|
I've seen a lot of questions posted here on zerg econ management. How many drones do I need? What do I do with my excess minerals? How come my opponent can produce so much shit with so few workers?
I used to build as many drones as I can get away with. But when I hit B-, my opponents started to out macro me consistently. Several times I've been told by my opponents that I make too many drones, which was really confusing to me. But the idea is simple, if you make more workers than you need, eventually you end up having lots of excess minerals in the bank for no good purposes, but in the mean time you waste valuable resources and time making workers, and your army is smaller than it should be.
Recently, I played a good Korean zerg who later taught me the 9 pool build like I've never seen it before, and what amazed me was how calculated and exact his econ management was, perfect, not a single drone more than he needs, not a single drone less. That exactness of Koreans in their econ management is precisely why they can pump out so many more units. So, afterward I did an experiment with all zerg units and figured out the amount of income needed to maintain constant production of each unit. And I will put all my findings into this zerg econ guide to share it with others. Be reminded that my max rank was B- a couple seasons ago, so I'm not gosu or anything, just hope this helps.
How many drones do I need to maintain constant production of a ..... from 1 hatchery?
Drone: 3 drones mining minerals
Zergling: 3 drones mining minerals (duh, it costs 50 minerals just like the drone)
Overlord: this unit is unique since you don't constantly produce it....you just need to start a new one when the old one pops out. 2 drones per overlord is good, if later in the game you need to make 2 overlords every round, then 4 drones, no more will be needed, ever.
Hydralisk: 6 drones on minerals + 1 drone on gas ( remember, we are talking about 1 hathcery)
Mutalisk: 4 drones on minerals + 3 drones on gas. It should be noted that this combo gives you the same amount of minerals as gas, 100/100 for muta, and the same combo is used for lurker and ultralisk.
Note, in zvz for muta/ling combo, it's 6 drones on minerals + 3 drones on gas per hatchery, mainly because mutas are limited by gas, not larva, so you will have excess larva and no gas for the zergling.
Scourge: 1 drone on mineral + 3 drones on gas. This unit is only constantly produced in the ling/scourge strategy in zvz. And for that 3 drones on minerals + 3 drones on gas should maintain constant production of ling/scourge from 1 hatchery.
Queen: Omitted from experiment due to lack of use.
Lurker: 4 drones on minerals + 3 drones on gas. This includes the hydralisk cost.
Defiler: 3 drones on minerals + 3 drones on gas, you get a little excess minerals, but 2 drones on minerals is not enough.
Ultralisk: 4 drones on minerals + 3 drones on gas.
How to make use of this information strategically?
Nowadays every game looks like one you have seen many times before, and the reason is because the Korean exactness is taking over the game. Every build they use has a hidden econ side that not everyone notices, but if you look into it you will find that every build ultimately yields the exact amount of workers needed for the strategy of that build. I will do a complete build analysis later. But here let's consider some simple cases:
I want to do mass hydra so how many drones do I need?
Well, to maintain constantly production of hydras from 1 hatch, you need 6 drones on minerals + 1 drone on gas (see section above). So it really depends on how many hatches you want to have in your particular strategy.
If you are doing a 2 hatch hydra break, once the upgrades are done, you can take off a drone from the gas, and have 14 on minerals + 2 on gas. (6 per hatch for hydra + 2 for overlord) If you are doing a 5 hatch mass hydra, then you need 5*6+2=32 drones on minerals and 5 drones on gas. Calculating this way gets you the exact amount of drones you need, and then you can maximize your army size.
It's zvz and I never know when to make drones and when to make units?
Simple, if you want to do ling/scourge, multiply the amount of hatcheries you want by the drone combo, 3 drones on minerals + 3 on gas. So if you are doing 1 base ling/scourge with 2 hatcheries, you need 6 drones on minerals and 3 on gas.
If you have 3 hatcheries and 2 bases, and you want to do muta/ling, you will need 6+6+3=15 drones on minerals and 6 drones on gas. Where is the overlord? Well, due to the intense nature of zvz, you can always skip a couple of lings to make an overlord, the only thing that matters is the amount of mutas you have anyway. But if you think your opponent won't be aggressive and you want steady growth, by all means add 2 drones for that overlord.
The same calculation can be done for any unit combo. So, if you woke up one morning and a nice strategy comes to mind, you can now easily calculate how many drones you need to make that strategy happen!
