|
United Kingdom20345 Posts
On July 13 2012 13:44 Xapti wrote:
This is a bit off-topic, but it's still related to zerg economy:
You know what I find annoying? The quite large discrepancy that exists between economy in some maps (with same mineral field counts). I don't know if this happens with other races aside from zerg (I would think it does, maybe not to the same degree), but I've noticed around a 5-7 second difference on different maps when building an overlord and pool.
Xel'naga caverns gives the fastest economy I've seen (be it top position or bottom — although top is maybe 1 second less), while many other maps are significantly slower such as daybreak — even if the minerals are mined optimally.
I find this to be most annoying as a zerg, since many builds early on require strict timings, and 7 seconds (or more if you also don't mine optimally) can really make significant holes in builds such as losing larvas. The best example I can think of is 11 overpool build, where the difference in economy can be between losing most of a larva and virtually no fraction of a larva.
It is extremely annoying for protoss too, especially as all of the robot zergs stuck in diamond (there are many of them, seems to be the majority) mostly got there from blindly following timings, like X minute gas, roach warren, evo chamber, etc, and when any aggressive timing can be delayed by mineral counts and the enemy is following predetermined time stamps without scouting, it is notably harder to do damage
|
On July 13 2012 09:01 Maytsh wrote:Show nested quote +On July 13 2012 06:39 Danglars wrote:
I don't care if your mining rate with the third drone is 5 minerals per minute less than the first two. It's not 5 minerals less per minute for the new drone, it's 5 minerals per minute for all drones. The statistic is a bit misleading. So taking OP's numbers - if 16 drones mine 670 minerals and 20 drones mine 750, this means that the four last drones only contributed 17 minerals per minute - which means that each of the first 16 drones is almost three times more useful than the 17th one upwards. Or put another way - while the first 16 drones collect back their cost in roughly 70 seconds, the 17th drone upwards takes a whopping 3 minutes to do the same thing. It's a pretty safe bet that you'd have more useful things to invest in if that's your window. I think there's a pretty good reason for calling anything past 16 drones oversaturated.
If you read the context, I was correcting improper use of the word fully saturated for 16 drones on minerals for the screenshot. I was not suggesting that the difference does not have a big contribution on a zerg's income management (It does) or the wisdom of under-saturating bases in favor of drone optimal mining efficiency (It is a very wise decision PRECISELY for the cost of drones in supply, cost, and opportunity cost for the drone's minerals).
The OP has changed the misleading titles on the screenshots so it's all resolved. There was no misunderstanding as to what he said, and the reality of the situation he describes. This is not the first thread on the topic nor will it be the last. My learning came from Piousflea's research as well as that found on LiquiPedia.
There is change from 24 minerals per minute to 39 minerals per minute going from the first/second worker to third worker, amountint to a ~38% penalty for the third. He will earn his money spent back (not considering supply in overlords) at ~125 seconds compared to ~77 seconds according to my calculations. Not as overstated as you suggest at 3 minutes but substantial enough to give credence to what OP says. I both agree with his conclusions and suggest higher level zerg players to consider their 3base saturation levels when defending Protoss aggression.
|
Nice guide. if basic. thanks again Tang.
Actually, a real comprehensive guide to SC2 economics would be a real boon to the community.
Its not just the amount of workers, but the flow speed and volume of all 3 races in different siituations that needs further examination.
e.g. how does larve injects + droning compare with muling, compare with chronoboost? I imagine P would have the smoothest exponential graph, Z with the fastest and largest, and T with the biggest spikes.
The way income comes in definitely affects the unit choice output / ratio and perhaps can better teach us what units we should build and when.
|
if ur in the lower leagues stick to a full page of drones to measure saturation (3 rows on bottom=24) and use the crtl click method, believe me, if you are playing a game and get to the 10-15 mins, check again, you have about a full line missing. why? youre making buildings, spine and spores.
Just get to a full page, less to think about, easy to quickly glance at, no need to bother boxing as most of the time you may get ovies in there queens and stray units, its too much micro for you to be dealing with. control click, add 2 if you know u have the 6 on gas, hell even click the gass and check 3 are on it. I read too much into this when i was getting out of bronze and silver, a simple replay check i hadnt made enough drones in the first place.
Tang is the man to listen to though, i just think these pros overestimate the ability of us lowbies, yes we can do it but its more distracting for us, we just cant do it accurately.
