EPT[D] harvester optimization - Page 2
| Forum Index > StarCraft 2 Strategy |
|
Aether
Canada123 Posts
| ||
|
AssuredVacancy
United States1167 Posts
On April 20 2010 13:55 Aether wrote: The problem is that the cycle isn't long enough to support 3 miners at that distance. The cycle will always be interrupted by the third probe wandering off. So what you're saying is, the AI intentionally moves the third worker away, causing a decrease in mining rate? What if you had patch num x 3 workers mining where there would be 3 workers per patch? Would the third worker still move away? | ||
|
Aether
Canada123 Posts
| ||
|
HTX
Germany265 Posts
nicely done! | ||
|
NeoLearner
Belgium1847 Posts
 Seem very counter intuitive but I so want to try this for myself. Does someone see a way Blizzard could fix this? If they knew this, I would presume they would try and fix it. They are quite strict in starcraft 2 in making the intuitive thing the absolute best it seems... | ||
|
-orb-
United States5770 Posts
Also, do you mean like one spot away diagonally? Or like one spot away in one direction? | ||
|
Aether
Canada123 Posts
| ||
|
NB
Netherlands12045 Posts
do you want more workers so that your armysize will be smaller or less worker with a huge army size? the cap still 200 so i dont think increasing number of workers will help you that much unless u want to be like movie vs Flash and exp like crazy and try to mine out b4 your foe | ||
|
Aether
Canada123 Posts
On April 20 2010 15:59 NB wrote: this actually gv us a consideration as a zerg player, number of workers population. do you want more workers so that your armysize will be smaller or less worker with a huge army size? the cap still 200 so i dont think increasing number of workers will help you that much unless u want to be like movie vs Flash and exp like crazy and try to mine out b4 your foe That's why I said it is impractical to use it on normal mineral fields. High yields though only result in 3-4 extra workers, and if you're in a macro battle that leads to you mining out an entire high yield, your food is probably in the high one-hundreds anyway. | ||
|
ptell
United States103 Posts
| ||
|
Aether
Canada123 Posts
| ||
|
Thamoo
Canada234 Posts
On April 20 2010 16:12 Aether wrote: I have and there is no noticeable difference. It still peaks at 122 per guyser. If there's a difference at all it's like a fraction of 1 gas per minute. I even tested it with 4 guysers to see if a small difference would make itself more apparent through multiplication, but the numbers stayed consistent. 488 peak 122 per guyser. Isn't it supposed to be 180 per guysers (60 per worker)? Thought I read somewhere that workers mines mineral and gaz at the same rate (60/minute). Great work btw! Will definitivly incorporate that in my play when I get a gold expo up =D | ||
|
Sadistx
Zimbabwe5568 Posts
Replays please. | ||
|
kostja
Germany7 Posts
| ||
|
Aether
Canada123 Posts
The important thing is that the numbers are consistent relative to each other. I observed that on a normal field 8 = 325MPM 16= 650MPM and guysers = 122 GPM which all work out to a little over 40 resources per worker per minute. | ||
|
s031720
Sweden383 Posts
It sounds like youve found something noone ever thought about before; building your expo one step away increases your income with quite a bit as long as you can saturate enough. This sounds like something terran can easily exploit, building his expo as usual, but once it is saturated lift it off and move it one square for an extra 100minerals per minute. Awesome! | ||
|
sk`
Japan442 Posts
| ||
|
Aether
Canada123 Posts
On April 20 2010 16:26 kostja wrote: Mining on Metalopolis thanks I did most of my tests with protoss by the way. All workers MUST collect minerals at the same pace, so that should be irrelevant, but it's nice to see it done with another race. I would've uploaded my replay of my tests, but I didn't think anyone would want to watch me mining out mineral fields on an empty map when they could just test it themselves. I was minimizing while the workers mined too, so there are parts where nothing happens in the replay for like a half an hour because I'm watching a show or something, heh. As you can see from his replay though, it finishes a little under a minute faster, AND that is with the normal placement being over-saturated. There are 4 extra drones basically going to waste due to oversaturation. I should note one thing though, the results of these tests don't account for Terrans using MULEs. I haven't done any kind of calculations for that, I'm thinking that it would still be better than the normal placement, but the gap wouldn't be as large. Possibly even counter-beneficial depending on how the MULEs are timed and how many of them you intend to use. | ||
|
kenhennen
29 Posts
thanks for replay kostja | ||
|
Aether
Canada123 Posts
On April 20 2010 16:42 sk` wrote: So... if the node is in closest spacing, ideal worker count is roughly 2.5. +1 off and it is 3 and yields more than 2.5. Plausible... potentially plausible in SC1 as well... Yeah, I know it seems very counter-intuitive. The thing that really changes things is the fact that probes won't wait at the closer mineral fields. If they waited and only mined the node you assigned them to this wouldn't work, but because they ultimately wander off to look for another node increasing the distance to allow for full saturation actually helps. | ||
| ||
