Hello. I'm a student of Artifical Inteligence and recently I have chosen a master thesis. Goal of my work is to choose some strategic turn-based board-game (like Agricola, LeHavre, Carcassone, etc.), which is implemented allready, and then create AI player with variety of approaches, algorithms and strategies. Can someone of you refer me to the appropriate game which is implemented without AI players? The best would be implemented in Java or C#.
Thank you very much. (Sorry, if the topic is not in appropriate forum, but I'm new here)
Searching for openSource implementation of strategic game
Let me get this straight, you want us to tell you of a game, like one of the above, which currently doesn't have a version with an AI player? OR a game that doesn't have an online presence at all?
I understand now. I don't really play many board games online or on a tablet so I wouldn't be able to help you there. Anybody else know of any online games without AI?
Nope, not really. I think most of them have AI and are not open source. You might have better luck trying the forums of Brettspielwelt or BoardGameGeek. If you can't find anything you might want to consider coding a simple game yourself.
Any game that has a VASSAL module ought to do. Should be possible to add custom code (in Java) to the module, and implement the AI that way.
http://www.vassalengine.org/wiki/Category:Modules
http://www.vassalengine.org/wiki/Programming_Tutorial
You should probably ask somewhere like the VASSAL site forums first to make sure this is actually possible, but my guess is that it is.
Actually since the entire VASSAL engine is open source, you could also add the AI anywhere you want to. Looking at it that way, VASSAL is an open source implementation of many different boardgames, exactly what OP asked for and more.
Sligtly related to the subject. I am curious about what you can come up about board game AI. You see one project that I would be interested in doing in the future without actually knowing if it would be done would be to do a java framework to implement board games as a video game.
One of the biggest problem so far is the AI. The best solution would be to have an AI into the framework to prevent the game designer from desiging one. I am not sure if there could be some common concepts or mechanics behind AI design that could be applicable as a generic AI, so that the game designer only needs to bother about implementing the rules and the components of his game. The framework would take care of the interface and the AI.
I have seen some sort of academic languages for formally defining game rules, that could be used by an AI to understand how to play that game, but those specifications are complex and you would probably be happy if you managed to use it to describe anything much more complex than Tic-Tac-Toe.
Any kind of practically useful game AI will be very closely tied to a specific game. Many (most, probably) computer games that have different scenarios or maps have AI-related information manually inserted in every individual scenario/map, because even making an AI that can handle different terrain layouts is much more difficult than making one for a specific map.
For example, if you ask the player to play a card from their hand. The framework could simply chose which is the most important card to play.
On the other the designer will have to determine what makes a game more valuable than the other. In case of "I don't know", simply set the value of the cards to a random value. One each cards has a value, the AI can select which card to play when he is asked to do so.
So the idea is simply to extract the common AI concepts from the AI routine and leave only what the designer has to define. Use a random value if not defined.
Sure, but that framework would still only support games with essentially the same rules. Introduce a new rule and you need to tweak the framework and/or game-specific AI.
The best alternative would be a turn-based strategic game, which is allready implemented with OpenSource code of course, but without AI. That's because the subject of my diploma thesis is only to create AI, not implemetnig of gui, rules and whole game logic.