EPTHow does it compare to Starcraft 2? I heard with better AI its a lot more complicated, and in some instances having more workers might lower income? I don't know much about this so I'll let the more knowledgeable post their opinions
SC2 Optimal worker saturation
| Forum Index > SC2 General |
|
SubtleArt
2710 Posts
How does it compare to Starcraft 2? I heard with better AI its a lot more complicated, and in some instances having more workers might lower income? I don't know much about this so I'll let the more knowledgeable post their opinions | ||
|
magz4
United States9 Posts
| ||
|
Volshok
United States349 Posts
6: Zerg: ~248 Protoss: ~248 10: Zerg: ~419 Protoss: ~419 14: Zerg: ~590 Protoss: ~590 18: Zerg: ~705 Protoss: ~705 22: Zerg: ~762 Protoss: ~762 Here is the replay of the test game proving this: http://www.mediafire.com/?uyqjjdymmhy edit: also here's the ratios of going from step to stop going up 4 workers: 6>10 worker ratio: 1.6667 Mining ratio: 1.6895 10>14 worker ratio: 1.4 mining ratio: 1.4081 14>18 worker ratio: 1.2857 mining ratio: 1.1949 18>22 worker ratio: 1.2222 mining ratio: 1.0809 as you can see you're getting basically 1:1 bang for your buck up to 14 workers, but after that your payoff per worker added drops and it gets less and less efficient to add workers. Along with Zeke's info Also, note that each worker has a walking speed of 2.8125. Every single one. There are also NO known differences in mining speed. | ||
|
tabako
United States35 Posts
| ||
|
Zavior
Finland753 Posts
On February 25 2010 01:48 tabako wrote: Why would it vary by race? I must be missing something pretty basic here... My guess would be that it varies by race because... the workers are different! At least zerg workers move a bit faster on the creep if I'm not completely mistaken  | ||
|
Integra
Sweden5626 Posts
On February 25 2010 01:08 Volshok wrote: I don't have Terran. Both of these numbers were taken from posters on this board, but I didn't grab the names. Sorry to the OPs, I copy-pasted these for friends on another forum. ill add another perspective that supports the statement above about Zerg: According to the income tool in the replay on a 8 mineral field 8 drones gather around 300-350 minerals. 16 drones will gathers around 2x 300-350 =600-700. The soft cap of mining per base is close to 8x3 drones, after this each drone mine 50% less until you reach 8x4 which is the hard cap where additional worker won't make more resources. Once you have reached 8x2 drones each new drone you produce after that will gather roughly half the minerals of a drone, if you make 4 additional drones after the 8x2 you will in reality only get minerals from 2 drones. If you make 8 additional drones after the 8x2 you will only be getting minerals from 4 drones. any more drones after 2x8 plus 8 is a waste of resources and supplies. | ||
|
SubtleArt
2710 Posts
On February 25 2010 01:08 Volshok wrote: I don't have Terran. Both of these numbers were taken from posters on this board, but I didn't grab the names. Sorry to the OPs, I copy-pasted these for friends on another forum. Holy shit why is it THAT much different for Zergs? Was this intentional :O? | ||
|
Drunken.Jedi
Germany446 Posts
On February 25 2010 01:57 Zavior wrote: My guess would be that it varies by race because... the workers are different! At least zerg workers move a bit faster on the creep if I'm not completely mistaken I'm pretty sure that drones do not move faster on creep. At least it used to be like that when the movement speed bonus for zerg units was first implemented. Also, from watching streams and vods, drones do not seem to be any faster on creep. Apart from that, movement speed has nothing whatever to do with the amount of minerals you get from full saturation. All that increased movementspeed would do is to decrease the number of workers required for full saturation. My best guess as to why the numbers for zerg and terran are so different is that one of the two players (or both) that tested this somehow made a mistake or that they used different methods of measuring income. | ||
Zelniq
United States7166 Posts
i think the numbers are mistaken | ||
|
MannerMan
371 Posts
I don't know just speculating. | ||
|
Queequeg
Germany263 Posts
16 SCVs: 610-667, avg 639, ~39.9 per SCV 20 SCVs: 705-800, avg 753, ~37.6 per SCV, 4 more SCVs yield 29 each 24 SCVs: 781-876, avg 829, ~34.5 per SCV, 4 more SCVs yield 19 each 28 SCVs: 781-876 ==> 24 = saturated So if you have 20 SCVs mining, and you add another SCV, it will only be half as effective. | ||
|
daywiss
United States83 Posts
| ||
|
sorech02
United States15 Posts
