Gogo Chess
Posted: May 10, 2019
Introduction
Since the late 1990's I have tried to implement the game of chess with artificial intelligence.
This implementation in Java finally resulted in a computer player that comes up with reasonable moves in a reasonable
time (at least, on my modern, fast laptop). The goal to implement Gogo Chess is to build a chess game that can beat me.
Admittedly, I am not a strong chess player at all. All chess education I received consists of reading one book explaining
the rules of the game and perusing David Levy's Computer Chess Compendium.
To make the application easy to use, I created a graphical user interface, which shows a large chess board. After starting up the application, the white player is the human player and the black player is the computer player. Use the buttons at the bottom right to toggle between humand and computer player. The human player can enter moves by dragging pieces.
Download
You can download Gogo Chess 1.7 by clicking this link. Double click the downloaded file to start Gogo Chess.
To run Gogo Chess you need a Java Runtime (version 8 or newer). Most modern computers have a Java Runtime installed. If your computer does not have a Java Runtime, then download one from the website of Oracle.
If you are a software developer and want to check out how I implemented this game, check out my project on Github.
Manual
The following sections explain how to indicate which player is played by a human and which by the computer and how long the computer will think. You can change these settings before the start of the game and also during the game.
Human player or computer player?
Use the buttons at the bottom right of the screen to setup whether the white player is a human or computer player and whether the black player is a human or computer player. For example in the image above, you can see the outline of a human on the white button, indicating that the white player is a human player. The outline of a computer on the black button indicates that the black player is a computer player. Click on the white or black button to toggle between human or computer player. Beware: if you first toggle the white player, then both players will be computer players and the computer starts playing the game. If you want to play as human player with black against the computer with white, first toggle the black button to human and then toggle the white button to computer.
How long does the computer player think?
The computer will think ahead a number of moves ahead. In certain situations, many moves need to be investigated, which might take a lot more time than when just a couple of moves need to be investigated. The computer player has two possible modes to control how long the computer player thinks for a move:
- The computer thinks a configurable maximum number of seconds
- The computer thinks ahead a configurable number of moves
In mode 1 the computer will think ahead a number of moves, but the computer will change this number to limit the thinking time to the configured maximum time.
At the bottom right you can control how long the computer will think. The hourglass button indicates that the computer thinks for a maximum number of seconds (mode 1). By default the computer will think for 15 seconds. Click on the arrow up or down buttons to let the computer think longer or shorter.
Click on the hourglass to change the mode to mode 2, where the computer thinks ahead a number of moves. By default the computer will think ahead 3 moves in this mode. Click on the arrow up or down buttons to change the number of moves the computer will think ahead.
Undo and redo moves
Click on the button with the arrow to the left to undo the last move. It is possible to undo as far as the beginning of the game and redo moves. Click on the button with the arrow to the right to redo a move.
The currently selected move is shown bold in the moves panel. When undoing moves, both players will be controlled by humans as a way to pause the computer from thinking up new moves again.
If instead of redoing moves a new move is made (either by a human or computer player), all moves following the current move are removed and the new move is added as last move.