Case Study - zvt 3 hatch muta build analysis
Let's first calculate how many drones you need for this strategy. We know that the muta contain is only a transition phase into lurker/ling, then when hive comes, you expo and add defiler into the army, and when you secure more gas, and your carapace up is looking good, switch to ultra/ling.
Too much depends on the actual game so we won't do this match up to the end game, but unless somethings goes horribly wrong with your muta harass, you should enter a 5 hatch lurker/ling phase without any problem. Since lurker production is limited by gas, not larva, with 3 gas you can keep 3 hatches producing lurkers, which equals to 4*3=12 drones on minerals + 3*3=9 drones on gas; now you have no more gas, but you want to keep 5 hatches pumping lings (again, lurker production will yield excess larva for the hatchery since it's limited only by gas), that means 3*5=15 additional drones. Don't forget the overlord, that's another 2 drones on minerals. Let's sum things up, we have 12+15+2=29 drones on minerals, plus the 9 drones on gas for a grand total of 38 drones. But wait! you want to expo again sometimes, right? Put another 3 drones on minerals to afford your 4th expo, and when the expo is up, add more drones depending on whatever strategy you want to use for the end game.
So, that is our analysis. Do the pros actually do that? Let's work through a typical 3 hatch muta build. (There are billions of pro replays out there that feature the exact same 3 hatch muta build, you can find a rep yourself...)
12 hatch
11 pool
14 hatch
16 gas
25 gas
Spire when lair finishes
33 hatch (note that if you made 3 sunkens you will not be able to afford this, cuz you spent it on 2 of your sunkens...)
Shoot up to 35 supply, then make 3 overlords. Overlords finish at the same time as your spire, and you should have 900/900 to pop the mutas, yeah yeah everyone knows...
Assume that you lost no units for the sake of easy calculation. After 11 mutas your supply should be at 35+22=57. Subtract the mutas and 12 zerglings you should have 29 drones, 6 of which are on gas, so 23 on minerals. Looking at our goal of 29 drones on minerals + 9 on gas, are we only 9 drones short? Not quite, remember that a drone is lost for every building you make. At this point we need a hydralisk den, an evolution chamber, another geyser, another hatchery, plus a queens nest, and in all likelihood another 2 sunkens, so that is 7 drones lost that we have to replace, plus the 9 we needed earlier for a total of 16 drones that you need to make.
Coming back to the build analysis, at 57 supply where we left off you are gonna use 6 drones on buildings (queen's nest is usually gotten at the same time as lurkers, so that's after you make your hydras), that puts your supply back to 51, and after making 13 drones while you are muta harassing, at 64 supply it is time to make hydras and go lurker/ling like we wanted. Later, you can add a 3 more drones to meet your production requirement, and make a few additional drones for more buildings and save money to expo again. In fact, a lot of progamers go to 68 supply before they make hydras, but understand that it requires gosu muta micro to contain the terran for your delayed lurkers, and I don't recommend it unless you are feeling good about your muta harass, much easier is to make lurkers first, and then add a few drones. The econ boost from a few earlier drones is significant at the pro level, but not to most of us.
OK, that concludes our zerg econ guide, I've given you the drones needed for each unit, walked you through some simple calculations so you can use the numbers for yourself, and I've analyzed the econ side of the popular 3 hatch muta build as a case study. The rest is up to you. I hope I can improve myself with this new insight, and I hope you guys can benefit from it as well. GGs
Bonus Terran Section:
This might look counter-intuitive, aren't you helping the zergs you might ask? Well, I'm trying to help the whole community, it's just that I play zerg and I'm not familiar with the other races' econ management. But I've done some work on the Terrans as well.
Marine: 3 scvs on minerals
Bat/Medic: 3 scvs on minerals + 1 scv on gas
Vulture: 4 scvs on minerals
Tank: 5 scvs on minerals + 2 scvs on gas (Note that in terran production the unit's production time is the time you have to wait before making the next one, unlike the zerg where all larva spawn equally fast. So, theoretically, 6 scvs on minerals + 3 scvs on gas will get you the same amount, but your tank won't finish in the factory so you can't make the next one right away, that leads to excess money, which means you wasted time on extra scvs, and your army is not as big as it could be.)
Science Vessel: (god I hate this unit...irradiate is so imba) 2 scvs on minerals + 3 scvs on gas.
Goliath: Omitted due to laziness, hey I'm not responsible for you Terrans.
Battle Cruiser/ wraith/ valkyrie: Omitted due to laziness.