On February 25 2010 01:08 Volshok wrote: I don't have Terran. Both of these numbers were taken from posters on this board, but I didn't grab the names. Sorry to the OPs, I copy-pasted these for friends on another forum. I'm guessing that since he pasted this together from two sources, they each had a different time limit, as well as other potential factors I'm not thinking of. The two sets can't be directly compared in this case. | ||
|
HolydaKing
21254 Posts
On March 17 2010 02:54 sorech02 wrote: I'm guessing that since he pasted this together from two sources, they each had a different time limit, as well as other potential factors I'm not thinking of. The two sets can't be directly compared in this case. obviously you are right. it's 2 different time limits, but that doesn't matter as both show when you reach the saturation. | ||
|
da_head
Canada3350 Posts
zerg's always required fewer workers because they don't need to maynard workers unlike the other races | ||
|
Chairman Ray
United States11903 Posts
| ||
|
Tristan
Canada566 Posts
| ||
|
-orb-
United States5770 Posts
WTF? | ||
|
TheBB
Switzerland5133 Posts
On March 17 2010 03:04 da_head wrote: lol drones don't move faster on creep...have u guys even played sc1? zerg's always required fewer workers because they don't need to maynard workers unlike the other races The reference to SC1 is hardly relevant, and I'm pretty sure the reason why Zerg require fewer workers is wrong. This results from using the hatchery to make both drones and army units. First off, to increase production capacity, you need to build hatcheries, which you might as well do at expansions, and if you expand more, you need fewer workers. Secondly, you will have fewer workers because you need the larvae for other units. | ||
|
PizzaHash
Netherlands76 Posts
Though terran should be buffed with minerals maybe - their worker is not mining when building, but will be back when building finishes, so they lose mining time while protoss doesnt. | ||
|
lepape
Canada557 Posts
Also, for terran, when is a planetary fortress a better thing to build over the comsat thing? Is it only late game, when your new expos are already saturated from the start? | ||
|
xnub
Canada610 Posts
i Tho was somthing blizzard said right off the bat when they said creep would inc zerg speed | ||
|
da_head
Canada3350 Posts
On March 17 2010 04:37 TheBB wrote: I guess one of them was timed in-game for one minute, while the other was timed with the game clock in a replay. The reference to SC1 is hardly relevant, and I'm pretty sure the reason why Zerg require fewer workers is wrong. This results from using the hatchery to make both drones and army units. First off, to increase production capacity, you need to build hatcheries, which you might as well do at expansions, and if you expand more, you need fewer workers. Secondly, you will have fewer workers because you need the larvae for other units. how is it hardly relevant? you don't need to maynard workers because zerg can produce workers at a much higher rate than the other races due to having to make hatcheries at each expansion. yes there's also the issue of using larvae for attacking units, but that's obvious and i didn't feel like going into that. though perhaps it was necessary given the responses given in this thread... | ||
|
semantics
10040 Posts
On February 25 2010 01:08 Volshok wrote: I don't have Terran. Both of these numbers were taken from posters on this board, but I didn't grab the names. Sorry to the OPs, I copy-pasted these for friends on another forum. I did the zerg one I didn't use a replay to tell me the resources like the toss was did I recorded my game in that game i started with 32 drones on a feild waited 2 mins for workers to get into optimum area then recording how many mins i got in a min took away drones waited 1 min as workers should already be on a pretty optimum patch then recorded 1 min repeated it again and again did 18 and 19 drones twice to see consistency. anyways about 18-19 drones is about optimum cost per get back which seems to be true about the toss so we can assume is the same for terran. | ||
|
semantics
10040 Posts
On March 17 2010 05:01 xnub wrote: Drones do not move faster on creep .... i Tho was somthing blizzard said right off the bat when they said creep would inc zerg speed They don't which probably to keep the ability to scout zerg possible early game else drones would rape scouting workers if they ever got on creep. | ||
|
Exteray
United States1094 Posts
On March 17 2010 05:08 da_head wrote: how is it hardly relevant? you don't need to maynard workers because zerg can produce workers at a much higher rate than the other races due to having to make hatcheries at each expansion. yes there's also the issue of using larvae for attacking units, but that's obvious and i didn't feel like going into that. though perhaps it was necessary given the responses given in this thread... It's hardly relevant because you said "lol drones don't move faster on creep...have u guys even played sc1?".. and in sc1, nothing moves faster on creep, while in sc2, most zerg untis move faster on creep. So whether drones move faster on creep in sc1 has nothing to do with whether drones move faster on creep in sc2. | ||
