Conquer Club

I've been away from conquer club for several years now. I remember when I left playing that The bots were absolutely horrible as far as strategy. Artificial intelligence in backgammon and chess play at world-class levels so is there some reason that the bots for game colony don't have a higher skill level? About the only thing bots are good for is really just getting to understand how a map is played but it would be nice if they could play at such a level they may even teach humans certain tactics. Does programming bots at Conquer club pose such a problem that they can't be taught to play better?

premio53 wrote:I've been away from conquer club for several years now. I remember when I left playing that The bots were absolutely horrible as far as strategy. Artificial intelligence in backgammon and chess play at world-class levels so is there some reason that the bots for game colony don't have a higher skill level? About the only thing bots are good for is really just getting to understand how a map is played but it would be nice if they could play at such a level they may even teach humans certain tactics. Does programming bots at Conquer club pose such a problem that they can't be taught to play better?

Chess computers are programmed by professional teams of research-scientist-level programmers, with consultation from world-class Grandmasters. CC is run by one self-taught, part-time programmer.

Your question is like asking "Why can't Joe the retired farmer with his rusty spade dig a trench as fast as MegaConstructCo with its 625 bulldozers?"

I'm sure there are programmers that use this site that would love to create a strong bot. I understand what you are saying though.

I must point out though that there are many amateurs who have written very strong chess programs individually. There are world class maps that are farmed out to individuals who play on Conquer Club that are just as good or better than any professional map makers. It seems it would create a great interest to allow programmers using this site to compete against each other in creating the strongest Bot. The strongest Bots could even compete against each other for medals.

Right now the bots are at a point where they can't even obey the rules of the game. They are not programmed to recognize team games, polymorphic or trench warfare.

bigWham did put out a callout a few years back looking for AI developers (mostly for the Triqqy side games) but there were no takers.

That makes sense. The Bot for Chess is very weak, for example, by putting its queen in danger or not recognizing its likely capture.

DoomYoshi wrote:Right now the bots are at a point where they can't even obey the rules of the game. They are not programmed to recognize team games, polymorphic or trench warfare.

bigWham did put out a callout a few years back looking for AI developers (mostly for the Triqqy side games) but there were no takers.

[quote="jusplay4fun"]That makes sense. The Bot for Chess is very weak, for example, by putting its queen in danger or not recognizing its likely capture.]

There are numerous chess programs made by amatuer programmers who themselves are very poor chess players that can defeat Chess Grandmasters. Backgammon programs use dice and they are at par with the very top human Backgammon players in the world. A programmer on this site wouldn't have to program for every map and options available. They could take say just one map like "Classic" with standard rules such as using flat rate cards or no cards and tweek it until it plays to a very high level before moving on to other options.

As someone who enjoys chess and backgammon programs, I would love to use a Bot as a learning tool here at Conquer Club. If I had any programming skills at all I would jump at the chance to at least try to develop such a program.

Backgammon is probability-heavy, while chess relies on a database of moves. AI is good at these things, whereas risk seems a lot more complicated and nuanced.

You may be right because I'm not a programmer. But backgammon requires not only probability of dice but a great deal of strategy when it comes to using the cube. I understand that a Master level backgammon player will beat an average player 9 out of 10 times in a 7 point match.
Conquer club maps require probability of dice and strategy as well. I'm sure someone could probably figure out how to put the probability in with the strategy concerning maps on conquer club.