|
Looking for some different viewpoints, though I realize it varies based on map/race/other factors I'm likely not considering. Basically I'm taking a look at 'optimal' number of workers in the mid/late game without going over the top and costing yourself potential army size.
The specific time frame I'm looking at is about the time you take your 4th base (likely main is either completely mined out or getting quite close, both on mins and gas), because up to that point constant worker production likely won't cause issue with pop cap.
The thought I came to is ~65-70 workers is about as many as would be needed on a typical map, as you're most likely to be mining off either 1 or 2 8-mineral patch bases with the other 1-2 6 mineral patches (on some maps 7-min patch bases make an appearance, but I digress), as well as 6 vespene geysers at any given point in time.
So if we figure, in general, ~22 mineral patches, you get 'good' saturation at 44 workers, capped saturation at about 66, with 18 in gas. If we aim for 2 workers/patch 3/gas we would hit our goal with 62 workers (3/patch would be 84).
In any event, adding additional worker(s) for building/repairing/morphing is how I determined 65-70 workers should be sufficient for mining 3 bases at a time, and you'd be able to cut workers after reaching this point to devote all the available population to army.
I realize army comp/positioning is far more important than total size, but in a situation of a 200/200 confrontation, one with 65 workers vs 100 workers is obviously going to be a 35 pop difference in army size, which is substantial (bit of an extreme example).
What are your thoughts/opinions on when you stop producing workers so you don't cut into max army size?
|
Late late game if i find myself with more than 3 mining bases of workers i wind up having to kill a few off to keep up. This is generally when it goes above 80.
I'd say a safe amount is 2 and a half mining bases saturated, meaning somewhere around 70 workers it's time to stop.
|
On September 17 2010 08:21 Everhate wrote:
I realize army comp/positioning is far more important than total size, but in a situation of a 200/200 confrontation, one with 65 workers vs 100 workers is obviously going to be a 35 pop difference in army size, which is substantial (bit of an extreme example).
What are your thoughts/opinions on when you stop producing workers so you don't cut into max army size?
this might sound crazy, but in the situation you describe I would actually put my money on the smaller food-count army. The smaller army has more money, which likely means more upgrades and more unit producing structures, and more static defenses. With more expos/buildings to destroy etc. he can sustain much more economic damage than his opponent and still be able to bounce back. I also find that with a smaller army, I tend to micro a lot better, and I suspect other people are the same way.
further, if you have a stronger economy than your opponent and you beat him to the supply cap, why wouldn't you attack? you're just giving him time to catch up.
|
The ideal amount of workers is just enough to maintain your ideal level of unit production, teching, and upgrades.
Let's say I'm going mass muta/ling as zerg. I like to have 2 saturated bases and 2 bases mining gas, with half a dozen or so drones on minerals at each expo. That way, you have a ton of gas for mutas, and some excess minerals off your saturated expos for lings.
|
On September 17 2010 09:01 Zerokaiser wrote:
The ideal amount of workers is just enough to maintain your ideal level of unit production, teching, and upgrades.
Let's say I'm going mass muta/ling as zerg. I like to have 2 saturated bases and 2 bases mining gas, with half a dozen or so drones on minerals at each expo. That way, you have a ton of gas for mutas, and some excess minerals off your saturated expos for lings.
Certainly something I didn't consider enough, with zerg being limited not only by having to use drones for buildings, but also being capped on production with larvae. Some of the math on these things has been done already, but a lot of it is still left to 'feel' as everything is still fairly new.
If, for example, you were running 4 hatches on 3 bases, the needed numbers of drones would change, as with terran/toss as additional production structures were added...looks like I need to do a bit more math on some of the builds I'm working with to determine just how much income I need to get the production capacity I'm looking for on 2.5-3 bases worth of full production.
Has already given me quite a bit to consider in just a few replies.
|
It isn't under 100, especially on games with lots of harass instead of big head on head battles.
Just make sure you are upgrading a lot when it starts to get high.
|
If your macro and multitasking while being very agressive with main ball and harassments can keep up with 100 workers go for it, there is no reason to wait for 200/200 and then A-Move for such a player. The player with the strongest economy should use this economy to trade it against damage to the enemy army/building/economy.
A player with 100 workers should not wait for 200 to attack, he might lose more in his battles but he should be able to still be ahead if it is 100 workers against 60.
On a side note I would just say 60 to 80 workers look good to me. It depends on how much action the game as (no action until 200 => few workers). This is 2 to 3 mining bases.
|
certainly not below 80. depending on the matchup, the map and the playstyles it might be well above 100.
also keep in mind that the strength of an army starts to grow sublinearly with its size once a certain treshold troop count is passed. for example, on a passage with 10 tiles width, a marine only army would certainly not gain significantly if u increased its size from 60 to 70......
in general, all three races got access to splash dam in the later stages of the game, so that quality starts to beat mass once the corresponding tech is unlocked. to sustain a 70 pop hightech army and reproduce it nonstop can easily require 100 or more workers.
edit: there is one exception though: if it goes to the very late game and there are not many battles going on, it might be worth considering lowering the scv count and producing additional orbital fortresses to get the same eco with mules, freeing up the pop count of the skipped scvs.
|
I don't know why you wouldn't want 3 saturated bases. Factoring races in, as Zerg I always want my bases fully saturated, and then just continually pumping units every production cycle. Throw a wave of units to attack some expo somewhere and keep making more. Even if you were T or P, just throw some drops around the map, or if you are P and don't want to throw drops, just put down a pylon and warp some harass in or something.