I do want to note however, that zerg has the highest mining efficiency due to it's expansion tendencies, and mining efficiency doesn't really affect zerg. But a terran will experience diminishing mining efficiency, so when you do the calculations, make sure you add a few scvs to balance out the efficiency factor. I know pros make about 4-5 more scvs for that purpose.
So if you are the terran who always has 3000 minerals in the bank and you never have big enough of an army, always miss your attack timing due to army size, it is likely that you are making too many scvs. Do the calculations yourself to see if that is the case. If not, you need to practice keeping all your barracks flashing all the time...that's not what this guide is for. Peace
|
Cool guide, thanks smi.lols
it's a bit too specific and I can only see some limited use as people won't count everything so exactly but it could give a general feel of what people need  good testing
|
Amazing... Its way to indepth. :D
|
Damn, so many mistakes, I think I've corrected most of them at this point. If you find any more mistakes either in calculation or interpretation. Let me know and I'll correct it.
Wolf, Iike Artosis said in his guide "The Difference between Koreans and Foreigners", he mentioned that Koreans practice the same thing over and over again. And that is the point, it's not realistic to calculate everything in game, but if you practice it over and over again, the numbers will become natural to you, and you can get a good estimate in any situation by intuition alone based on experience.
In fact, in both Savior and JD's FPVODs, I noticed that both of them sometimes selected all their drones at each base, it can only mean that they are counting and calculating their next step.
|
This is really useful information, but I'm a little confused - If the lurker cost includes the hydralisk cost, how can it be only 4 drones on minerals, when hydralisk alone is 6? An ultralisk is 200/200, so how can it be supported with 4 on minerals and 3 on gas, when a mutalisk is 100/100 and needs 6/3? Something isn't adding up right here.
Anyways, thanks for the post. I've seen many references o Savior's "management style", but was always a bit confused on what exactly it meant. I used to not pump enough drones early on, and now I sometimes find myself with too many, so this guide will definitely be useful.
|
This helps me insanely much. Thank you.
@ ^ Good points though. It would be cool if you made that clear.
|
The build time is longer for ultras/lurkers.
|
Wow, nice. Thanks for all the work you put into this, must've taken forever lol. I'll try and keep these numbers in mind and test it out to see if it helps at all.
|
Underwhelmed:
You probably read the thing before it was fully corrected, I added some notes.
For hydras, 6 + 1 gives you constant production, meaning every time a new larva comes out you make another hydra.
For lurkers, you are limited by gas, you can't make the next hydra before you get enough gas for it, and the larva spawning rate is faster. So if you have 6 + 3, after you morph your first lurker, you will have an idle larva but not enough gas for the next hydralisk, overall you will end up having excess minerals. 4 + 3 gives you the constant production of lurkers based on gas rate. And because you have excess larva, if you add 3 drones you can make zerglings, so check that section again, the total drone count should work out.
For mutas, it's like the lurkers, 4 + 3 gives you the money for the muta, but you can't make the next one before you collect 100 gas again, the idle larva that spawned faster should be made into a zergling, and it turns out that 6 + 3 can keep constant muta/ling production from 1 hatch.
The number for unit combos are different from pure units because you want to take the gas mining rate limit into account and find a way to use the idle larvas. Sorry for the confusion.
Nevake:
You are right. But again these units are limited by gas, how fast you mine gas decides how fast you can produce them. The larva regeneration rate is much faster than all of them, so you add drones to make zerglings with the idle larva.
|
Be reminded that my max rank was B- a couple seasons ago, so I'm not gosu or anything, just hope this helps.
psh gosu 
|
good information to know!
one thing I would say is I think your argument for not making too many drones is a little off the way I see it, there is no such thing as too many drones. If you have too much you make more hatcheries and macro more. The issue is knowing when its 'safe' to pump up your econ and when you need to be producing units.
One of the hardest parts of playing zerg as far as I am concerned is knowing when you should be producing drones/when you should be producing units due to the lack of a main worker production building.
|
I think this could be a little more useful if you consistently went with larvae spawn rate as the limit rather than gas rate. e.g. if you're playing on a map with 2+ geysers, or just have a gas expansion. e.g. if i knew i wanted to do muta ling off 2 bases with 1 hatchery worth of mutas and 1 hatchery worth of zerglings, how many drones should i have on gas between the two bases?
|
dream-_-:
It has to do with timing. Of course having more drones now will yield more money later, but more importantly you want to have a maximized army at all times, so having the exact number of workers at any given time allows you to have the largest army at that time, as nothing is "wasted" on excess workers.
