Start a new topic
Add Reply
as the Topic Description says,I wanna write a player for a game named Ataxx,which is defined here: http://en.wikipedia.org/wiki/Ataxx
I know how to write A player,but I'm trying to write an expert player,a player with a winning strategy...
but I have no idea about the whole algorithm part,any suggestions?
Ataxx
Page 1 of 1
Ataxx
writing a player for Ataxx
Rate Topic:
MultiQuote
#2
ali_ali
- New D.I.C Head
-
- Group: New Members
- Posts: 5
- Joined: 06-December 08
Dream Kudos: 0
Posted 06 December 2008 - 11:41 AM
as the Topic Description says,I wanna write a player for a game named Ataxx,which is defined here: http://en.wikipedia.org/wiki/Ataxx
I know how to write A player,but I'm trying to write an expert player,a player with a winning strategy...
but I have no idea about the whole algorithm part,any suggestions?
I know how to write A player,but I'm trying to write an expert player,a player with a winning strategy...
but I have no idea about the whole algorithm part,any suggestions?
Was This Post Helpful?
0
#3
William_Wilson
- lost in compilation
-
- Group: Moderators
- Posts: 4,692
- Joined: 23-December 05
Dream Kudos: 3275
Expert In: Java, C, Javascript
Posted 06 December 2008 - 11:55 AM
Please don't create duplicate threads.
What language are you planning to do this in?
What language are you planning to do this in?
Was This Post Helpful?
0
#4
ali_ali
- New D.I.C Head
-
- Group: New Members
- Posts: 5
- Joined: 06-December 08
Dream Kudos: 0
Posted 06 December 2008 - 12:25 PM
William_Wilson, on 6 Dec, 2008 - 11:55 AM, said:
Please don't create duplicate threads.
sorry I don't know how that happened...
William_Wilson, on 6 Dec, 2008 - 11:55 AM, said:
What language are you planning to do this in?
JAVA...
Was This Post Helpful?
0
#5
ali_ali
- New D.I.C Head
-
- Group: New Members
- Posts: 5
- Joined: 06-December 08
Dream Kudos: 0
Posted 07 December 2008 - 11:55 AM
in the last couple of days,I've been spending more time on the problem and came up with a bunch of new ideas which seem to work,but they have the lack of time... :-/
it's kind of like the old fashioned "brute-force" (checking all the cases and choose the best move possible)...
BUT,I'm trying to optimize it a bit,trying to decrease the time,for example decreasing the "depth" of the algorithm:
let's start from the bottom,the "Greedy algorithm" only considers the current move,now a better algorithm is to also consider the next move (the opponent's move right after ours) and then to choose the best move possible...
now the Greedy algorithm has depth 1,the second algorithm has depth 2 and so on...
now obviously the more depth our algorithm has,the higher becomes our chances of winning...
BUT,the lack of time is a serious issue here...
also another way to decrease the time,is to work on the "searching"(finding the best move possible) part,optimizing the search algorithm...
what do you guys think?
it's kind of like the old fashioned "brute-force" (checking all the cases and choose the best move possible)...
BUT,I'm trying to optimize it a bit,trying to decrease the time,for example decreasing the "depth" of the algorithm:
let's start from the bottom,the "Greedy algorithm" only considers the current move,now a better algorithm is to also consider the next move (the opponent's move right after ours) and then to choose the best move possible...
now the Greedy algorithm has depth 1,the second algorithm has depth 2 and so on...
now obviously the more depth our algorithm has,the higher becomes our chances of winning...
BUT,the lack of time is a serious issue here...
also another way to decrease the time,is to work on the "searching"(finding the best move possible) part,optimizing the search algorithm...
what do you guys think?
Was This Post Helpful?
0
#7
ali_ali
- New D.I.C Head
-
- Group: New Members
- Posts: 5
- Joined: 06-December 08
Dream Kudos: 0
Posted 07 December 2008 - 09:56 PM
LaFayette, on 7 Dec, 2008 - 02:05 PM, said:
you should take a look a the minimax game tree structure.
I've considered that,but there is two things about that:
1.the time is really important to us,each move must be made in less than 30 seconds,so I think we must decrease the depth of the algorithm or something...
2.these algorithms are MADE for the worst case ever,so they work the finest if our opponent is also strong and smart,but what if he isn't?I mean our algorithm/code prepares it self for the worst case ever and then performs a move.but maybe if he knew that the opponents move isn't that good,then there would be a better way to win,the question is that does this algorithm win against a weak opponent as well?
Was This Post Helpful?
0
Page 1 of 1
| 