The way I see it, if you want to keep producing units, don't kill workers to make room for food. If you kill a worker or two, you can only make one extra unit, in which case by that time your army is so big that the one extra unit won't be as beneficial as using a soldier to scout or something.
If your production gets halted because of food cap, just throw something at your opponent. Do damage to your opponent and you'll be more ahead than if you pumped out one more marauder.
|
I think the example of the 200/200 army is skewing what I'm looking for, and may have been a poor example for that reason. The real issue, at least for me, comes down to what point in the game you can consider your worker count sufficient given reasonable opportunities to continue to expand.
For example, it is unlikely you will be able to maintain 4 fully mining bases, though 3 is reasonable in many cases, and going over a certain worker count doesn't add additional income unless you add an additional base (actual count varies based on number of mineral patches and, to a smaller extent, distance of geysers from primary structure).
So, given the size of the maps, and total number of bases available, it seems reasonable that you reach a point (my thoughts were in the 65-70 worker range) that additional workers may not add additional income for the remainder of the game, due to limitations of available bases (population count is a sidenote and, as such, I shouldn't have brought up, though the replies to this point have steered my thinking that way, so thank you for that =D )
Basically, I have a general idea of where I consider myself 'comfortable' on worker count. Continuing to look for opinions on where you consider yourself comfortable.
|
Its Sort of a balance... if your constantly army trading/fighting your not really hitting 200 so having 60-70+ workers is good so your money jumps back so you can unit produce.
If its really late game and your both sort of doing a little harassing, both maxed at 200, and your money is piling up, it might be worth it to throw some workers away to get more units.
Most games end before i stop making probes - 15 mins or so(unless im cutting probes to do a timing push, or stop a timing push). 15min or longer, depending on how much they harass my min line/ i may stop producing if my eco is strong.
|
On September 17 2010 10:21 Everhate wrote:
I think the example of the 200/200 army is skewing what I'm looking for, and may have been a poor example for that reason. The real issue, at least for me, comes down to what point in the game you can consider your worker count sufficient given reasonable opportunities to continue to expand.
For example, it is unlikely you will be able to maintain 4 fully mining bases, though 3 is reasonable in many cases, and going over a certain worker count doesn't add additional income unless you add an additional base (actual count varies based on number of mineral patches and, to a smaller extent, distance of geysers from primary structure).
So, given the size of the maps, and total number of bases available, it seems reasonable that you reach a point (my thoughts were in the 65-70 worker range) that additional workers may not add additional income for the remainder of the game, due to limitations of available bases (population count is a sidenote and, as such, I shouldn't have brought up, though the replies to this point have steered my thinking that way, so thank you for that =D )
Basically, I have a general idea of where I consider myself 'comfortable' on worker count. Continuing to look for opinions on where you consider yourself comfortable.
these thoughts are correct imho. for a base with 8 mineral patches and 2 gas, all fully saturated, u need 24 + 6 = 30 workers. so thats 90 workers for 3 fully operational bases. so the general guideline of 90-120 seems about right, depending on if u can get 4 or only 3 bases mining at once. imho it is not toooo uncommon to get to 4 mining bases.
|
As zerg, I think anything over 60 is counterproductive.
Get 100+, stockpile 10000 minerals and 10 minutess of larvae off 5 hatches, then suicide all the drones and rebuild a near 200 supply crackling army. I actually did this against a turtling terran, and won. 
|
what?
8 patch 2 geyser base calls for 22 probes.16 for minerals, 6 for geyser. your natural should have 22 probes also. so 44 probes. you should be MAYNARDING your main's probes to your 3rd. you definitely dont make 22 more probes just to saturate that new expansion. you might make like 5 or 6 more probes at the 3rd, MAYBE. simply because when you take your 3rd, your main will be, at the very least, half-mined out. in most cases it will be 3/4 mined out or 2/3 mined out, so you wont need 22 probes there anymore, but maybe just 14. therefore you maynard your main's probes to your 3rd, and your natural's probes to your 4th.
plus, 100 probes is overkill. you could have 50 probes and 25 extra HTs+ 25 extra zeals. seriously, in the late game every single bit of supply matters, and you want as many of these cost-efficient units as you can get. 100-supply army wont cut it.
i dont see how 100 probes could possibly benefit you more than 50 would.
(this doesnt apply for zerg)
|
As a Zerg player, I aim for 60-70 workers to effectively mine 3 bases at a time.
|
|
|
assuming game wasnt a fast game, then 90. fully saturate 3 bases
|
i dont see how 100 probes could possibly benefit you more than 50 would
The more resources you bank the better. When you are constantly base trading you are going to lose big chunks of probes at once so they work as a buffer. At 50 probes you are not going to instantly replenish to 200/200 so the supply does not become a big deal unless you are specifically being passive.