|
Wintermute
United States427 Posts
Until then, people should calm down. Looks to me like 3 workers per patch is the absolute saturation point and around 2 1/2 is sort of a sweet spot. For zerg I'd certainly try to maintain at 2 1/2 per patch, so as I need to pull drones off to make stuff, I am still in the sweet spot. For terran and protoss I'd say it depends how soon I expect to get up another expo. | ||
|
danieldrsa
Brazil522 Posts
However i did only test playing as toss You should take on acount the possibility of losing workers, so i think 4 or 5 more doesnt hurt so much | ||
|
MorroW
Sweden3522 Posts
just put 2 drones on each mineral patch and start mining. then after about 10 seconds they will have a good rhythm and then u can exit after 5 seconds and load the replay to see avg recourses ill go test the 3 races myself cause i have a hard time believing they mine differently edit: ah actually that didnt work so well, it didnt want to show me recourses since i had no opponent xd is it true drone mine faster than scv?? | ||
|
semantics
10040 Posts
| ||
|
HydroZ
United States34 Posts
However, one might make the argument that Zerg should require less workers to reach saturation, given that they cannot produce both Drones and other units at the same time (while Protoss/Terran can produce Probes/SCV's at a constant rate). At the same time, holding Zerg to the same requirement for saturation seems to be consistent with their unit production system. Yes, you can't produce Drones constantly; but at the same time, you can produce other units 3, 4, 5 at a time, which the other races can't exactly do (at least not with one building). So the inability to constantly produce Drones is something Zerg gives up to be able to quickly produce any unit of their choice in large bundles. This would suggest that having unequal saturation numbers isn't fair (or the test was incorrect). | ||
|
semantics
10040 Posts
On March 17 2010 06:59 HydroZ wrote: Although I haven't seen the test that claims these are the correct rates, it doesn't intuitively make sense to me why one race would require less workers for saturation. It would make sense to standardize the collection rates so that no race gets an unfair advantage. However, one might make the argument that Zerg should require less workers to reach saturation, given that they cannot produce both Drones and other units at the same time (while Protoss/Terran can produce Probes/SCV's at a constant rate). At the same time, holding Zerg to the same requirement for saturation seems to be consistent with their unit production system. Yes, you can't produce Drones constantly; but at the same time, you can produce other units 3, 4, 5 at a time, which the other races can't exactly do (at least not with one building). So the inability to constantly produce Drones is something Zerg gives up to be able to quickly produce any unit of their choice in large bundles. in sc1 drones were the slowest miners i believe but it didn't matter as zerg expoed the most. in sc2 from what i've tested zerg and toss are pretty much the same enough to just be the same. | ||
|
ProoM
Lithuania1741 Posts
On February 24 2010 23:30 SubtleArt wrote: How does the worker saturation / diminishing returns system work in starcraft 2? In BW it was basically just more workers > faster mining although the rate of increase gradually went down. How does it compare to Starcraft 2? I heard with better AI its a lot more complicated, and in some instances having more workers might lower income? I don't know much about this so I'll let the more knowledgeable post their opinions Just a note, in BW SCVs mined faster than probes or drones ;]. Also, in BW perfect probe saturation was ~24. | ||
|
Khalleb
Canada1909 Posts
On March 17 2010 07:07 ProoM wrote: Just a note, in BW SCVs mined faster than probes or drones ;]. Also, in BW perfect probe saturation was ~24. the problem is this is the drone who mine faster and this is the zerg who expand the most so ya we have a problem here | ||
|
Zeke50100
United States2220 Posts
2 per patch isn't saturation unless every patch happened to be close enough to the HQ (which doesn't exist as of now). 3 per patch is probably average (in regards to max saturation for each map). It doesn't matter if any more than 2 per patch has diminishing returns percentage-wise; it still brings you closer to optimal. | ||
|
MeruFM
United States167 Posts