Perfect saturation i think should only be thought of past plat as then you have to start to know the game past solid mechanics. Its just what ive found over the years. you need to make your game more simple.
ive seen too many lowbie streams where they do the box, forget to deselect and move all of their units OFF the lines . . yeah . . . boxing=goodtactic, you just lost 20 secs of mining, forgot to put the 4 back on gas, created more of an issue for yourself, missing injects, not looking at minimap . . it spirals out of control.... GG.
EZMODE= Cntrl click workers, see full page. Move on. Really.
|
On July 13 2012 06:33 BoggieMan wrote:
the title is kinda missleading since alot of this can be used by all races, great advice for the players that don't know all this already, i however think you should write that people in the lower leagues should not focus on this at all, cause its kinda silly wasting time practicing stuff like this when its your general gameplay that has huge flaws.
I couldn't disagree with you more. The problem with newer player's general gameplay is that their mechanics are weak, they typically under or over-drone to a severe degree, and they have little to no understanding of how to build up and manage their economy.
This is practically the ONLY thing new players should be focusing on because it's an excellent starting point to learn how to play. Knowing how to check worker saturation and knowing what ideal saturation even is is practically a ticket out of Bronze league.
On July 13 2012 09:30 Staboteur wrote:
For the sake of lower level players such as myself, I can only imagine that achieving and maintaining an optimal saturation are the only aspects of any practical use, as the biggest problem for many lower level players isn't so much having not enough minerals as too much of them. Though you could obviously further optimize things by preventing bouncing drones, building drones from the larva closer to the actual undersaturated base or doubling up the closer patches earlier... unless it achieves a consistent result that is necessary for your build, I can only see it resulting in me floating 1500 minerals in the midgame instead of 1000, which doesn't actually help me win games.
If you EVER see yourself with 1000 minerals, and you aren't saving up for a round of 10 mutas when your spire pops or something, put down a hatchery. Always. If you can't spend your money, you need to be producing more units, and that means you need more production structures. It's 1000 minerals that aren't doing anything for you. Take a drone and go expand, or throw down a macro hatch close to your base. Even if you don't feel particularly safe expanding, just do it. If you lose the hatch because you couldn't defend, you haven't really lost anything, because typically you wouldn't be spending those minerals anyway.
On July 13 2012 09:30 Staboteur wrote:
I can't see managing the finer points of economy being any practical use of APM for anyone below high masters. Am I wrong in this?
League isn't really the relevant factor here. If you have the extra APM to spare, then it's just another little thing you can do to give yourself an advantage. Honestly though, a lot of the things the OP mentioned aren't huge APM timesinks. Checking saturation, pulling drones to make a building from the side closest to where the building will be, and grabbing drones from the far mineral patches rather than close ones are all pretty much one-action tasks.
|
Very good guide. I think I've reached the point, though, that the absolute best thing that I could do right now is get a better computer. My computer is so slow I'm unable to do so many things like maintain more than two or three bases, use significant army sizes, or really do anything in the late game at all. I play at an average of 8 frames per second, lower if I'm in a battle.
Irrelevant complaining aside, though, great guide~
|
Is worker efficiency always the best kind of efficiency to go for?
Seems to me you have a choice between "per-worker" efficiency (16 workers on minerals) and total income per minute efficiency (build workers until there's no more gain to income per minute).
Both have pros and cons, so is limiting mineral harvesters to 16 per base always the best?
|
I recommend not using boxing to count saturation, too much shit around the hatchery which might be boxed, or too high risk to miss a drone or two. Instead, just hold control and click on a drone. 16+4 (6 for gas but only 4 visible at a time since 1 is always inside the gayser) and you're perfect.
|
Impressive and super guide Tang
Well Done.
|
Usefull guide,
this should be pretty helpfull for some of us, although you could go a little deeper, i would have loved a section about building drones over 16 17 ingame seconds (~20 to compensate for traveltime) before you need them for structures and gas (~5:10 2 drones for extractors - 5:40 6 drones for gas - 6min full gas saturation).
|
On July 13 2012 13:56 BriMikon wrote:
Thank you for your great wealth of information Tang. I am interested on you viewpoint toward which is the best queen inject method. I use the backspace method, though i switch the base camera hotkey to spacebar, and it has worked wonders for me. Do some of the korean bawss' like Nestea use something like individual queens on 5,6 and 7 because it has strategic advantage or maybe stuck in BW ways?
Thanks again for your time.
Well because I don't use camera hotkeys very often, I prefer to hotkey my queens as 5-6-7 etc. I like it for three reasons:
1) I've been doing it since beta, and it's become habitual. I don't have to think about it, I have it timed in my head.
