Tablut is an ancient board game which was popular in nothern Europe and whose story is still uncertain. The game board is grid of 9x9 squares where two players alternate in moving their checkers:
- attacker (black soldiers)
- defender (white soldiers and one king)
This repository contains an Java-based intelligent agent able to play to the game and it was designed to take part in the Tablut Students Challenge 2020 of the Fundamentals of Artificial Intelligence and Knowledge Representation course held at the University of Bologna.
The agent is based on one of the many versions of the rules for this game, known as "Ashton Tablut" (whose rules are explained in this paper and more briefly in these slides) and it interact with a Server maintained by tutor Andrea Galassi.
Both white and black players have been implemented using Iterative Deepening search combined with MinMax algorithm and AlphaBeta cuts. Simple heuristics described in these brief slides guarantee a deep exploration of the game tree.
1° place in the group stage and in the final standings.
Out of a total of 16 games, our player won 13 games, drew 2 and lost 1. Congratulations to all participants, but especially to team tabloidi for being the only team to defeat our player. You can see the awards and results obtained in the competition in this folder or at the offial page.
First of all, you need to download the project through cloning.
git clone https://github.com/gmurro/Tablut.git
cd TablutYou need to have JDK >= 11. From Ubuntu/Debian console, you can install it with these commands:
sudo apt update
sudo apt install openjdk-11-jdk -yTo start playing, you will simply move in jars directory and run the Server with:
java -jar Server.jarThen you will run either black and white artificial intelligent player:
./runmyplayer black 60 localhost./runmyplayer white 60 localhostYou can run players also with different parameters:
./runmyplayer <role> <timeout-in-seconds> <server-ip> <debug>
<role> : role of the player in the game (black or white, it is mandatory)
<timeout-in-seconds> : time taken by the player to find the best moves (default: 60, not mandatory)
<server-ip> : ip address of the server (default: localhost, not mandatory)
<debug> : if this argument is explicit, player print logs during the search of the next move (default: none, not mandatory)We use Git for versioning.
The name of group enrolled in the Tablut Students Challenge 2020 that presented this project is brAInmates.
| Reg No. | Name | Surname | Username | |
|---|---|---|---|---|
| 997317 | Giuseppe | Murro | giuseppe.murro@studio.unibo.it |
gmurro |
| 985203 | Salvatore | Pisciotta | salvatore.pisciotta2@studio.unibo.it |
SalvoPisciotta |
| 1005271 | Giuseppe | Boezio | giuseppe.boezio@studio.unibo.it |
giuseppeboezio |
This project is licensed under the MIT License - see the LICENSE file for details