The workers mine the same speed. 16 is optimal. 2 per patch. 24 is saturation. 3 per patch. Adding 8 more workers gives about 20% more mineral. So 16 workers can already get 80% of the minerals at an expansion. | ||
|
Zeke50100
United States2220 Posts
On March 30 2010 09:54 MeruFM wrote: This thread has lots of misinformation. The workers mine the same speed. 16 is optimal. 2 per patch. 24 is saturation. 3 per patch. Adding 8 more workers gives about 20% more mineral. So 16 workers can already get 80% of the minerals at an expansion. Actually, Optimal can mean Saturation depending on the time period you observe the income for. For example, if you are comparing 1 base play into late-game with 16 workers to 1 base play into late-game with 24 workers, the 24 worker choice is optimal. I know this is an extreme example, but it's the easiest way of illustrating it ^_^ | ||
|
Tamerlane
Canada424 Posts
But the 24 = saturation of minerals is erroneous, some mineral spots (the closest to main obv) can be gathered at full time with only 2 workers on it. I'm fairly sure most maps have at least one spot for which this is the case, and a lot have 2 spots => that's why sometimes the real worker saturation comes at 22 workers... Although, I don't see how there could be a difference in income between 24 and 32 workers :o | ||
|
AndyJay
Australia833 Posts
Zerg 16 drones 925 17 drones 945 18 drones 990 -- 995 19 drones 1040 -- -1050 22 drones 1080 32 drones 1060 19-20 Drones is full saturation on 8 minerals The Zerg numbers are incorrect. Stop spreading misinformation. The actual Zerg numbers are very close, if not the same as the Protoss numbers. | ||
|
lololol
5198 Posts
The income from a saturated base is close to 800 minerals per minute(a replay timer minute to be exact). From my test it appears that optimal saturation is at 3 workers for every slightly distant mineral patch and 2 for every close one(so usually around 22, I tested just on the blistering sands left main). The close ones aren't actually mined with 100% efficiency, but if you add a third worker, it will just shuffle between patches every time it brings minerals back and barely improve mining efficiency(if at all). | ||
|
Steve496
United States60 Posts
| ||
|
SBelmont
United States122 Posts
The reasoning: The time it takes for a worker to deliver minerals and return to them is slightly longer than the time it takes for a worker to start mining and then start heading back. With 2 workers a patch, there will (almost) always be a slight delay in which the patch is not being mined. For the closest patches to your CC/Nexus/Hatchery, it only takes 2 workers to always have it being mined. Obviously this can also be noted that 18 workers is complete saturation at a gold mineral patch (3 x6). This isn't accounting for different in worker movement speed. I believe all workers have a "mine" time that is equal to each other, so with 3 per patch income should be even. | ||
|
us.insurgency
United States330 Posts
| ||
|
StorrZerg
United States13919 Posts
| ||
|
-orb-
United States5770 Posts
6: Zerg: ~248 Protoss: ~248 10: Zerg: ~419 Protoss: ~419 14: Zerg: ~590 Protoss: ~590 18: Zerg: ~705 Protoss: ~705 22: Zerg: ~762 Protoss: ~762 Here is the replay of the test game proving this: http://www.mediafire.com/?uyqjjdymmhy edit: also here's the ratios of going from step to stop going up 4 workers: 6>10 worker ratio: 1.6667 Mining ratio: 1.6895 10>14 worker ratio: 1.4 mining ratio: 1.4081 14>18 worker ratio: 1.2857 mining ratio: 1.1949 18>22 worker ratio: 1.2222 mining ratio: 1.0809 as you can see you're getting basically 1:1 bang for your buck up to 14 workers, but after that your payoff per worker added drops and it gets less and less efficient to add workers. | ||
|
Gescom
Canada3363 Posts
| ||
|
Zeke50100
United States2220 Posts
Until you hit 17 workers (that is, once you hit 17 workers, the remainder of this sentence is false), you have the exact same efficiency for every worker barring patch distances. The 17th worker, however, will have less efficiency due to not having a time slot it can perfectly fit into. @Orb: The reason your ratios/efficiencies are not accurate is because you only measured them at every 4th worker, starting at an arbitrary 6. If you measured efficiency for EVERY worker, you should find relatively maximum efficiencies for the first 16 workers (the variations will be due to patch distances), and then start seeing decreases with the 17th worker. By jumping from 14 to 18, you are assuming that each worker between there will have the same efficiency. Also, note that each worker has a walking speed of 2.8125. Every single one. There are also NO known differences in mining speed. | ||
|
_EmIL_
Sweden138 Posts
| ||
|
NiiPPLES
United Kingdom201 Posts
| ||
|
SichuanPanda
Canada1542 Posts
| ||
|
DrSmoke
United States175 Posts