2) I can easily go between my main, natural, and third by clicking 55, 66, 77.
3) I can quickly select a queen if I need to micro her.
I really think it's just best to do the method that you're most comfortable with, each has its own advantages.
|
On July 13 2012 22:43 Gajarell wrote:
Usefull guide,
this should be pretty helpfull for some of us, although you could go a little deeper, i would have loved a section about building drones over 16 17 ingame seconds (~20 to compensate for traveltime) before you need them for structures and gas (~5:10 2 drones for extractors - 5:40 6 drones for gas - 6min full gas saturation).
I actually really like these suggestions.
|
On July 13 2012 14:39 Cyro wrote:Show nested quote +On July 13 2012 13:44 Xapti wrote:
This is a bit off-topic, but it's still related to zerg economy:
You know what I find annoying? The quite large discrepancy that exists between economy in some maps (with same mineral field counts). I don't know if this happens with other races aside from zerg (I would think it does, maybe not to the same degree), but I've noticed around a 5-7 second difference on different maps when building an overlord and pool.
Xel'naga caverns gives the fastest economy I've seen (be it top position or bottom — although top is maybe 1 second less), while many other maps are significantly slower such as daybreak — even if the minerals are mined optimally.
I find this to be most annoying as a zerg, since many builds early on require strict timings, and 7 seconds (or more if you also don't mine optimally) can really make significant holes in builds such as losing larvas. The best example I can think of is 11 overpool build, where the difference in economy can be between losing most of a larva and virtually no fraction of a larva. It is extremely annoying for protoss too, especially as all of the robot zergs stuck in diamond (there are many of them, seems to be the majority) mostly got there from blindly following timings, like X minute gas, roach warren, evo chamber, etc, and when any aggressive timing can be delayed by mineral counts and the enemy is following predetermined time stamps without scouting, it is notably harder to do damage
Sorry, but you just don't know what you're talking about. If there's any robot players in diamond, they're almost all protoss. 60% of the time if you ask a protoss player "Hey, how do I hold what you did this game" they say " I dunno" because they just don't have any understanding of the game. If you ask the same question to the zerg player, 90% of the time he'll have an answer like, "get earlier blink and make cannons until your archons are out" or something. Just my personal experience.
You can't win with zerg by blindly following ANY timings. What does attacking with roach at 12 min work against? Protoss taking a 3rd and not being prepared for it. That's the only thing. Zerg needs to have the correct reaction for EVERYTHING ELSE. Even if they hold your all in, they still need to play it out really well, if you just turtle correctly and get out a mothership deathball (on a map like daybreak, where a zerg attack is impossible.) then you can still win with 3 base vs 8 base.
|
On July 13 2012 23:18 EnE wrote:Show nested quote +On July 13 2012 14:39 Cyro wrote:On July 13 2012 13:44 Xapti wrote:
This is a bit off-topic, but it's still related to zerg economy:
You know what I find annoying? The quite large discrepancy that exists between economy in some maps (with same mineral field counts). I don't know if this happens with other races aside from zerg (I would think it does, maybe not to the same degree), but I've noticed around a 5-7 second difference on different maps when building an overlord and pool.
Xel'naga caverns gives the fastest economy I've seen (be it top position or bottom — although top is maybe 1 second less), while many other maps are significantly slower such as daybreak — even if the minerals are mined optimally.
I find this to be most annoying as a zerg, since many builds early on require strict timings, and 7 seconds (or more if you also don't mine optimally) can really make significant holes in builds such as losing larvas. The best example I can think of is 11 overpool build, where the difference in economy can be between losing most of a larva and virtually no fraction of a larva. It is extremely annoying for protoss too, especially as all of the robot zergs stuck in diamond (there are many of them, seems to be the majority) mostly got there from blindly following timings, like X minute gas, roach warren, evo chamber, etc, and when any aggressive timing can be delayed by mineral counts and the enemy is following predetermined time stamps without scouting, it is notably harder to do damage Sorry, but you just don't know what you're talking about.
I don't mean to preach, but telling people they don't know what they're talking about will not convince them to listen to your counter-argument (regardless of who is right or wrong).
|
On July 13 2012 14:39 Cyro wrote:Show nested quote +On July 13 2012 13:44 Xapti wrote:
This is a bit off-topic, but it's still related to zerg economy:
You know what I find annoying? The quite large discrepancy that exists between economy in some maps (with same mineral field counts). I don't know if this happens with other races aside from zerg (I would think it does, maybe not to the same degree), but I've noticed around a 5-7 second difference on different maps when building an overlord and pool.
