I'm really surprised that nobody really covered this technique on Teamliquid when it's been covered extensively on the Korean forum, PlayXP.com. While worker splitting is not as important as it was in Starcraft 1 because Starcraft 2 comes with a pretty good AI that automatically splits the workers for you, it's not exactly the best and manually splitting them still results in optimal mining.
Also, there's one more way to make your mining even more optimized in the beginning stage of the game. I'd like to coin the term "stack mining", since I don't see any other term for it & it's the best I could come up with :/
This is only a minor advantage and not really going to be the decisive factor to your gameplay if your level is lower than Diamond, but in higher levels and in matchups such as PvP where getting the warpgate research asap is crucial, this method could prove to be quite advantageous to your gameplay.
For proof on effectiveness, watch the TSL games of oGsMC vs White-Ra
I. Worker Splitting
Your goal is to have all the workers mine in the closest spot, as fast as possible without having them bounce at all. Keep in mind that this will take some practice to get it as perfectly as possible. Here is the way I do it:
1. Locate "Key" minerals. This will be the mineral patches which you will make the right click on to optimally separate your workers into.
2. Select all your workers, and right click them onto one of the key mineral patch. Then make a small box to select three workers which will go to the other remaining key mineral patch.
3. Your workers then will automatically be split and be efficiently mining from the mineral patches asap
II. Stack Mining
The key thing to remember in mining minerals is that the closest mineral patch yields faster mining and if done right, you can stack two probes on a single mineral patch without having them bounce even if the other mineral patches are not being mined.
From this, it is without saying that if you can get the probes to mine from the closest mineral patches first, you will result in seeing maximum mining efficiency.
1. Your worker that is mining on the farther patch than the closer one is going to be desynchronized eventually, meaning that a situation of one worker is returning mineral to the base where the other worker is still mining will happen. Keep in mind that the mining pattern is non-linear and slightly random, so you have to locate the better de-synchronized worker. In this case, it's the worker that has been selected. This de-synchronization happens when your 8th worker is being produced.
I personally prefer to stack the mineral patch that's away from the gas, because it's easier to build with the outer-lying probe.
2. Force the worker to mine on the closer mineral patch by vigorously right clicking on it to prevent it from bouncing to another patch. Make sure to be right clicking as closely to the base (Nexus) as possible to make sure the two workers are perfectly synchronized to the same patch. You also don't want to be losing too much mining time from doing this, so make sure to do this when the worker you've selected is perfectly or near-perfectly de-synchronized.
3. Repeat the procedure for the other mineral patches. In this case, you can see that the probe selected is almost desynchronized.
4. Eventually, your mineral patches should look something like this
From the above youtube link, you can see that MC had a lead by a probe in all three of his games by doing this.
Comments, criticisms and questions are always welcome.
Does splitting in SC2 actually affect the mining capability of the player? i.e would a player that doesn't split have the same amount of minerals at a certain point in time?
On April 13 2011 05:32 islandman wrote: Interesting.
Does splitting in SC2 actually affect the mining capability of the player? i.e would a player that doesn't split have the same amount of minerals at a certain point in time?
Yes it does affect, but not to a ridiculous degree.
On April 13 2011 05:35 MrArarat wrote: I don't think that in our level of gaming (not pro) a 30/40 minerals lead will change the outcome of the match.
Regardless of what you think, having good habits is something you should strive to achieve if you want to improve. I actually do exactly what this guide shows every game I play.
On April 13 2011 05:35 MrArarat wrote: I don't think that in our level of gaming (not pro) a 30/40 minerals lead will change the outcome of the match.
In most pro matches it also doesn't decide the outcome, but if u do this u dont just mindless spam ur 1-2-3 buttons and actually do something + gaining a small advantage. What is there to lose?
On April 13 2011 05:35 MrArarat wrote: I don't think that in our level of gaming (not pro) a 30/40 minerals lead will change the outcome of the match.
Regardless of what you think, having good habits is something you should strive to achieve if you want to improve. I actually do exactly what this guide shows every game I play.
I do it also! Don't get me wrong. I mean, that its not that HUGE advantage.
