2012 aiide predictions?

[abcdefg]

This is the biggest bwapi tournament so I wonder what are your predictions for strategies and AI techniques we will see this year? I know ualbertabot posted on TL that he has general algorithm to micro any kind of unit which their bot will use to perfectly micro all its units. Will be very hard to stop rushes if bot has perfect micro.
It would be nice to see familiar bots returning with improvements I think ualbertabot and skynet will place very high like last year but I also hope a strong zerg or terran can show up this year. Protoss seems to dominate which makes sense I guess given how much harder zerg/terran are. I think though zerg/terran has more potential with good micro.
What bot takes it this year?

[Taranok]

[quote name='abcdefg' date='19 July 2012 - 10:38 PM' timestamp='1342755530' post='10992']
This is the biggest bwapi tournament so I wonder what are your predictions for strategies and AI techniques we will see this year? I know ualbertabot posted on TL that he has general algorithm to micro any kind of unit which their bot will use to perfectly micro all its units. Will be very hard to stop rushes if bot has perfect micro.
It would be nice to see familiar bots returning with improvements I think ualbertabot and skynet will place very high like last year but I also hope a strong zerg or terran can show up this year. Protoss seems to dominate which makes sense I guess given how much harder zerg/terran are. I think though zerg/terran has more potential with good micro.
What bot takes it this year?
[/quote]

I think Skynet and Ualbertabot will do very well at the least. I'm really hoping a really good Zerg shows up though.

It'd really sweet to see some carrier micro lol, but I don't think anybody's even tried that at this point.

[krasi0]

I hope for good T and Z bots, too. I am tired of Protoss domination

[Taranok]

are you competing krasi0?

[Equalizer]

Well going by this ladder Skynet has a considerable head start compared to other existing bots.

The Overmind from 2010 still is a considerable force so a beefed up version of what they did could have potential.

Mind you I don't really have much idea of what has been in the works.

I get the impression that focusing on one strategy and executing it well has the most success against other AIs.

From what I've seen army control is the main thing holding the bots back (and this main deficiency probably is also what is helping the Protoss bots to win against bots of the other races)

[krasi0]

Taranok,probably not in this competition. I am still in the process of rewriting my bot to Java.
Equalizer, I mostly agree. I don't think anyone is still working on the Overmind project although it was very promising.
I don't know of any (current) bots that randomly switched between strategies in order to surprise their opponents but Laccolith (Skynet's author) mentioned that he's working on this for the AIIDE 2012 competition. Surprises are to come I hope!
IMO it's high time we see some non-scripted behavior (closer to theoretical AI) show some real potential in the competition.

[abcdefg]

It seems skynet page says the new 3.0 bot might not be in this due to time constraints so it might be the same skynet from last year. That being said I know the ualbertabot has been getting worked on a while and his algorithm for optimal micro was impressive so I think its possible that bot could win this time. They also said the bot would play random.

[Equalizer]

[quote name='abcdefg' date='31 July 2012 - 08:53 PM' timestamp='1343793232' post='11028']
It seems skynet page says the new 3.0 bot might not be in this due to time constraints so it might be the same skynet from last year. That being said I know the ualbertabot has been getting worked on a while and his algorithm for optimal micro was impressive so I think its possible that bot could win this time. They also said the bot would play random.
[/quote]

Playing as random maybe a handicap for ualbertabot unless they're bring a new approach to the table. Do they have page up for their past work/plans?

[abcdefg]

Not that I know of. We just have to wait and see I guess.

[abcdefg]

We just need to see some improved versions of the bots from previous years. I think if the bots worked on better engagements that would make very big difference but I know its easier said then done. A big battle with fog of war is not an easy feat to figure out things like how many siege tanks the enemy has and things like that.