Chapter 1 Introduction

1.1 Background

Chess is one of the most popular and oldest games in the world which could be dated back to the 15th century. In today’s world, more than millions of people play chess every day. According to a press release by AGON, 605 million adults play chess regularly, which is comparable to the number of regular users of Facebook. Thanks to the two biggest platforms for online chess, Chess.com and Lichess.org, either new players or experienced players could match-make with people with similar skill levels.

While chess rules are simple and easy to learn, chess is full of ideas and strategies. The number of positions grow exponentially every move. After white and black moves one piece each, the game can reach 400 positions. After the second and third move, 197742 and 121 million positions can be reached. The complexity of chess contrary to its simple rules has led it to be one of the hot topics in AI research. For this project, we took advantage of the open database of Lichess.org that provides up to 1 billion games of data per month to analyze and discovered insightful takeaways for future researchers on this topic.

1.2 Scope of the Project

Since the game of chess itself is profound enough that many has done years of research on it, it is not possible for this project to visit every aspect of chess. For this project, we will focus on the relationship of time and moves statistically instead of studying series of moves that would be related to strategy. We hope the results of this project would provide players a new perspective on choosing a move when he/she is under time pressure. Thus, we have implemented an idea of Time ratio which would be discussed in later sessions. With different plotting techniques, we expect to visualize some meaningful findings.

1.3 Introductory Chess Knowledge

As most chess games do, online chess is also played in a specific time format. You are given aspecific amount of time and have to make moves within that time. If your time reaches 0, you immediately lose. Time formats are usually notated in the form “a + b” where a is the time you start with and b is the time you get after every move also called increment.

There are also classification for chess moves. In this project, we will be mostly dealing with bad moves which classify into 3 categories, dubious, mistake, and blunder. Dubious moves are objectively bad but possibly have some merits. Mistakes are bad moves that should not have been played. Blunders are bad mistakes that end up in a loss of material or a mating sequence.

Dubious Move Mistake Blunder

1.4 Specific Libraries Used

1.4.1 Rchess

A package developed by Joshua Kunst that wraps the chess.js javascript library and the chessboardjs library for analyzing chess move, piece movement and check detection.

For more info, please visit: {https://www.github.com/jbkunst/rchess}