It really only matters in PvP in my opinion, and even then, it factors in to about .01% of the whole game. It has been discussed to death, however. Here's an example of my split that I believe is the best way to do it, but isn't optimal on ST, TDAltar, or SPlateau because of the mineral patch locations:
My opinion is that it is better to occupy all mineral patches before doubling up, since if you make a mistake you have drifting probes which equal no mining at all. I prefer in this case to play it safe and not lose mining time from the beginning of the game.
It does have an affect, if you build your worker before splitting. In my expeirence if you do your split correctly you will be able to que, another worker before the first one even finishes. This is pretty good for zerg because they can have more then one worker building at a time
On April 13 2011 05:44 sgtjimmy wrote: It does have an affect, if you build your worker before splitting. In my expeirence if you do your split correctly you will be able to que, another worker before the first one even finishes. This is pretty good for zerg because they can have more then one worker building at a time
I'm able to get 10 minerals leftover when I order my second probe.
On April 13 2011 05:44 sgtjimmy wrote: It does have an affect, if you build your worker before splitting. In my expeirence if you do your split correctly you will be able to que, another worker before the first one even finishes. This is pretty good for zerg because they can have more then one worker building at a time
I'm able to get 10 minerals leftover when I order my second probe.
That's freakin good man I think I only ever get 5, probe, then 5 more... You must be fast :D
On April 13 2011 05:49 sgtjimmy wrote: Are you splitting? and this is more from a zerg standpoint
I split just like the OP details. And it's important in all matchups. If I get those 10 minerals sooner, and am prioritizing closer minerals early game, along with good building placement close to my nexus and on-time chronoboosts, I could perhaps get a free Assimilator (and I actually do this PvP to prevent gas steals with certain openings). Or perhaps I can just slightly delay a stalker in favor for a Zealot doing a double stalker opening in PvZ.
On April 13 2011 05:35 MrArarat wrote: I don't think that in our level of gaming (not pro) a 30/40 minerals lead will change the outcome of the match.
Regardless of what you think, having good habits is something you should strive to achieve if you want to improve. I actually do exactly what this guide shows every game I play.
well if you screw up this micro you could end up 30-40 minerals behind, so its kinda a double edged butterknife
I wonder how much of a difference this makes. Maybe someone with some time on their hands could test it? Either way, I will be doing this. I'm always looking for little ways to get ahead. Thanks for the nice guide.
These actually make little difference but it's something to do. You should also mention mining harrasment ie worker displacement. This is when you mine their minerals to send their probe to a different patch and waste time by moving. You can actually queue commands, mine, stop, mine diff patch, stop, mine etc and push thier workers around.
Like the OP said it is a really only realistic for high masters/pro's but this is something everyone should be practicing as it is a good habit and it works on you unit selection precision its just good practice. I also do this every game every small advantage adds up (especially as zerg trying to get a fast hatch)
Just watch high level streams, you will see them moving probes to closer mineral patches and doubling up. First rule of manual mining: Don't waste mining time!
I do this but sometimes even after the workers have been doubled up after about two or three trips back and forth they'll move over to the farther patch again. Any idea why that happens. Is there something that I'm doing wrong?
On April 13 2011 06:16 FryBender wrote: I do this but sometimes even after the workers have been doubled up after about two or three trips back and forth they'll move over to the farther patch again. Any idea why that happens. Is there something that I'm doing wrong?
Some mineral patches are so close that 2 workers is actually over saturation.
If this is the case, your workers will slowly get to the point where one worker gets to the mineral patch early enough that the AI decides it would be faster for it to move to an un-mined patch.
Learning to recognize these patches and properly manage them is also very critical to proper mineral stacking.
The thirty-forty mineral difference can help you smooth our builds, I've found. Especially on certain maps where a lot of the mineral patches are not very efficient like Tel Darim you can get a decent lead over your opponents with good syncing. It can shave a few seconds off a 4gate, for instance, and in a PvP where you get your core a few seconds sooner and your additional gateways a few seconds sooner can easily decide a game.
well i don't know if it is still worth it if the scouting worker block those close patches and forces the mining workers to go to other patches. whole effort seems wasted, though the idea behind is obviously to gain a slight minerallead to get faster and better timings.
