EPTPTR: New Worker AI? - Page 3
| Forum Index > SC2 General |
|
worked13
Australia93 Posts
| ||
|
SiguR
Canada2039 Posts
On February 28 2011 14:00 Encrypto wrote: I seriously went and tested it a couple of hours ago, and it did not do what you are saying. Check it out again. I believe you are mistaken. He's not, as far as i am aware. It might be inconsistent, but I know it definitely does rerout directly out of the CC/nexus/hatch before getting to the minerals sometimes. It has seemed pretty random at times. Maybe they simply made it more consistent. | ||
|
LegendaryZ
United States1583 Posts
On February 28 2011 09:02 Sueco wrote: Why don't you take that giant brain of yours and use it to win with superior gameplay when the game has actually started? I mean if you're so good you can split workers then the midgame must be a cakewalk right? It's a faux skill test, it serves no purpose other than showing off mindless hand speed. What really should win you a game are decisions, not hand speed, and Blizzard seems to agree. Isn't hand speed naturally part of "superior gameplay" as it applies to Starcraft? I think one of the greatest things about Starcraft has always been the need to be able to think well, but also be able to translate those thoughts into in-game commands fast enough. I don't understand this war on "mundane" or "mindless" clicking and hand speed as if the necessity for physical dexterity in the game is somehow a bad thing. I've always liked how it wasn't so much major decisions, but rather the sum of little differences over time that really showed the difference between pros and amateurs. I suppose this is a personal thing, but I don't believe this is a "faux skill test". It's just another one of the small nuances in gameplay that have been removed for no good reason. I agree with being able to rally workers to mineral fields, but I feel that there should always be room for a player to spend a little extra effort to improve on anything that is automated to make it more efficient. Certainly having the APM to spend time making sure your workers are on separate patches doesn't (and shouldn't) define your skill at the game overall, but I view it as a useful positive thing in the game because it rewards the players for putting extra effort into otherwise small details that others might not think about. | ||
|
Funkatron
United States725 Posts
| ||
|
RageOverdose
United States690 Posts
Basically, when the rallied mineral patch is full and the worker coming out would run into the patch that would be mined and would have to wait, the AI will move to an empty patch coming straight out. It can be a bit inconsistent, much like how sometimes a worker won't even try to change patches even if it has to wait to start mining. EDIT: Actually it won't even be an empty patch necessarily, just one that isn't the rallied one. | ||
|
Keitzer
United States2509 Posts
On February 28 2011 14:09 RageOverdose wrote: Not new. Basically, when the rallied mineral patch is full and the worker coming out would run into the patch that would be mined and would have to wait, the AI will move to an empty patch coming straight out. It can be a bit inconsistent, much like how sometimes a worker won't even try to change patches even if it has to wait to start mining. EDIT: Actually it won't even be an empty patch necessarily, just one that isn't the rallied one. see that completely disproves this whole post | ||
|
latan
740 Posts
| ||
|
RageOverdose
United States690 Posts
On February 28 2011 14:18 Keitzer wrote: see that completely disproves this whole post I'm not going to be bothered to do OP's research documentation for him. If you don't believe that I just tested it to make sure my post wasn't full of shit, that's fine but I've noticed this since Beta. Based on the responses it borderlines common knowledge. So if the OP really wants to prove his discovery is new then he needs to make a video of the current live setting under very similar conditions. Meh, I may do it anyway. | ||
|
NrG.ZaM
United States267 Posts
| ||
|
Encrypto
United States442 Posts
On February 28 2011 15:03 NrG.ZaM wrote: It's been like this since beta, just made a quick video showing what patch I'm running and this "new worker AI" + Show Spoiler + http://www.youtube.com/watch?v=ALZ52ZFc_5E I stand corrected. I swear I tested this and did not receive the same result. Maybe it used to be more inconsistent than it is now? I don't think I will be testing that, however. | ||
|
getpicture
141 Posts
| ||
|
t3tsubo
Canada682 Posts
| ||
|
Zarahtra
Iceland4053 Posts
On February 28 2011 15:42 Encrypto wrote: I stand corrected. I swear I tested this and did not receive the same result. Maybe it used to be more inconsistent than it is now? I don't think I will be testing that, however. I don't know exactly how it is on PTR, but it has always seemed rather random when he goes straight for empty patch and when he bounces of the rallied patch. I did some testing of this either in beta or early release, trying to figure out under which condition it happens, but I didn't get any nearer, partly since making the worker not go for the rallied patch was so random an rarely happened. If it is consistently avoiding the rallied one now, some change they did, might've effected the algorithm that handles that decision making, but I very much doubt it was a conscious change on blizzards half in the least. | ||
|
dragonblade369
Canada464 Posts
On February 28 2011 14:05 worked13 wrote: There's also a new change that alerts the player when a MULE/Inject/Chrono has completed... this game is becoming increasingly easier. You know that this will not decrease the skill ceiling right? So in the end, there's no point in whining... | ||
|
Overpowered
Czech Republic764 Posts
 | ||
|
ArdentZeal
Germany155 Posts
| ||
|
ALang
Canada288 Posts
| ||
|
TheGrimace
United States929 Posts
Also, rallying probes to certain patches isn't a skill I care too much about. Since observers hardly bounce between the bases at the beginning to show both players skillfully mining and making predictions for the match based on that, I don't think anyone was ever impressed by that. It was super important in BW, but it's a non-issue in SC2. | ||
|
nakam
Sweden245 Posts
| ||
|
latan
740 Posts
it has always been like this, i don't see anything new in the video. although maybe it has been optimized a lil bit, i haven't tested it, but the video shows nothing different than usual. so in my observations it goes like this: if a worker (rallied or not) is order to mine a patch that's currently occupied it will find another free one directly (without bouncing). but sometimes another worker will get there first so the other one will bounce. so this may explain the confusion as to why sometimes workers bounce and sometimes go straight to a free one. | ||
| ||