Also, all timing attacks absolutely depends on the exactness of workers since you only have one shot, and you really want the largest army you can possibly make for that timing window.
Polemarch:
I think gas rate is more realistic as additional gas requires more expos, and expos are not given to you without a fight, so realistically gas is the main limiting factor on unit production.
|
On September 17 2008 11:26 w3jjjj wrote:
So if you are the terran who always has 3000 minerals in the bank and you never have big enough of an army, always miss your attack timing due to army size, it is likely that you are making too many scvs. Do the calculations yourself to see if that is the case. If not, you need to practice keeping all your barracks flashing all the time...that's not what this guide is for. Peace
lol This paragraph is the only one that I felt was informative to me but thanks for the whole guide in general as well. The last paragraph describes my play 100% and I really need to work on it.
|
On September 17 2008 12:47 w3jjjj wrote:
dream-_-:
It has to do with timing. Of course having more drones now will yield more money later, but more importantly you want to have a maximized army at all times, so having the exact number of workers at any given time allows you to have the largest army at that time, as nothing is "wasted" on excess workers.
Also, all timing attacks absolutely depends on the exactness of workers since you only have one shot, and you really want the largest army you can possibly make for that timing window.
Polemarch:
I think gas rate is more realistic as additional gas requires more expos, and expos are not given to you without a fight, so realistically gas is the main limiting factor on unit production.
this is true, guess i didnt read enough to see you were talking about a timing rush
|
Just watched a Jaedong ZvP replay where he had like maybe 2 or 3 more drones than the number of mineral patches at each expansion, he had mass hatcheries and so many units. I think this is what's missing from foreigner ZvP, they make too many drones and they get run over by midgame push and/or they let Protoss take too many expansions. I'm going to start experimenting and focusing more on this aspect from now on.
Dimaga, the new foreigner zerg monster, seems to have a good concept of this judging by his recent replays.
|
Holy crap, this is really really indepth and really useful to know. :O
That Overlord timing, if correct, just solved like 90% of my problems with Zerg, knowing when to build Overlords, I always either build too much or too little at the worst possible times T_T
|
On September 17 2008 12:54 dream-_- wrote:Show nested quote +On September 17 2008 12:47 w3jjjj wrote:
dream-_-:
It has to do with timing. Of course having more drones now will yield more money later, but more importantly you want to have a maximized army at all times, so having the exact number of workers at any given time allows you to have the largest army at that time, as nothing is "wasted" on excess workers.
Also, all timing attacks absolutely depends on the exactness of workers since you only have one shot, and you really want the largest army you can possibly make for that timing window.
Polemarch:
I think gas rate is more realistic as additional gas requires more expos, and expos are not given to you without a fight, so realistically gas is the main limiting factor on unit production. this is true, guess i didnt read enough to see you were talking about a timing rush
He's not just talking about a timing "rush", in a sense you can see the entire ZvP game as a game of macro/army timings.
Btw thanks a lot sMi.lols for sharing, this is some next-level shit for me.
|
its too... mathematical. i think its more tuition than anything when youre making drones / army.
|
How much is that tuition?
It may seem too mathematical but that's because of how complex Zerg macro really is. If you rely on just intuition, you will never be able to maximize both your economy and your army size with the same proficiency that Pros have, and that's what separates them. I have no doubt that things like this are running through Jaedong's mind constantly when he's playing ZvP, which may also explain why it took him so long to get so godly at ZvP: mechanics alone won't help with this management ability.
|
Wow this is really interesting and in-depth.
I wish someone would make one for protoss units lolololol
|
On September 17 2008 14:38 -orb- wrote:
Wow this is really interesting and in-depth.
I wish someone would make one for protoss units lolololol
yeah lets give the ez race more help
|
omfg they stole ur bucket
but in all seriousness
"you want your army to be maxed at all times" -smi.lols
why? i find several instances where the other player plays too defensively or safe ; to not make a huge army; like when i deny scouting; i usually opt for a greedy econ or tech build
i'm not trying to insult your logic or anything i just don't understand timing and "Exact" build orders too well and hoping for some enlightenment
like for example... you have like the best.. "exact" 3hatch muta build for pretty much any 1rax-> fe terran BO,
but u notice terran made only 2rax and is teching super super hard, so why not dual expo, mass drone, double hydra ups and hydralurker that up;
as opposed to making a ~ a dozen speedlings for the anticipated break on ur 3rd gas
The econ boost from a few earlier drones is significant at the pro level, but not to most of us.
i like having more drones 
|
On September 17 2008 15:04 alphafuzard wrote:Show nested quote +On September 17 2008 14:38 -orb- wrote:
Wow this is really interesting and in-depth.