On April 13 2011 06:31 .kv wrote: ugh you shouldn't click the mineral patch you posted in your image but actually the third mineral patch IMO...it's closer
Determining which mineral patch to click entirely depends on how the workers are desynchronized, not how "closer" the mineral patch is.
I simply chose to take that mineral patch first because it would boost my mining efficiency without losing too much mining time.
On April 13 2011 05:37 TheAwesomeAll wrote: sometimes when I stack mine, the probe decides to wander off regardless, a bit later. Anyone else has this or am i doing something wrong?
The way the worker AI functions from what I can tell is that there is a "estimated wait" threshold that dictates whether the worker decides to wait at a patch or go off and find a better patch. If you send a worker to mine a patch and it gets there and discovers that the probe mining the patch will take longer than the threshold time, it will go off and ruin your day.
In the case of where there are open mineral patches, the threshold is very low, so you have to spam a bit. Furthermore as the op hints it, if the travel time of the spammed probe is a bit longer to the base than the original probe (e.g., because the spamming shifted the probe a little off center), then on its second mining trip, it may decide to go find a new home because of the delay.
Because of this, it is much less error-prone to mineral stack after you have one probe at each patch because the threshold time is bigger. However, you lose a bit of the benefit of stacking since probes 7 and 8 will be mining far patches instead of near ones.
Great guide, this is how I split my workers too! Only downside to this was the first time I played on Tal'Darim Altar I was so thrown off by the mineral layout that I fucked up my split >_>;;
Personally, I prefer to start stacking my workers starting on the 9th worker, to avoid frustration with workers re-desyncing after lining them up properly. As of late, I've also been trying to perfect the habit of only taking far mineral patch workers when putting them into gas. It can be troublesome at times though I must admit.
On April 13 2011 05:44 sgtjimmy wrote: It does have an affect, if you build your worker before splitting. In my expeirence if you do your split correctly you will be able to que, another worker before the first one even finishes. This is pretty good for zerg because they can have more then one worker building at a time
I'm able to get 10 minerals leftover when I order my second probe.
On April 13 2011 05:42 tehemperorer wrote: It really only matters in PvP in my opinion, and even then, it factors in to about .01% of the whole game. It has been discussed to death, however. Here's an example of my split that I believe is the best way to do it, but isn't optimal on ST, TDAltar, or SPlateau because of the mineral patch locations:
My opinion is that it is better to occupy all mineral patches before doubling up, since if you make a mistake you have drifting probes which equal no mining at all. I prefer in this case to play it safe and not lose mining time from the beginning of the game.
I find that it is equally useful in Terran MUs because of the need for workers to be away from mining to build, it gives a bit of an edge when you can squeeze a bit more out of your early game.
On April 13 2011 05:42 tehemperorer wrote: It really only matters in PvP in my opinion, and even then, it factors in to about .01% of the whole game. It has been discussed to death, however. Here's an example of my split that I believe is the best way to do it, but isn't optimal on ST, TDAltar, or SPlateau because of the mineral patch locations:
My opinion is that it is better to occupy all mineral patches before doubling up, since if you make a mistake you have drifting probes which equal no mining at all. I prefer in this case to play it safe and not lose mining time from the beginning of the game.
I find that it is equally useful in Terran MUs because of the need for workers to be away from mining to build, it gives a bit of an edge when you can squeeze a bit more out of your early game.
it's not just PvP; for Zerg (as someone mentioned) the faster you can hit 50 minerals in the early game, the faster you can make drones. with proper stacking i can make 2 drones at the same time after my 9 ovie spawns, with too much bouncing i have to wait a second or two between the drones. it's not a big deal, but as cecile mentioned, it's best to practice good habits!
I've been doing this forever, as have a bunch of pros, the difference it makes is insignificant, but it is fun to have something to do before my rax finish =D.
On April 13 2011 05:35 MrArarat wrote: I don't think that in our level of gaming (not pro) a 30/40 minerals lead will change the outcome of the match.