Xel'naga caverns gives the fastest economy I've seen (be it top position or bottom — although top is maybe 1 second less), while many other maps are significantly slower such as daybreak — even if the minerals are mined optimally.
I find this to be most annoying as a zerg, since many builds early on require strict timings, and 7 seconds (or more if you also don't mine optimally) can really make significant holes in builds such as losing larvas. The best example I can think of is 11 overpool build, where the difference in economy can be between losing most of a larva and virtually no fraction of a larva. It is extremely annoying for protoss too, especially as all of the robot zergs stuck in diamond (there are many of them, seems to be the majority) mostly got there from blindly following timings, like X minute gas, roach warren, evo chamber, etc, and when any aggressive timing can be delayed by mineral counts and the enemy is following predetermined time stamps without scouting, it is notably harder to do damage
To be fair, there are quite a few builds designed to counter the blind-roach aggression in the midgame. The +1/+1 immortal/sentry/stalker push is a good example, I remember mana 2-0'd Stephano with this build. There's also ways of putting on pressure with things like a 1Zealot/1Stalker "poke" into a 4gate zealot warp-in to throw Zerg off their timings and force them to produce units before reaching full 3base saturation, which makes it harder for them to shut down protoss's third. Some players use Stargates pretty successfully to pressure before robo/third.
|
On July 13 2012 23:18 EnE wrote:
You can't win with zerg by blindly following ANY timings. What does attacking with roach at 12 min work against? Protoss taking a 3rd and not being prepared for it. That's the only thing. Zerg needs to have the correct reaction for EVERYTHING ELSE. Even if they hold your all in, they still need to play it out really well, if you just turtle correctly and get out a mothership deathball (on a map like daybreak, where a zerg attack is impossible.) then you can still win with 3 base vs 8 base.
I'm not sure I agree with this. If you follow the build order steps / timings of the Stephano-Style 12min maxout and execute it well, it can be quite difficult for Protoss either to win with a 2base all-in or secure their third. There were a few months where Protoss players were left scratching their heads against this style of play, it seemed almost impossible. I do agree that you need to do some scouting (most importantly the timings of Protoss expansion geysers), but for the most part the style revolves around getting the correct upgrades and reaching that full 3-base saturation by around 8/8:30, so you can move out by 10:30/11:00 with +1 missle, roach speed, and 150~ supply roaches.
|
I really disagree with the notion of 'optimal mineral saturation'.
You stop at (16+16+16)+(3+3+3)= 57 drones on 3 base (including gas) and you will find yourself starved in all 3 match-ups.
Yes, you want to make sure your bases are always even on workers, at all times, but I don't think the idea of optimal saturation has any practical impact on the game. You want to strive for 70+ drones...
|
On July 14 2012 00:41 Belial88 wrote:
I really disagree with the notion of 'optimal mineral saturation'.
You stop at (16+16+16)+(3+3+3)= 57 drones on 3 base (including gas) and you will find yourself starved in all 3 match-ups.
Yes, you want to make sure your bases are always even on workers, at all times, but I don't think the idea of optimal saturation has any practical impact on the game. You want to strive for 70+ drones...
While you are correct in your latter statement, in your first statement you assume that you stop at 3 bases. Tang is explaining optimal saturation per worker, not optimal saturation on X amount of bases. Of course having 70+ workers on 3 bases is better than having 57 drones on 3 bases. At the same time, having 70+ workers scattered over 4 bases is better than having 70+ workers on 3 bases.
Really great guide Tang! Thank you for sharing this. It would also be interesting to see some kind of evidence for which worker transfer (manyard) technique is the best, while you are still on this topic.
|
Very informative guide Tang, Economy Management is a very important thing to focus on when trying to improve your overall gameplay. Great read as always ^_^
|
On July 14 2012 00:41 Belial88 wrote:
I really disagree with the notion of 'optimal mineral saturation'.
You stop at (16+16+16)+(3+3+3)= 57 drones on 3 base (including gas) and you will find yourself starved in all 3 match-ups.
Yes, you want to make sure your bases are always even on workers, at all times, but I don't think the idea of optimal saturation has any practical impact on the game. You want to strive for 70+ drones...
Actually it's 66 drones for 3 bases with all the gas:
(16x3) + (6x3) = 66
Having more drones than that is inefficient, as you'll end up with triple mining patches or drones just bouncing around. 70+ is beneficial on 4base.
|
|
|
|
|
|
EPT