I wish someone would make one for protoss units lolololol yeah lets give the ez race more help
There is no such thing as the "easy" race it's just it's easier to start a protoss and get decent but most Protoss will cap at C iCCup level
|
damn this hits the nail on the head.
i've been wondering about the ratio, when to drone, timing too.
keep up the good work
|
Great, this information seems to be very usefull if you ZvZ or if you plan to go for Hydra-all-in ZvP or 3hatch-speedlings-all-in ZvT...
All the builds that can end a game quickly and where every single unit counts when you attack.
I am Terran but I play Z from time to time and will definitly test some of this builds with your information
|
I did the same experiment a few years ago but never found any real use of it as you have to modify your drone count to afford all these buildings.
|
HeavenOnEarth:
First, if your army is maximized and your opponent's is not, why not just roll him over with superior numbers? If he is playing passive in his base, just match his production speed and keep pumping units. Theoretically, if you have the same production speed and you have more units to begin with, he'll never be able to catch up to you.
I didn't say 3 hatch muta is good or bad, I just used it as a case study to reveal the hidden econ side in popular builds. If you want to use a different strategy, it would be good if you understand how much econ you need to execute your strategy. The numbers are intended to be used for strategic planning.
Zerg-Legend:
I'm still trying to find a way to use this myself lol. Right now I'm focused on calculating production speed. I'm experimenting on the other two races to learn their numbers, and when I do I should be able to tell exactly how many workers my opponent has by scouting his production facilities, and that should tell me how many drones I need to match his production. The idea here is to cut drones just slightly above the number that I need, so I can have the same production speed but keep a small econ advantage. In theory, everything else being equal, that small advantage will grow to become decisive, well, in theory. I'm testing this strategy.
|
Theoretically, if you have the same production speed and you have more units to begin with, he'll never be able to catch up to you.
So theoretically, all games are decided in the first 5mins, and whoever falls behind first , BO and econ management wise.
|
However, that's rarely the case as any good player can pull ahead ; but evidently if we stick to this formula? ;
i mean sure u might ever so slighty ahead in economy, but that won't make up for the fact you're going pure hydra( although perfectly macro'd ) vs a plethora of storms and zealots ! you would probably want to switch to lurker ling... meaning more drones less hydra for the time being, and you wouldn't be able to ... keep the maximum army requirements at that particular time ( since you are switching from hydra to lurker ling)
Several times I've been told by my opponents that I make too many drones, which was really confusing to me. But the idea is simple, if you make more workers than you need, eventually you end up having lots of excess minerals in the bank for no good purposes,
lurkerling > delayed +2 temp push > hydras > 7min zeal >newbie zergs
Let's assume that^ for the moment... and let's assume you want the maximum army, just in case of 7min zeal , yet still have enough production, to match delayed +2 temp push for lurker ling
Ok, i'm starting to grasp this idea, but is that possible?
To go .. "low econ" hydra and switch to "high econ" lurker ling , you would have to stop... somewhere and pump drones; therefore you would have a a timing where his army> yours
BUT ; how do we avoid that? i don't think with even the most precise of precision macro drone management you could do that v/s a veterean toss
.... And therefore you mix economy... with tech... and it's literrally impossible to keep a scout alive to calculate this into the ground as the game progresses that far... so how can you be so sure what you're doing is actually putting you ahead?
like for example, zvz is simple (assume 9pool vs 9pool) ; you make 1 more drone than he does, and the timing from his base to your base , makes the zergling count equal.
Therefore you pull ahead and outnumber his mutas etc and win
BUT in ZvZ you generally have a good idea of what he's doing, either you have speedlings running around or an overlord in his base; otherwise you wouldn't generally risk making extra drones
BUT NONE OF THAT MATTERS if you are facing a completely different bo, from your own, IE overpool into mutas... vs 12hatch in main turtle, into more mutaling, you have to adjust completely different
or maybe 9pool ( you made like 10lings , or something) vs 12pool sunk, ( no ling) ... way behind in economy , no way ur going to make enough drones, so you have to rely on ling pressure force more sunkens lings than need and those faster mutas to do damage and micro the muta/scourges? therefore taking the tech route
like you said
to reveal the hidden econ side in popular builds.
it works in the popular FE vs FE strategies, but not for all cases
|
Pretty awesome guide I must say, but you should probably write in it somewhere that it only works against a Terran going for 3 rax --> fac (the standard of course). Against 4 rax you need to get lurkers much faster, same with 2/1 rax --> fac, where you also need hive faster to counter their tank/vessel push.