On April 27 2010 02:48 SichuanPanda wrote: I've found 17 DRONES to be sufficiently saturated for Zerg. However, Protoss I've found 24 Probes provide full saturation, and 22 SCVs provide full saturation for Terran. The number of workers I put on gas depends how far the geyser is, the ones that are as close as possible to your CC/Nex/Hatch I used 3 workers on, while the ones that are one space further away I've found 4 can mine it without having a worker waiting outside for more than .3 seconds. 100% wrong | ||
|
agleed.agleed
Germany110 Posts
On April 27 2010 02:48 SichuanPanda wrote: I've found 17 DRONES to be sufficiently saturated for Zerg. However, Protoss I've found 24 Probes provide full saturation, and 22 SCVs provide full saturation for Terran. The number of workers I put on gas depends how far the geyser is, the ones that are as close as possible to your CC/Nex/Hatch I used 3 workers on, while the ones that are one space further away I've found 4 can mine it without having a worker waiting outside for more than .3 seconds. yeah sorry dude, never seen some wrong in one post. | ||
|
moapy
Australia5 Posts
Can I please have someone who actually knows what they are talking about tell me which one of the 50 responses to this thread is correct? My understanding is 16 is a nice number and 24 is max saturation? (this is not including 6 gas workers). Is this correct? | ||
|
Warp
United States166 Posts
The best example of saturation is watching HuK play.... because he ALMOST always stops probe production at 30. Way to revive the thread btw | ||
|
Shika
Sweden1711 Posts
2.5 workers per patch = 20 workers (not counting gas, naturally). You could add 4 more but the gain will be small and you are probably better of sending those to an expansion or using the larvae for attack units. | ||
|
Stropheum
United States1124 Posts
| ||
|
semantics
10040 Posts
On August 25 2010 08:09 Shika wrote: For zerg, this is what i've gathered to be quite effective. 2.5 workers per patch = 20 workers (not counting gas, naturally). You could add 4 more but the gain will be small and you are probably better of sending those to an expansion or using the larvae for attack units. This 20-22 gives around the max return for your value, then ofc 3 for each gas. Anything past that is usually used for maynard. | ||
|
heishe
Germany2284 Posts
any more drones bring absolutely minimal improvements, because the freelancers sometimes don't perfectly cover the far patches, but the difference will not make a difference in any game. the difference from 16 to 19-20 drones also isn't too big, about 3-5% if I reckon correctly. So when you set up two bases, maynard workers so that both bases have 16 on minerals before you fill both up to 20 | ||
|
Snowfield
1289 Posts
| ||
|
Najda
United States3765 Posts
On August 25 2010 09:38 Snowfield wrote: Man, there's even a ingame tip for this, it says 3 workers pr mineral field, but since zerg has creep, you can have slightly less. Drones have the same movement speed on and off the creep. Don't believe me? Go into a game and look at the drones movespeed when on the creep (2.81) then move it off the creep and check again, still 2.81 . The 2nd post of this thread gives all the information one could really need. There is no magic number though, because its always useful to keep making workers for a transfer. | ||
|
Santiago4ever
Sweden299 Posts
| ||
|
MythicalMage
1360 Posts
| ||
|
stepover12
United States175 Posts
I think the maximum number of mining workers you should have at 1 base is 30 (3 per mineral patch, 3 per gas.) That's because at any given time only 1 worker can gather at a mineral patch, and 3 workers per patch ensure that the mineral patch is constantly being mined . (Same thing with gas geyser.) So, if you have 1 base, aim for 30 workers, 2 fully mine-able bases - aim for 60 workers, and so on... There's a trade-off if you get to the rare situation of 4+ fully mine-able bases. Theoretically you'd need 120+ workers for fastest mining speed. But an 80 food army may not be able to cover and defend all your bases. Thus practically I'd say stop at around 100 workers no matter how many bases you have. | ||
|
DarkSeerTurbo
United States105 Posts
LISTEN: 1. All workers (probe, drone, scv) move the same regardless of creep 2. They all mine at the same pace 3. There is something called diminishing returns (econ principle). It basically means that for every extra worker you add on, you will get more money in absolute terms but less in a relative sense. Basically, this means that every extra worker you add will be LESS efficient. The threshold for diminishing returns starts after 2 workers per mineral chunk (ie after 16 workers, you will start hitting diminishing returns for every worker you make). 4. After 3 workers per mineral chunk (for a total of 24 workers), you will get zero return on your investment. Do not build more than 3 workers per mineral chunk. Conclusion: 16 workers is optimal (most efficient) for a mineral patch consisting of 8 nodes. 24 is the max that you can have. Anything between 17-24 will get you more money than 16 but at a lower efficiency rate. | ||
