|
I tested a small, specific part of this months ago, because i used bamboo nearly every game and I very oftne found myself in the situation where you have about 40 SCV and only a single mineral field to mine from.
Now, what I rnoted about those tests is that 40 SCV still mine (considerably) faster than 30 SCV, even if they are all on one mineral field and it looks like it's just overburdouned with workers . These numbers are exact, it was 40 workers vs 30 workers.
Now, let me ask you this: when will you ever have 40 scvs on each of your expansions? not too often huh?that's where the "never stop worker production" theory comes from
Now let me ask you this. If you have 4 mining bases and you want to mine with as close to optimal efficiency as possible (i say as possible because maybe 50 SCV mine even faster, I don't know), you need 160 SCV's. That leaves you 40 supply for units when you are maxed out. This look good? no?
"Never stop workers" <- bull shit
as to when you actually should stop workers, I quite simply don't know. However, my long TvP games vs BigBalls have convinced me that:
a) I build too many scvs late game
b) BigBals is closer to the right amount
does bigballs make too few or could he be better with even less? We don't know (yet)
|
is awesome32282 Posts
Protoss need to stop probe production unless caught in a weird situation or going carriers in a PvT.
once your income becomes pretty good (meaning you have shitloads of cash), usually when each player holds half the screen you should stop probe production or even suicide probes if theres not many bases to mine from (seen this in a lot of reps).
In PvT you can just keep throwing units at your adversary if you have the right amount of money and break him.
But i think the first question was more about scv per base early, mid game. As late game depends a lot in situations you are caught into.
|
If you ask me, it's going to be random everytime because the workers are always going to fly around to try to find an open place to mine, and sometimes another one gets there before it, so they have to fly over to another, etc.. in a perfect setting, two per patch would be fine, because it takes almost exactly the time to gather minerals as it does to take one back and return to the stack
|
It is one thing to know the saturation limit, but in respect to GroT's loss, it was clearly better to find the optimal value for highest efficiency of mineral collection. Obviously there are other variables to GroT's loss, but he clearly stated his SCV production was a factor. I read a lot of replies of 2.5-3worker per patch (an estimated of ~20-24 workers), so this is possibly a reasonable amount in term of efficiency. We're only considering 8 mineral patches condition by the way.
|
-
There've been a thread like this before, anyway here's my answer to this troubling question.
There should be 8 peons mining mineralers, 8 peons returning minerals and 8 peons that's going for minerals (3x8). That way you'll allways have a peon on every mineral patch, nonstop.
-
|
Are you suggesting a rate of 3peon/patch? This has already been posted.
|
On November 13 2004 21:42 LOcDowN wrote:
The asian guy @ UCI
Isn't everyone at UCI asian?
For whatever reason, as GroT mentioned, it seems that up to around 40 peons will keep getting you money faster, even though the numbers suggest that ~25 should be the absolute maximum.
|
On November 13 2004 20:24 This-is-not-a-smurf wrote:I meant for Zerg, one would never always make Drones early-mid game, as it would disrupt the balance of larvae, putting you at a loss. Also, gosu, the static defense is as Zerg right? I don't know of a single situation where you would even mass defense your main as any other race, and it's only as Zerg if you don't expo, or mass static defense at the nine choke. You're probably better than me, but you don't seem to know mch about Zerg 
nah i was talking about protoss and terran.
i hate playing zerg.
i can never balance lings and drones.
i was wondering.
what is the rate that larva appear.
i know by the time you make 1 drone another larva will appear almost all the time.
|
and as a protoss player i stop making probes after i have about half the screen.
i usually have enough probes for 2 places or maybe 3.
by the time the third one is up the first one should have been weakened a lot by now so i move a lot of probes out.