Correct me if I'm wrong, but I've noticed this in pro replays and when I'm playing.
EDIT: If Mods read this, this topic really needs to get stickied and put in the recomended guides section.
|
On September 17 2008 11:32 Superiorwolf wrote:Cool guide, thanks smi.lols it's a bit too specific and I can only see some limited use as people won't count everything so exactly but it could give a general feel of what people need good testing
I actually believe it will prove out very usefull for those who are perfectionists and have the patience to think so indepth in advance of their games.
Regarding economy, with this model you can now DO calculations yourself prior to the game, and depending on map, to establish exactly how your build order should go, and other details specifically economy related (workers and unit training facilities)
All in all, this is great, thank you very much OP.
|
I used this has a rough guide for a new zvt i'm workin on and it improved how fluid I went from 3 hat muta to 5 hat lurkerling ... I wasn't exact, but it definately brought a clarity to my econ management.
|
Vatican City State491 Posts
I find it strange that 6 drones are needed to maintain hydras and only 4 for mutalisks :O
There used to be a program used to create build orders via the windows menu - it calculated the gas/minerals input and provided information on time (the only drawback was the fact that it did not include the timing lost via transfer/scouting/or when your opponent would kill your drones). I forgot it name and cant find it; I found something called "evolution forge 1.62" but it's not the one I remember
|
This is def how zerg should play, only as a many drones as needed. Way too many zergs overdrone and they are left with too few units at key times because they simply dont get it
On September 17 2008 22:23 closed wrote:
I find it strange that 6 drones are needed to maintain hydras and only 4 for mutalisks :O
There used to be a program used to create build orders via the windows menu - it calculated the gas/minerals input and provided information on time (the only drawback was the fact that it did not include the timing lost via transfer/scouting/or when your opponent would kill your drones). I forgot it name and cant find it; I found something called "evolution forge 1.62" but it's not the one I remember
You're limited by gas in the case of mutas whereas he stated you only need 2 drones per 50 mineral things (drone/ling) so 4 per 100 mineral thing makes sense ;P
|
I think this most of all would really help me in my 2v2 game. because every worker unit counts. i always wonder how pro's use only 4 workers and still make massive armies!
|
To all who promotes different zvt strategies. I apologize for the confusion, but this is not a strategy guide. I only used 3 hatch muta as a case study to show how the build eventually gets you just the right number of drones for 5 hatch lurker/ling. All other builds work the same way, if you follow through the build perfectly you end up having just the right number of drones to execute that strategy.
Some people are arguing for adaptation in game, well that's precisely what this guide is for. If your opponent does strange things, you want to adapt to it but you won't have a build order laid out for that particular case, this is where you run a short calculation in your head and design your own build order.
For timing and army mix arguments. Every strategy has it's own timing. So depends on which build order you are using, your timing will be different in each case. For zvp everything is reactive, so the strategy you choose depends on your opponent's strategy. If you made mass hydra vs storms, that's a bad strategic choice for you, it has nothing to do with econ management, you should have scouted earlier and went for a different econ build up to support a different army mix.
|
I make too many drones.
:O so specific though; does anyone really take specifics into mid game? i find myself struggling to just macro and micro , to even consider this
also, i know you only mentioned the drone count; but what if let's say it's 9 minerals per expansion and you already have 9 drones; but that would affect mining time wouldn't it?so i'm assume you calculated all this just using an unsaturated mineral patch, but where there's already like 13 or so drones, wouldn't the "equation" you made , lack a bit of minerals. also, if you have more expansions, you could need even less workers, since all patches are open to mine
If you made mass hydra vs storms, that's a bad strategic choice for you, it has nothing to do with econ management, you should have scouted earlier and went for a different econ build up to support a different army mix.
i was just using it as a case study ; let's just say you opened with a hydra build and you scout no rush , +2 -> mass templars or w/e , and adapt into lurkerling.
All toss needs to do is make a goon upgrade +1 ; kill your scout ovie, and you are in the dark
Not to turn this into a "omg how to i scout toss after early game" but ... i generally don't have any scouts after my initial overlord; which is usually on a cliff and doesn't see much besides the forge either way.