|
farseerdk
Canada504 Posts
On August 25 2010 10:49 DarkSeerTurbo wrote: wow guys, there is SO much bad information here! and the bad info just keeps on repeating! LISTEN: 1. All workers (probe, drone, scv) move the same regardless of creep 2. They all mine at the same pace 3. There is something called diminishing returns (econ principle). It basically means that for every extra worker you add on, you will get more money in absolute terms but less in a relative sense. Basically, this means that every extra worker you add will be LESS efficient. The threshold for diminishing returns starts after 2 workers per mineral chunk (ie after 16 workers, you will start hitting diminishing returns for every worker you make). 4. After 3 workers per mineral chunk (for a total of 24 workers), you will get zero return on your investment. Do not build more than 3 workers per mineral chunk. Conclusion: 16 workers is optimal (most efficient) for a mineral patch consisting of 8 nodes. 24 is the max that you can have. Anything between 17-24 will get you more money than 16 but at a lower efficiency rate. Almost right. This would be true if all mineral patches were exactly the same distance away from the nexus/cc/hatch. For the close patches, the optimal number is exactly 2. There is no wasted time (don't believe me? go and set up a mining pair on 1 close patch and watch). On the FAR patches, however, 3 is the max, but there will be some idle time. As such, the actual maxima for marginal returns would depend on the number of close vs. far patches, but the MAX it could be is 16. | ||
|
Keitzer
United States2509 Posts
On August 25 2010 09:55 MythicalMage wrote: According to the Blizzard tips, it's three per mineral patch/geyser. im pretty sure it's 6 for gas (3 per) and 2 per mineral patch (so 16).... this makes 22 the "optimal" assuming 3 per mineral is oversaturated i like to have about 25 at each in case there are a few distant mineral patches | ||
|
Comeh
United States18918 Posts
b: It is highly likely that you will lose workers to harassment / pulling to defend at some point c: Some workers will be sent off to build buildings, lowering the amount of workers from the "optimal" number at some points. However d: workers take supply away from your army, particularly important in the late game. overall, build more workers than you think you need. | ||
|
ctOph
Canada6 Posts
<15? Make more workers 16-23? You can make more if you like, but beware that anything after the 16th worker is subject to diminishing returns, and will have less minerals mined per minute than the first 16 workers. 24? You've reached the cap. Any more workers after this would be stupid, unless you have an expo about to pop and you can move some off. Any more workers after 24 on minerals will be either subject to very very harsh diminishing returns, or even not increase your minerals per minute AT ALL, since once you have too many, there will always be a worker wandering around for an open patch of minerals that it will never find. That's pretty much it. You will see an increase in minerals per minute by adding workers past 16, while still keeping them under 24. The only thing is, like I already said, any workers past 16 will be subject to diminishing returns, since they will be wandering part of the time looking for an open patch. When you have 16 workers, one will mine the patch, then return to the HQ, as its returning, a second worker will mine the same patch, by the time its done, the first worker will be returning, leaving a very short amount of time for a third worker to come in and mine that patch. That is why you should have at least 16, but no more than 24. edit: Also, I'd like to point out that if you expand, take off workers from your main to your expo, but don't bring your main down to less than 16 workers. Having a main w/ 16 workers, and an expo w/ 16 workers is better than having 24 at your main and 8 at your expo. Kind of common sense but I thought I'd point it out. If you take a third, you should probably only have 16 workers on minerals per base, because having full saturation (30 per base, 24 on min and 6 on gas) cuts into your food supply hard. Creating workers into diminishing returns while having more than 2 bases seems pretty illogical to me. Then again, if you need the minerals, go for it, but at that point in the game you might want to have a bigger army to defend your economy. /thread | ||
| ||