Regardless of what you think, having good habits is something you should strive to achieve if you want to improve. I actually do exactly what this guide shows every game I play.
well if you screw up this micro you could end up 30-40 minerals behind, so its kinda a double edged butterknife
This topic is set up to provide a way to improve, be it by a fraction. I totally have to agree that, if you want to improve, you should eventually also strive to perfect worker splitting and stack mining. You don't have anything else to do at this stage of the game anyway and it might help you getting focused at least more than if you'd have done nothing. By the way, I've been using this technique since I was in bronze, in search of a way to perfect my almost perfect pre-opening and I found it quite hard to screw up this micro after a few games, even at that time.
I just wanted to clarify a slight mistake in the OP. It seems to suggest that sending 3 workers to the middle, further patch is fastest, but actually it is not. For example, in the picture provided, you would want to select the most left patch, and for the second 3 probes, send them to the most right patch. It does make a small difference in the speed that each probe reaches the closest mineral patches.
Using the OP's method causes all workers to travel to and bunch up on the furthest patch when it is not necessary. You can see a real difference if you test it. It's also just easier to click the furthest patches too, in my opinion.
On April 13 2011 06:31 .kv wrote: ugh you shouldn't click the mineral patch you posted in your image but actually the third mineral patch IMO...it's closer
The point is not to get the 3 probes to a close mineral patch as soon as possible. The goal is to get as many close mineral patches mining as possible. Assume, like on XNC, the mineral fields are spread like this;
1. Close 2. Far 3. Close 4. Far 5. Far 6. Close 7. Far 8. Close
His guide says to click 3 probes onto 2 from the start. This will result in automatic splitting of these 3 probes among 1, 2 and 3, thus giving mining from 2 close and 1 far. If you instead do what you say, click the 3 probes on 3, the auto splitting will place the probes on 2, 3 and 4, mining from 1 close and 2 far.
Clicking 3 probes to 2 and 7 gives 4 close and 2 far mining nodes, resulting in the overall most efficient mining. When you then synch up the workers eventually (added bonus; this is easy to do when theres already a desynched probe at 2 and 7 you can place on 1, 3, 6 or 8 all depending on which lines up better).
Great writeup OP, surprised that I havent seen this from someone already :D
I may not have blue post or high masters rank or been in hundreds of leagues... But I can tell you this. Idra never lost a game cause he didn't split his drones. This basically means that sure, you get a "lead" but that lead is instantly lost once you hit the mid-game. Once you hit certain variables where 5 minerals no longer is an issue (maynarding probes, oversaturation, loss of probe etc) You having that lead means absolutely nothing. It is a good habit, and it actually was worth doing in broodwar because your units did not auto split, but because they do so in SCII there is absolutely no reason to worry about that kind of economic lead.
Purest form of this example, if 2 zergs 6p the one who splits his drones and the one who doesn't split his drones will have lings at the exact same time OR if this example holds true and you get an extra lead (by no means can you actually apply it but w/e) you should have 3 sets of lings FASTER than the other guy. But the point is you both 6pooled, which means you both have 6 lings anyways, which means that you didn't actually get ahead or behind by having those lings 0.05 seconds sooner.
I don't mean to downvote your post, I think it is a great idea to split guys, it gets into good habit and causes a certain flow to your builds. I would look more to how that helps your timing for your first buildings more than if it helps you get more workers... It doesn't create economic lead as you both have the same exact minerals regardless at the end of the day.
On April 13 2011 07:28 Kornholi0 wrote: I may not have blue post or high masters rank or been in hundreds of leagues... But I can tell you this. Idra never lost a game cause he didn't split his drones. This basically means that sure, you get a "lead" but that lead is instantly lost once you hit the mid-game. Once you hit certain variables where 5 minerals no longer is an issue (maynarding probes, oversaturation, loss of probe etc) You having that lead means absolutely nothing. It is a good habit, and it actually was worth doing in broodwar because your units did not auto split, but because they do so in SCII there is absolutely no reason to worry about that kind of economic lead.
Purest form of this example, if 2 zergs 6p the one who splits his drones and the one who doesn't split his drones will have lings at the exact same time OR if this example holds true and you get an extra lead (by no means can you actually apply it but w/e) you should have 3 sets of lings FASTER than the other guy. But the point is you both 6pooled, which means you both have 6 lings anyways, which means that you didn't actually get ahead or behind by having those lings 0.05 seconds sooner.
