| Author |
Ataxx
|
Arno Reper
Ranch Hand
Joined: Mar 14, 2006
Posts: 286
|
|
hello,
i had to do for school an ataxx game, what i did...but now i would the computer could play against me ( because me vs me ... become a bit boring with the time ).
Do you have an idea the way it must be done? i took a look to M Doug Lea site but i don't like the way his ataxx works...so sloooooowly...
It'll be nice someone could help me.
thx a lot
arno 
ps: this is personnal work, im not cheating for school 
|
The man who makes no mistakes does not usually make anything<br /> <br />>>> SCJP 5.0 >> SCJD B&S <<< In progress
|
 |
Brendan Fosberry
Ranch Hand
Joined: Dec 16, 2005
Posts: 33
|
|
|
What are the rules of Ataxx?
|
|
Arno Reper
Ranch Hand
Joined: Mar 14, 2006
Posts: 286
|
|
hi Brendan,
its like othello...take a look at :
http://g.oswego.edu/dl/applets/micro.html
thanks for your answer,
arno 
|
|
Brendan Fosberry
Ranch Hand
Joined: Dec 16, 2005
Posts: 33
|
|
I havent done any AI yet but what I would do is when its the computers turn it weighs all its options so it says "ok...so I can move piece 5 ..to these places.." and then individually works out the gain it would get from each of those moves and goes with the best results.
if you wanted to be smart then you could consider the move after the one youre planning for..and make the AI think about how YOU can move based on where he moves to. Come up with some kind of weighting formula, maybe taking into account how desperate the AI is to finish quickly.
|
|
Layne Lund
Ranch Hand
Joined: Dec 06, 2001
Posts: 3061
|
|
Originally posted by Brendan Fosberry:
I havent done any AI yet but what I would do is when its the computers turn it weighs all its options so it says "ok...so I can move piece 5 ..to these places.." and then individually works out the gain it would get from each of those moves and goes with the best results.
if you wanted to be smart then you could consider the move after the one youre planning for..and make the AI think about how YOU can move based on where he moves to. Come up with some kind of weighting formula, maybe taking into account how desperate the AI is to finish quickly.
This is the basic idea behind an algorithm called minimax search. You can also make improvements in the search by doing alpha-beta pruning. I hope this gives you ideas if you are interested in googling for more information.
Layne
|
Java API Documentation
The Java Tutorial
|
|
Arno Reper
Ranch Hand
Joined: Mar 14, 2006
Posts: 286
|
|
hi layne,
could you explain more about the "alpha-beta pruning" ? I read an article about it on wikipedia...but it remains dark 
I also found this about alpha-beta : http://www.computerchess.us/gtchess/
do you have a tutorial or something else about?
thx
arno 
[ April 18, 2006: Message edited by: Arno Reper ]
|
|
Arno Reper
Ranch Hand
Joined: Mar 14, 2006
Posts: 286
|
|
hi,
i made an automove class and it almost works...lol, it plays well only i got some bug. Now i gonna fix them.
Thx for your help
arno
|
|
 |
|
|
subject: Ataxx
|
|
|
0
