b KQkq - 1 1 All logs are saved in files and afterwards in a (relational) database.
We will analyse all generated FEN on Igrida Cluster with Stockfish UCI Engine (e.g., depth 20 with multi-pv 1). We parse all games with a Java parser in order to analyze different moves (promotions, king castling, captured pieces.) and generate FENs. We can also get an iterator on moves to get information about Pieces Captured Count or Pieces Moves Count. The static analysis first parses each file and inspects each interesting headers information like White Elo Rating, Result or Date. Rg3+ Kf7 1/2-1/2Īs you can notice, each PGN file is separated in two parts: (1) headers (2) moves We are writing a technical report on various statistics of the database (e.g., number of unique positions)Ĭhess games are saved in PGN file, for exampleġ.
processed games with Spark SQL in order to generate CSV files together with R scripts to compute statistics. parsed various PGN files and structure each game in a (relational) database. We qualify the analysis as "static" (as opposed to "dynamic", see below) since we do not analyse moves with chess engines. We are essentially analysing headers information (related to players' ratings, dates, openings, etc.). organize a community of potential contributors for fun and profit. Chess analysis software#
investigate software engineering/scalability issues when computing millions of moves. provide open data and procedures for exploring new directions. replicate state-of-the-art research results (e.g., on cheat detection or intrinsic ratings). Our goal is to propose an open infrastructure for large-scale analysis of chess games. For instance we would like to answer a question like "Who are the best chess players in history?"įor doing so, you typically need to analyze millions of moves with chess engines it requires lots of computations. In fact numerous applications can be and have been considered such as cheat detection, computation of an intrinsic, "universal" rating, or the determination of key moments chess players blunder. 
We hope to gather various interesting insights on the skills, ratings, or styles of (famous) chess players.
We are using Igrida Cluster for large computations with Stockfish UCI Engineĭo not hesitate to participate or contact us! Objectives. We are writing a technical report on various statistics of the chessgame database (e.g., number of unique positions). This repository contains different resources (e.g., Java code) to analyse chess games. 5 millions of chess games (300+ million of chess positions) have been recorded from the very beginning of chess history to the last tournaments of Magnus Carlsen.
0 kommentar(er)
