A Survey of RTS Game AI Research and Competition in StarCraft (2013) [pdf]

2Pacalypse- wrote: post #4

qznc wrote: post #3

Is there a human vs bot comparison?

Yes. Here's a video[0] of the winner in 2014 StarCraft AI Competition against one of the better human player. More can be found at their official website[1] under Media/Files tab. [0] https://www.youtube.com/watch?v=DQI1wyRmeUQ [1] http://webdocs.cs.ualberta.ca/~cdavid/starcraftaicomp/

Seems the TvZ part of the bot could be improved by getting up turrets in time. Maybe they could get the bot to do a statistical analysis of builds they are likely to face in each match-up and use a build order that counters it.

Re: A Survey of RTS Game AI Research and Competition in StarCraft (2013) [pdf]

#6

Post by qznc » Sat, Nov 22, 2014, 8:55 PM UTC

2Pacalypse- wrote: post #4

qznc wrote: post #3

Is there a human vs bot comparison?

Yes. Here's a video[0] of the winner in 2014 StarCraft AI Competition against one of the better human player. More can be found at their official website[1] under Media/Files tab. [0] https://www.youtube.com/watch?v=DQI1wyRmeUQ [1] http://webdocs.cs.ualberta.ca/~cdavid/starcraftaicomp/

AI tries to defend with siege tanks against airborne mutalisks??

Re: A Survey of RTS Game AI Research and Competition in StarCraft (2013) [pdf]

#7

Post by z3t4 » Sat, Nov 22, 2014, 9:40 PM UTC

I don't know any games besides chess and tic tac toe where a bot have been able to beat a good player ...

Re: A Survey of RTS Game AI Research and Competition in StarCraft (2013) [pdf]

#8

Post by Argorak » Sat, Nov 22, 2014, 9:49 PM UTC

z3t4 wrote: post #7

I don't know any games besides chess and tic tac toe where a bot have been able to beat a good player ...

TicTacToe can be easily played to a draw, unconditionally. Optimal play is simple, even for humans and optimal play leads to a draw.

Re: A Survey of RTS Game AI Research and Competition in StarCraft (2013) [pdf]

#9

Post by tinco » Sat, Nov 22, 2014, 9:57 PM UTC

Glad to see the scene is still alive and well. Back when it was picking up I made a Ruby client to one of the server based BWAPI clients, I built a really neat Ruby API to it too. Unfortunately I invested so much time at that point the actual 'AI' I made just 6-pooled the opponent, in just 44 lines though[1]!

Lately the release of mruby made me think maybe I should try again :) (mruby seriously reduces the amount of effort required in getting Ruby to talk to BWAPI.)

1] https://github.com/d-snp/RProxyBot/blob/master/basic_ai.rb

Re: A Survey of RTS Game AI Research and Competition in StarCraft (2013) [pdf]

#10

Post by Malician » Sat, Nov 22, 2014, 9:59 PM UTC

ctchocula wrote: post #5

2Pacalypse- wrote: post #4

Earlier quoted context omitted.

Yes. Here's a video[0] of the winner in 2014 StarCraft AI Competition against one of the better human player. More can be found at their official website[1] under Media/Files tab. [0] https://www.youtube.com/watch?v=DQI1wyRmeUQ [1] http://webdocs.cs.ualberta.ca/~cdavid/starcraftaicomp/

Seems the TvZ part of the bot could be improved by getting up turrets in time. Maybe they could get the bot to do a statistical analysis of builds they are likely to face in each match-up and use a build order that counters it.

Well, if you build enough to be safe against just about everything, your build will lose against any build which is properly executed. If you play the numbers and just have set builds which counter the most popular builds you might face, the human will pretty quickly figure out what the AI is doing and adapt their own build.

One of the most common and basic techniques in Starcraft is to bait another player into thinking you are doing a certain popular build, and then do something theoretically inferior but which requires a wildly different response.

Know your opponent is going to have turrets up at 5:30 on the clock? New build! Hits at 5:25! Have the AI use builds which have their turrets at (x-2) minutes just in case? Well, that's going to be an awful build against almost everything.

As shown in one example in the paper, even when the AI was extremely lucky and hard-countered the human's build, the human was able to adapt, respond, and win handily.

Since even the top humans are able to trick each other in these manners, you'd basically need general/strong AI to be able to compete.