Let's say you went 4hatch hydra... and now you want to transition into 5hatch lurker ling.
Like you said before you want to ideally have a larger army yet keep up with production rate, but realistically, in this case you would have to drone up and your army would become smaller than his, so i'm just suggesting you can't always have your army superior and yet still maintain the production speed required.
As opposed to what you said,
All other builds work the same way, if you follow through the build perfectly you end up having just the right number of drones to execute that strategy.
|
of course this is just a C newb zerg talking hurr ~
|
Thanks dude! This really helped me out alot (yes i suck, but now ill suck less! profit!)
|
So last night I modified Python a bit to help me measure resource collection rates - I'll skip over the details unless anybody is interested, but at 100% efficiency (1 worker per patch), one worker collects about 0.75 minerals per second of in-game timer, and gas is collected about at 1.5 times that rate (1.125). Larva spawn every 20 seconds of the in-game timer. So in the time it takes one larva to spawn, one worker will gather an average of 15 minerals and 22.5 gas (Obviously, resources are collected in multiples of 8 and aren't continuously added, but you get the idea. I also assume that a gas geyser with 3 workers on it is still at 100% efficiency, although I haven't tested it yet).
What this means is, by my calculations, to sustain constant hydralisk production alone, infinitely, from one hatchery, you'd actually need only 5 workers on minerals, and 2 on gas. HOWEVER, keep in mind that you will need to continually add Overlords (1 for every 8 hydras), which means you actually need 6 workers on minerals, and 1 on gas. But this actually produces a surplus of 110 minerals for every 8 hydras + overlord, meaning if you are pumping hydras from two hatcheries, you can get away with 11 on minerals and 2 on gas. On three hatcheries, since you'd need 16 on minerals and 3 on gas, and on four, you'd need 21 on minerals and 4 on gas. So really, it's more accurate to say 5.2 on minerals and 1 on gas for each hatchery producing hydras. I've tested this out in-game, and it seems to support my conclusions. Anyways, I hope this information will help people produce even more exact builds.
|
U deserve a medal!
-will this be present in sc2... oh hell =( -
|
This is... amazingly interesting, and probably has the potential to be deadly and extremely helpful... But remembering all of this is too much for a lowly weak player, especially me. I think most people are better off doing it by feel rather than by math.
|
I think most people are better off doing it by feel rather than by math.
Fuck that i've been looking for an starcraft easy button for years :D
sup dude ur easy i can just make 25 drones 4hatch hydra and rape u yawnrape savior style !
but in all seriousness i think doing it by math helps you get a clear, better understanding of the timing window.
interested where this thread is going to go :O
|
I found that 6/3 is perfect for muta.
But I guess you calculated the overlord separate at 2 so it evens out. (need a lord for every 4 mutas)
http://www.teamliquid.net/forum/viewmessage.php?topic_id=78202
As far as the thread goes I found that 10 drones on minerals at each base is the optimal, for a standard 3 hatch game with hatches regularly going up and unit pumping.
Any extra minerals or gas should go to expos, upgrades, or tech.
I mean, you don't go into a game going; "Ok, I need 4 drones on minerals here for lurker and 2 for ling so that's all i'm gonna ever need to win."
No, you gotta take into account that you need to expo and build tech stuff before you even get to that minimal point.
This info is mostly only useful in those clutch games where you lost drones or whatever and you can't afford to make anymore econ and/or you can seize an opportunity to win right there. Otherwise you'll probably already have all the requirements met and overkilled (so you can consistently expo/upgrade/tech as well)
PS- Queen data is basically the same as mutalisk 100/100 2 supply cost other than the build time.
|
I'm pretty uncertain about this. On the one hand, this kind of information is crucial in low-econ situations (ZvZ, tier 1 timing attack ZvP) where every single mineral needs to count. And I'd be shocked if Terran and Protoss players haven't had a similar table for their respective units for a long time already, and if they haven't designed their builds (not just timing attacks) at least partially off of those tables. I'd even go so far as to suspect that the recent near-perfection of Protoss and Terran macro couldn't have been done without a simple chart stating how many probes/scvs on minerals/gas are needed to pump tanks/carriers/vessels/unitx.
On the other hand, the unique manner by which Zerg units are produced seems to make playing on a knife-edge, pumping drones until the very last second and then shifting all hatcheries over to units, the most effective method of playing. Which is what most Zerg players try to do. This method is inherently unstable and it's easy to get either too greedy or too poor. But then again, a chart like the OP's might help with just that problem, though the point at which a player shifts to unit productions is still dependent on the progress of the game at hand.
