Archive for the ‘Artificial Opponents’ Category
“My insights into Fairy Tale strategy are markedly better than random play.” OK, it’s not that comforting.
Based on feedback, I’m skipping the Neural Network and just going with a correlation matrix (which will eventually evolved). I’ve hardcoded my default correlation matrix (80 x 40) and initial weighting of cards (1 x 40). [Player #1 is the non-random one].
INFO 9859 - GameState - Player #1 won 82 games. INFO 9859 - GameState - Player #2 won 9 games. INFO 9859 - GameState - Player #3 won 5 games. INFO 9859 - GameState - Player #4 won 8 games. INFO 9859 - GameState - Player #5 won 3 games.
Right now I’m just using a “Play the card you like best each round”, but perhaps I should look at the breakpoints … if I know I won’t be playing two of the cards at all, perhaps theirs a way to let it evolve ways of ordering the rest of the card play.
Now I just have to program a mutator, a serializor and a reaper. Then it’s off to the evolutionary races! (Eventually I’ll need to write a GUI so I can play against it, assuming it gets good enough …)
I’ll write some more bridge in a few days. And some actual game stuff later…
Or just help it out. If you’ve done Neural Nets in Java, can you lend a hand?
While I figure out what to do, I’ll start writing some test cases for obscure scoring issues for the game I implemented. Or not. (Don’t expect a release anytime soon, this is mainly a NN testbed right now. Doesn’t even have a GUI yet).
Pandemic is available on BSW.
Update: Because the game requires so much communication, BSW isn’t a great way to play. Pretty interface, though.
Despite harping on the negatives, I’ve racked up 100 games online. BSW provides so much benefit for free, and costs the games companies another purchase. Dr. Pangloss knew what he was talking about! Or it may just be that I’ve gotten to play so few games in real life … only 19 games played, and Falling was four of them? Vhoja Moi!
Better luck next month!
On the one hand, I think that a very good model for game designers is to get a passionate 20%. If 20% of the gamers love your game, you’ll get better sales than if 80% consider it OK to Good. (I first heard this idea for cars, where it certainly applies much more than games; since gamers don’t just buy a single unit every few years, but the idea holds).
I’m pretty sure I’m not in Dominion’s passionate crowd. I mean, I watched a game or two and then felt no strong urge to play. Despite there being a distinct lack of gaming recently. I did get in another game of Tribune, which I’m still enjoying and now ready to get past the Carpe Diem victory conditions…
Update:OK, so I was particularly grumpy last night. I’ll get around to playing Dominion on BSW today or soon, I imagine.
As part of my ‘bridge restart,’ I’ve been playing online on Bridge Base. Anyway, this weekend they had free “Robot Race” tournaments. Each table has one human and three copies of GIB. Play as many hands as you can in some time (25 minutes) and highest total points wins. This isn’t duplicate, so each table gets its own hands.
To make it more interesting, the human player gets a good hand every deal (a solid opening hand or better each time). So you are much more likely to declare, and your bidding/play is more important than the robots. GIB plays a fairly standardish system, but sometimes makes odd calls.
But it does play quickly. I seem to be able to get 10-12 deals done in 25 minutes. I’m not sure it’s real bridge, but it’s a pretty reasonable facsimile (and it does let you practice your declarer play opposite opponents who pay attention). And, without having to worry about partner’s pesky feelings, the stacked deck aspect is fun. I don’t think I’d pony up money for the experience, but I’d play free robot races again. And not just because I won the 2nd one I entered (with 5,000 points and 49 entrants).
Update: I see from the BridgeBase notes that the human player isn’t given an opening deal … instead you get the best hand dealt. Presumably you swap with the original recipient. So while that doesn’t guarantee an opening hand, it’s pretty close and it would make counting interesting when you get an 11 point hand.
Just a note to indicate that I finally got past Suffragette City (Hard) on Rock Band. Needless to say, the next 10 songs were all beaten on the first try. Enter Sandman and that Smashing Pumpkins song are the new stumbling blocks.
Boom Blox? Never heard of it. Enjoy the 4th.