EPTThat APM thing is questionable as well
StarCraft II: DeepMind Demonstration: Jan 24 - Page 10
| Forum Index > SC2 General |
Poopi
France12911 Posts
That APM thing is questionable as well | ||
|
Acrofales
Spain18262 Posts
On January 25 2019 03:50 Acrofales wrote: Well, that removes the "partial observable" part... guess they couldn't deal with that  Hm. Not sure now. TLO just said that proxy could have worked because the AI doesn't scout well... so they *don't* see the whole map? | ||
|
StasisField
United States1086 Posts
| ||
Waxangel
United States33592 Posts
| ||
|
IshinShishi
Japan6156 Posts
On January 25 2019 03:53 Garrl wrote: they would've put a protoss player against it if it could beat them. you guys seriously overestimate the level of play of GM players, I dont think it could beat a pro, but low-mid GM? For sure. | ||
|
imre
France9263 Posts
On January 25 2019 03:55 Poopi wrote: Still if this is the result we get when the AI can view full map (beating an offracing TLO), it'll be a long time before they can beat pros in fair conditions. That APM thing is questionable as well From what we saw the APM looked a bit below the standard pro player (especially game 1 a Korean Protoss would have had a way better prism micro) so this point isn't an issue. | ||
|
Waxangel
United States33592 Posts
On January 25 2019 03:55 Acrofales wrote: Hm. Not sure now. TLO just said that proxy could have worked because the AI doesn't scout well... so they *don't* see the whole map? the guy worded it really poorly; it doesn't have maphack, more like a "zoom-out-hack" (not EXACTLY this either) | ||
|
Xamo
Spain886 Posts
| ||
|
Dav1oN
Ukraine3164 Posts
| ||
|
Nakajin
Canada8989 Posts
| ||
|
Pandain
United States12989 Posts
| ||
|
Gandie
158 Posts
| ||
|
Kazi25
Philippines236 Posts
| ||
|
Musicus
Germany23576 Posts
| ||
|
Howard_Kao
China261 Posts
| ||
|
I wasbanned fromthis
113 Posts
| ||
|
Serimek
France2274 Posts
| ||
|
TheDougler
Canada8307 Posts
| ||
|
imre
France9263 Posts
| ||
|
Aeromi
France14461 Posts
| ||
| ||
