algorithm - Predict chances of winning or losing in a Chess Game -


i play chess days @ chess.com , wondering how algorithm might predict result of next game based on chess history @ site: ratings when playing previous games, opponent rating, , whether won or lost.

thanks, roberto falk

if searching way estimate skill , skill of opponent, elo-rating. developed purpose (chess). elo-rating changes based on games won , lost (and against opponent, based on elo-rating).

that means, if win against strong opponent (compared strength), elo-rating improve lot. if win against weak opponent, elo-rating improve not @ all.

the problem is, how elo-rating of opponents. if website not offer it, way compute yourself. however, need games of players on website find out how are. number of won/lost games worthless if don't factor in skill of opponent. if won 100 games against beginner, doesn't have good. if won 100 games against player, better.

if searching way estimate current state of game while playing (i.e. more win after moves executed), there several popular heuristics chess available. simple ones define value each piece (strong pieces have higher value). see player has stronger board, sum values of pieces.

for more heuristics (and other programming ideas), see here: https://chessprogramming.wikispaces.com
there several approaches estimate value of board, more reasonably write in answer.


Comments

Popular posts from this blog

asp.net mvc - SSO between MVCForum and Umbraco7 -

Python Tkinter keyboard using bind -

ubuntu - Selenium Node Not Connecting to Hub, Not Opening Port -