75 probes will always be better than 50 as long as you can effectively mine with them, pretty much no matter what the game situation.
|
I'd say it depends on your race. If you're Terran, then you can really use more SCVs than it takes to saturate however many bases you intend to have mining at any given time, as they are nice to have along with non-bio units to repair. As Zerg, you might want to over-saturate slightly to accommodate more buildings if/when you need them. As Protoss, you really don't need any extra as you can't do anything but build and mine with them, and 1 Probe can build everything you need if its within a practical radius of your base.
It also depends on how much you intend to expand. If you're gonna wait till your main is nearly mined out before getting an expo, then you don't need to over-saturate by very much at all since many of them will have nothing to do when the minerals run out.
|
On September 17 2010 13:54 EnderCN wrote:+ Show Spoiler +i dont see how 100 probes could possibly benefit you more than 50 would The more resources you bank the better. When you are constantly base trading you are going to lose big chunks of probes at once so they work as a buffer. At 50 probes you are not going to instantly replenish to 200/200 so the supply does not become a big deal unless you are specifically being passive. 75 probes will always be better than 50 as long as you can effectively mine with them, pretty much no matter what the game situation.
i think we can agree that 75 probes will not be efficient on 2 bases, but will be on 3 (actually 66 is efficient but w/e). so the issue is, when do you take your 3rd? you have two options: to take it really early (meaning not long after you take your natural), or to take it later, when you have a decent army. the problem with taking it early is that in order for your 3rd (and your nat) to be effective you need to pump probes from all 3 nexuses (even if you maynard you still have to do this), and this cuts your army size drastically (because you're spending 150 min every 17 seconds) to the extent that bio balls will kill it, 4 gate will rape it, and zerg will run it over. taking a 3rd later would be a better option because you will actually have an army to defend it, and also that you wont have to make more probes - you only need to maynard from your main, which should be close to mined out.
and late game, after you take your 3rd.. you shouldnt even be making probes. every bit of supply should go into your ARMY. unless you want to get run over by your opponents' army. it doesnt matter if you can reinforce your army faster than your opponent, because he can destroy your army, run over to your natural, and destroy your reinforcements while training units of his own.
i think we are looking at this the wrong way, for we are emphasizing "saturation of bases" over the timing attacks of the game. saturation and probe production are important, yes, but when we factor the timings in, we have to think differently about what we prioritize.
|
I had a think about this the other day because I found it maddening to hear "build more workers" as an honest to god rule to live by and I wanted to fidn the logic behind it. There must be a cut-off somewhere, right?
My thinking goes - you want to balance your production capacity against your income up until the point where either your income hampers maximum army size (200/200 with a strong worker skew) and your production capacity meets the needs of your maximum army size (surplus income spent on production facilities)
At that point, you can ask which of the following is a better "maximum army size":
- 4 bases/88 workers/112 food army which can be replaced as it is chewed up
- 3 bases/66 workers/134 food army with a short delay on replacing the lot
- 2 bases/44 workers/156 food army with a long delay if it's wiped
If you have plenty of banked resources, there's nothing stopping you from going 44/156 and simply having the production capacity and resources to replace that army if it gets wiped. It's just not going to be sustainable if your army is getting wiped all the time.
But I really don't have anywhere near enough experience to state what a good balance is. All this tells me is it's likely that 4/88/112 would be a pretty high ratio, used if you're sending wave after wave against the enemy, while 2/44/156 would be a lower gear suited for strong defense with harassment.
|
for one base, 30 worker is more than enough. And remember, you switch your worker to new expo too. So more than 80 ~ 90 is not necessary (until you are so good at multitasking that you spend your 90 support worth army wisely and rebuild them in second with crazy macro)
imo, checking the amount of worker during the game is not so bad idea. e.g. when you just made 2nd expo.
|
I agree, endgame zerg has some mobility, best way to free up food space is to attack an expo or something, not turn workers into static D, as long as you have enough hatches to resupply an army...
|
simple answer, there is no "too many".
As Fugue noticed a 4 base is saturated at 88 workers, which is very unlikely to happen. On that stage in the game it is very common to get some harass or expos getting attacked - so - you will loose some workers anyway.
This game is about economy and you win the game when your enemy is out of ressources and does not have any buildings (or in most cases will get all his buildings destroyed so he surrenders). The more bases, the more workers and - most important - the more ressources the map provides belong to you - the bigger your advantage actually is.
If you look at - for instance - at the Zerg larva: they can replace an 80 supply army within seconds, two, three or even more times - if they have the ressources to do so. What do you think happens to a 140 food army when it has to face an 80 food army three times in a row?
|
Just a thought, but I use Insane AI builds as a rough guide to what to do when I don't know (or cbb working out) the exact perfect mechanics.
So for workers, in their builds, I think Insane AIs go for a maximum of 24-ish workers per base.
I could be wrong, would be happy to have someone double check that. But if its max 24 per base, then you can work out from there what your max would be from 2 bases (48) or 3 bases (72).
|
|
|
|
|
|
EPT