I'm curious what would happen if a Zerg player forced themselves to treat certain hatcheries like a CC all the time, pumping workers basically non-stop, while treating certain hatcheries like Barracks/Factories/Starports. I honestly don't know whether playing like the other two races might help; even if it did work better than what's done now, it would make Zerg a lot less interesting. You know you're playing Zerg when they tell you've made too many workers.
|
On September 18 2008 03:42 HeavOnEarth wrote:of course this is just a C newb zerg talking hurr ~
This type of thing is used as a tool, mostly before you play a game or when you watch one and study to understand it. Just general comprehension on the subject will give you an advantage over alot of players.
|
One of best strategy threads eva!!!11
|
yeah this thread is really cool, i would have recommended it in the recommended threads but somebody beat me too it.
im mostly intertested in the terran section, i think it could help with my TvP
|
This thread is pure 24-carat gold.
Too many newer zergs don't understand the concept of overdroning and how much it hurts their macro. They treat their hatcheries like nexuses and try to keep up with probe production instead of army production. Eventually, all the extra drones and hatcheries become wasted investments against protosses with any kind of timing sense when the protoss pushes out with ~140 supply while the zerg is stuck at <100 with 50 drones and no army.
I think this thread will take a lot of Zergs to the next level.
Recommended?
EDIT: w3jjjj, can you recalculate the drone counts to where you are limited by larva only, not gas?
|
I just want to say thanks and how extremely useful a guide like this is.
I've been thinking of calculating and making a guide like this myself, but I never had the time or patience to do so. I'm one of those zergs that as Saracen says makes too many drones and ends up not having enough 'military' units when the push arrives. I've always wondered and been amazed at how players like July can have only a couple drones mining off multiple expos and yet produce an insane amount of units from them.
I should copy this down, haha.
|
since vessels take 3 scvs on gas per starport
does that mean its impossible to do sk terran (2 port vessels) with only 2 bases?
maybe it works out because of leftover gas + noob macro ^.^
|
On September 19 2008 22:52 Saracen wrote:
This thread is pure 24-carat gold.
Too many newer zergs don't understand the concept of overdroning and how much it hurts their macro. They treat their hatcheries like nexuses and try to keep up with probe production instead of army production. Eventually, all the extra drones and hatcheries become wasted investments against protosses with any kind of timing sense when the protoss pushes out with ~140 supply while the zerg is stuck at <100 with 50 drones and no army.
I think this thread will take a lot of Zergs to the next level.
Recommended?
EDIT: w3jjjj, can you recalculate the drone counts to where you are limited by larva only, not gas?
This thread is pure 24-carat gold.
Yes , it is
Thanks man =]
|
On September 18 2008 06:08 Underwhelmed wrote:
So last night I modified Python a bit to help me measure resource collection rates - I'll skip over the details unless anybody is interested, but at 100% efficiency (1 worker per patch), one worker collects about 0.75 minerals per second of in-game timer, and gas is collected about at 1.5 times that rate (1.125). Larva spawn every 20 seconds of the in-game timer. So in the time it takes one larva to spawn, one worker will gather an average of 15 minerals and 22.5 gas (Obviously, resources are collected in multiples of 8 and aren't continuously added, but you get the idea. I also assume that a gas geyser with 3 workers on it is still at 100% efficiency, although I haven't tested it yet).
What this means is, by my calculations, to sustain constant hydralisk production alone, infinitely, from one hatchery, you'd actually need only 5 workers on minerals, and 2 on gas. HOWEVER, keep in mind that you will need to continually add Overlords (1 for every 8 hydras), which means you actually need 6 workers on minerals, and 1 on gas. But this actually produces a surplus of 110 minerals for every 8 hydras + overlord, meaning if you are pumping hydras from two hatcheries, you can get away with 11 on minerals and 2 on gas. On three hatcheries, since you'd need 16 on minerals and 3 on gas, and on four, you'd need 21 on minerals and 4 on gas. So really, it's more accurate to say 5.2 on minerals and 1 on gas for each hatchery producing hydras. I've tested this out in-game, and it seems to support my conclusions. Anyways, I hope this information will help people produce even more exact builds.
I guess if u count it as 6 drones per hatch for hydra then it includes that 12 workers dont mine 2x more than 6 workers etc, cuz of minerals number ;
|
|
|
|
|
|
EPT