I don't mean to downvote your post, I think it is a great idea to split guys, it gets into good habit and causes a certain flow to your builds. I would look more to how that helps your timing for your first buildings more than if it helps you get more workers... It doesn't create economic lead as you both have the same exact minerals regardless at the end of the day.
I think it's more important to look at PvP, where a few seconds do matter. Plus, certain openings are suddenly viable and thus surely worth the effort.
On April 13 2011 07:28 Kornholi0 wrote: I may not have blue post or high masters rank or been in hundreds of leagues... But I can tell you this. Idra never lost a game cause he didn't split his drones. This basically means that sure, you get a "lead" but that lead is instantly lost once you hit the mid-game. Once you hit certain variables where 5 minerals no longer is an issue (maynarding probes, oversaturation, loss of probe etc) You having that lead means absolutely nothing. It is a good habit, and it actually was worth doing in broodwar because your units did not auto split, but because they do so in SCII there is absolutely no reason to worry about that kind of economic lead.
Purest form of this example, if 2 zergs 6p the one who splits his drones and the one who doesn't split his drones will have lings at the exact same time OR if this example holds true and you get an extra lead (by no means can you actually apply it but w/e) you should have 3 sets of lings FASTER than the other guy. But the point is you both 6pooled, which means you both have 6 lings anyways, which means that you didn't actually get ahead or behind by having those lings 0.05 seconds sooner.
I don't mean to downvote your post, I think it is a great idea to split guys, it gets into good habit and causes a certain flow to your builds. I would look more to how that helps your timing for your first buildings more than if it helps you get more workers... It doesn't create economic lead as you both have the same exact minerals regardless at the end of the day.
I've spent hours testing zerg openings for a few of my threads, and I can tell you that microing your workers at the start can make a SIGNIFICANT difference. I've had people send me their replays that surpassed the results I got in my testing, and the differences all came down to their methods of microing workers, where as I simply sent all workers and rallied to a single patch and let the AI do the rest in order to have consistency of results.
At the end of the day though, the real question is if there is any reason NOT to micro your workers. Clearly there isn't. You aren't doing anything else in the opening anyways, you might as well go for a mineral advantage and also practice your mouse control at the same time.
On April 13 2011 07:21 jdseemoreglass wrote: I just wanted to clarify a slight mistake in the OP. It seems to suggest that sending 3 workers to the middle, further patch is fastest, but actually it is not. For example, in the picture provided, you would want to select the most left patch, and for the second 3 probes, send them to the most right patch. It does make a small difference in the speed that each probe reaches the closest mineral patches.
Using the OP's method causes all workers to travel to and bunch up on the furthest patch when it is not necessary. You can see a real difference if you test it. It's also just easier to click the furthest patches too, in my opinion.
This is true, but this also leads to slower de-synchronization of workers, which is bad for stack mining. Eventually, it was determined that sending them to the middle was the best method.
On xel-naga you want to click those 2 middle patches like he showed in the OP. It makes 4/6 of your probes start mining the most efficient patches. It's true you could send 3/3 then break them off so they start mining them faster.. but this is the best way without having 6000 apm and the precision of a surgeon.
On April 13 2011 07:41 Alejandrisha wrote: On xel-naga you want to click those 2 middle patches like he showed in the OP. It makes 4/6 of your probes start mining the most efficient patches. It's true you could send 3/3 then break them off so they start mining them faster.. but this is the best way without having 6000 apm and the precision of a surgeon.
My 4-1-1 split that hit all close mineral patches; they all go to the closest patches naturally. Works on all maps but Tal Darim, Shakuras Plat, and Shattered Temp. Click all on middle patch, then split 1 to bottom edge then grab another split to top edge.
On April 13 2011 07:48 Alejandrisha wrote: That's probably the ideal split for that map but on xel naga you can't send 4 to one patch and have them grab the 4 most efficient patches.
Ah, I forgot different maps have very different patches. I don't know how many minerals I'd get in time when ordering the second probe, but I do know that the easiest to do/best benefit comes from a 3/3 split to the two farther mineral patches on XNC like the OP and Alej have stated.