ONLY do i make more probes if i get on an island.
also notice that protoss and terrans units cost a lot of control whereas i have almost never seen a game where zerg was constantly at 200 control.
they often attack at like 100 or even 80.
lings can be massed without having too much control.
|
Bill307
Canada9103 Posts
On November 13 2004 21:39 Peatza wrote:
-
There've been a thread like this before, anyway here's my answer to this troubling question.
There should be 8 peons mining mineralers, 8 peons returning minerals and 8 peons that's going for minerals (3x8). That way you'll allways have a peon on every mineral patch, nonstop.
-
You're wrong. Workers waste a lot of time moving between different patches, especially if the formation is "bad" or if there are even more than 8 mineral patches. I tested the saturation point of an 8-mineral formation a few times and found that it occurred between 24 and 32 workers (or 3-4 workers per patch). One interesting observation that I made was that as the # of workers approached the saturation point, the workers became increasingly inefficient, spending more and more time moving from patch to patch and often moving in lines, which is probably the epitome of inefficient mining.
|
On November 13 2004 22:12 choader wrote:Isn't everyone at UCI asian?
It was supposed to be a joke since UCI consists a high % of asian.
|
2.5 workers per crystal chunk is a good rule to follow.
|
On November 13 2004 15:34 ares01256 wrote:never stop making miners = best i think all miners mine at the same speed 
i don't think its wise to make miners the entire game... i think
keep making miners for a certain period of time = best
|
T_T You guys don't know macro starcraft.
|
never stop building workers is true to some extent. most games are over before you should stop producing workers. (for toss and terran)
for zerg you stop early. coz u dont need to saturate each expo wiht drones.
i dont have exact numbers but from what i read/heard before i think 2.5 miners per patch is what its supposed to be, so for lt its about 20 per expo (pretty much eveyrone has more prolly)
personally i dont think there is a set number as it greatly depends on how miners move when they get to a patch thats being mined already.
|
Probes mine faster than scvs btw <- fact
i suggest not worrying about # of workers unless zerg is ur race....
'optimal' workers is just an idea for silly people who can't think well
if u want to tech and need the money, cut back a couple workers, if u have lotz of free time and u are safe, make more workers...
This is true macro.
|
The worker count per patch is truly less predictable in BW - since in W3 u have 100% efficiency with 5 peons mining gold, though the lumber peons are less predictable and NOT always the same number.
In BW the most effective number is around 2,5 per patch - this means u still get a significant increase in income per worker to this number. By the time u have 3 per patch u still get increase but it would require a longer time for a worker to pay himself - so the efficiency is lowering and while at the beginning to about 2 per patch your probes mine 100% time possible, by the time u reach 3 per patch it could be less than 75% time.
|
is awesome32282 Posts
I emphasize. No dont need workers non-stop in TvT. I have seen Boxer reps where he dooesnt get hararsed, yet he gets little worker to get his unit production to the maxium.
|
On November 13 2004 17:24 worst.player wrote:
Why would you ever stop making workers? More workers = more expos = more money = more gateways = more wins.
NTT once gave me that advice
|
LoCDowN there is no right answer. its all situational, like everything else in starcraft. if your asking for the number of SCV's required to mine the fastest out of like a 8 patch mineral field, its INFINITE, because just by random chance, if you have 200/200 SCV's mining it, give it a few years of leaving SC running, and im sure for one split second, all the workers looking for a patch will be on the wrong side, and you wont be gathering from that one patch.
So, there IS NO ANSWER to your question, just use common sense. you build SCV's when you think the money is best spent investing in one. Take into account things like how many SCV's you already have mining, how many bases you have, and if spending the 50 mins on something other than an SCV would benefit you more than spending on an SCV.
The only way to learn this is to play a good deal, and think about why you're doing it.
A better question than yours would be "How much faster does 20 SCV mine than 15 SCV on a 8 patch mineral field?" or something like that, so your gauging of whether the new SCV is a good investment is more accurate. But until you have the other basic knowledge that you get from playing and understanding the game, thats kinda worthless information.
|
|
|
|
|
|
EPT