On xel'naga specifically I prefer to split 6:2:1 sending all to the inside close mineral patch then 2 of those 6 to the opposite outside close mineral patch and finally selecting 1 of those 2 and sending it to the inside close mineral patch.
This lets you stack all you're probes on close mineral patches after 2 trips.
This does make an impact for protoss, it lets you chrono boost at just over 38 energy without being supply blocked which means your first chrono boost finishes just as you get 25 energy on your nexus for your second chrono boost making your early game mining and probe production as efficient as possible.
This for instance let's you get a 13gate 14gas without a bump in probe production.
30-40 minerals advantage is huge in rush situations, think of it less as a mineral advantage, and more as a time advantage, you will start your gateway before them by a few seconds, also start ur cyber core, and your assimilator, and warpgate tech, etc... Basically this little trick translates into about a 3-5 second advantage at the 6 minute mark, which is huuuuge in PvP, and in ZvP it can help you get those few extra seconds for units when trying to defend a 4gate.
Usually i don't care about stuff like this, but the thing people don't realise is that this trick gives you a lot more than 20-40 minerals...
It gives you 150 minerals by 5 minutes
Thats a free goddamn gateway or roach warren Thats the difference between winning and losing the game with certain builds
The difference is so pronounced its actually shown to be a better idea to force onto the close patches REGARDLESS of whether you are desynched properly or not.
Nice guide, most pros do this but lots don't know about it. One thing I would consider, especially for Terran players, is that when you pair up your workers early game like this, you mine out the closest patches quicker than the other ones.
So you have to be really careful on not muling the same patches over and over again, and sometimes mule the far away patches, otherwise you end up actually decreasing your mining efficiency a bit later in the game.
A lot of people have been mentioning how this doesn't help much, but you need to at least realize that it's something to do during the boring early game, now that being said. It's not a game breaking difference, but I agree with the OP, it's a good habit to get into.
I actually do feel that the early game probe tricks and such are the difference between having that collo out in time to stop the 4-gate (I am only in plat so the 4-gates are most likely slow) or looking like an idiot as an army marches into your base without anything to stop it.
On April 13 2011 06:06 genius_man16 wrote: I'm surprised no one has made a guide for this yet. Very well done though
I've been doing this for awhile, but I never really thought it mattered that much, I just did it cuz the pro's did. Haha.
It would be interesting for someone to test and see if there really is that much gained from doing this.
One thing I've wondered is if it's better to queue the worker before sending the 6 to mine or after?
In strict theory this depends on where your first building is placed (for P and T, for Z its actually better to send before building). Technically, if you can place your first pylon instantly upon 100 minerals its better to build first, but if there's much travel time to the pylon or you delay it at all its better to split first.
However, the major point is that "queue->split" is much faster for most people than "split->queue", probably because the mouse movements required for the latter are significantly larger. As long as you can queue->split faster than the reverse, its a better option. The reverse also holds, but I don't know of anyone for whom thats true.
[edit] My question is - when do I stop? When the 4 best patches have paired workers, or when the 6 best patches have paired workers, or what?
[edit2] I've been practicing this on XNC lower spawn, and at least on that spawn doing this right generates a around a 1 second advantage on when you can chronoboost with no waste.
Basically when you are idle, micro your workers. On XNC there are 4 close mineral patches which means when you have 10 workers you should have 8 of them on close patches, and 2 of them on the other patches.... But do not sacrifice scouting, keeping your build neat etc. for this, though
By the way, there is a good way to split his probes faster against hellions rush.
When the hellions are coming, you need to select all your probes working and press S to stop mining. From there, you just spam F1 + click somewhere, everytime you repeat the process you have to click in different location, so your probes will go quickly a bit everywhere dodging the splash damage of the blue flame. Working pretty well to not lose all probes on a shot =)
On April 13 2011 09:45 SaJa wrote: By the way, there is a good way to split his probes faster against hellions rush.
When the hellions are coming, you need to select all your probes working and press S to stop mining. From there, you just spam F1 + click somewhere, everytime you repeat the process you have to click in different location. Working pretty well to not lose all probes on a shot =)
Lol we're talking about early-game splitting to maximize mineral efficiency.
oh, and one time I accidentally right clicked all my probes into gas against hellions... I cried.
2. Force the worker to mine on the closer mineral patch by vigorously right clicking on it to prevent it from bouncing to another patch. Make sure to be right clicking as closely to the base (Nexus) as possible to make sure the two workers are perfectly synchronized to the same patch.
Ehm...what does "it" refer to?
". . . by vigorously right clicking on [the patch] to prevent [the probe] from bouncing to another patch ?
And "as closely to the base as possible" refers to the patch closest to the middle? And the reasoning is that probes on closest patches will always stay synchronized?
When you spam click a mineral patch, the probe starts to shuffle around the patch to a point which is closest to your click location. If you click the mineral patch closer to the nexus, the shuffling is minimized, lessening the chance of an instant desync when the other probe returns to that patch.
Its probably the single least important part of the whole thing - most of the time, you need to recognize that a desync is going to happen again and deal with that probe as you would any other probe - once you've sync'd each probe to a patch individually they almost never desync unless the patch itself is too close.
2. Force the worker to mine on the closer mineral patch by vigorously right clicking on it to prevent it from bouncing to another patch. Make sure to be right clicking as closely to the base (Nexus) as possible to make sure the two workers are perfectly synchronized to the same patch.
Ehm...what does "it" refer to?
". . . by vigorously right clicking on [the patch] to prevent [the probe] from bouncing to another patch ?
Yep.
And "as closely to the base as possible" refers to the patch closest to the middle? And the reasoning is that probes on closest patches will always stay synchronized?
As close to the nexus means make sure the probe is in line with the nexus, rather than a bit to the side or whatever. You need the probe to return to the nexus in the shortest possible time(technically just in the same amount of time, but since the other one is already set on the shortest possible path..), or they'll just desync again.
Obviously this only makes a small mineral difference and will almost never make or break a game, but a great guide nonetheless and something I will strive to do in all my games from now on.
Also gives me something to do to warm up besides mindless spamming which tends to hurt more than help me.
Thanks, nice read and so well explained, actually it doesn't seems that much but when you do things that have a narrow window time (like in ZvZ and PvP) where every second counts this helps a lot.
What can you say about the study where apparently, the incomes of a player who splits (3-3 and 4-1-1) turns out to be the same as someone who simply right clicks on the middle mineral patches? Just wondering. It was also said that the only difference was a 5 mineral lead for a couple seconds, but that it was virtually equal after a while (idk how long a while is).
Also, are you sure it's more efficient to group all 6 probes, then split 3 from them, instead of 3-3?
Edit: Nevermind, I found the thread and it seems the reason why the incomes are basically the same no matter how you send your workers to mine because the map that the splits were tested on did not have a mineral line where the edges were closer (ex: Xel'Naga), but where the middle patches were the same/closer.
Using the stop command upon arriving at the close patch and then issuing a single gather command when the mineral crystal appears in the resident worker's hands is more reliable than spam clicking, in my experience
Splitting alone doesn't matter. Worker stacking undeniably does produce an advantage, and splitting is necessary to worker stack optimally, thus, splitting matters.
I've been doing this without even realising. when I split I put them there because they mine all 3 min patches in the same space before spreading out on the second run. As a low level player doesn't make too much of a difference, and sometimes I miss click and have 3 drones doing nothing, BUT when it works it looks boss and gives me a good 4 or 5 seconds boost!
On April 13 2011 07:28 Kornholi0 wrote: I may not have blue post or high masters rank or been in hundreds of leagues... But I can tell you this. Idra never lost a game cause he didn't split his drones. This basically means that sure, you get a "lead" but that lead is instantly lost once you hit the mid-game. Once you hit certain variables where 5 minerals no longer is an issue (maynarding probes, oversaturation, loss of probe etc) You having that lead means absolutely nothing. It is a good habit, and it actually was worth doing in broodwar because your units did not auto split, but because they do so in SCII there is absolutely no reason to worry about that kind of economic lead.
Purest form of this example, if 2 zergs 6p the one who splits his drones and the one who doesn't split his drones will have lings at the exact same time OR if this example holds true and you get an extra lead (by no means can you actually apply it but w/e) you should have 3 sets of lings FASTER than the other guy. But the point is you both 6pooled, which means you both have 6 lings anyways, which means that you didn't actually get ahead or behind by having those lings 0.05 seconds sooner.
I don't mean to downvote your post, I think it is a great idea to split guys, it gets into good habit and causes a certain flow to your builds. I would look more to how that helps your timing for your first buildings more than if it helps you get more workers... It doesn't create economic lead as you both have the same exact minerals regardless at the end of the day.
I've spent hours testing zerg openings for a few of my threads, and I can tell you that microing your workers at the start can make a SIGNIFICANT difference. I've had people send me their replays that surpassed the results I got in my testing, and the differences all came down to their methods of microing workers, where as I simply sent all workers and rallied to a single patch and let the AI do the rest in order to have consistency of results.
At the end of the day though, the real question is if there is any reason NOT to micro your workers. Clearly there isn't. You aren't doing anything else in the opening anyways, you might as well go for a mineral advantage and also practice your mouse control at the same time.
Don't get me wrong, there is an "advantage" but really you didn't win a game because you had 5 more minerals than the other guy one second sooner. I would say that this qualifies as being one of those things that gets you from being the 2nd best player in the world to like first best player in the world... Its like a non-existant difference of a difference for anyone under *last season* 4k masters... Like I also said it causes a fluency... I'd use it more for the fact that getting those minerals 1 second sooner helps my build come together better, not exactly faster. I guess smoother should be the object of this thread, not faster. You shouldn't always want to play faster, but if you play smoother or at least have a bunch of good timings you'll be set. I guess you could also think of this as like, if I do this at the beginning, I can transfer 1 more drone to my expo and then get an even larger economic advantage or something. Snowball effects + smooth effects + mental effects = win.
I started doing this a while ago after seeing it on LiquidHuk's stream. It gives you a bit of an edge and certainly helps a lot in a mirror matchup (like PvP).
dudes i dont want to rain on your parade, but i gotta say that this is so neglectable, that's it not even funny. what i also do is making sure that the first 6 harvesters cover the closers fields, and the 9th and 10th harvester double stack the nearer patches. an emphasis with the first few harvesters, rapidly clicking like a madman, brings no real advantage with it, it will rather make you lose concentration on the game, in my opinion.
plus you will get mined out on those patches faster ( even faster when u are terran, because you have to use the closer patches for mules ), so this is not worth it imo. maybe if i do a 6 pool.
instead of saying that pros are doing it and it gives you an edge, can you post 2 replays, one where you stack and one where you dont and do the same build in both (ie a 4 gate or something quick) for a comparison? I've been trying this and unless I'm doing it wrong I still end up at the same place regardless whether or not i mineral stack
I think the initial split is something of less significant importance. Definitely worth learning to do, but if you are having trouble your practise time is better spent on other things until the higher levels.
Worker stacking, however, can be a huge difference. This is anecdotal, but I often find that, in Protoss games, I can build my gateway a full supply count earlier when stacking workers, ie being able to build it on 13 with no interruption on probe making while constantly chronoboosting, where normally you'd have to do make it on 14 to get that.
I have been doing the worker stacking for some time now. Maybe at the beginning I was actually loosing mining time, but I slowly got better at it and now I kind of get it to work. And it is better than mindlessly spamming at the beginning of a game or doing nothing at all.
i already knew about that, but im not sure if splitting is really that good on all maps. xelnaga caverns e.g. is really easy cause of these 2 "key mineral patches" as you called them, and delta quadrant has the same layout of patches, but taldarim altar is imo really hard to split, although you can kinda see such key mineral patches as well. i sometimes think that its even as effectiv to just send all your probes to the middle patch (some high master players even do it and they can get out their second probe out in time and start the 1st probe before sending workers to mine).
Well it doesn't take a scientist to see that this is an advantage.. atleast from a terran players point of view..
If you get your minerals faster, you get your army faster or you get your expansion earlier.
No matter which of the above is your goal, you would always get it faster. If you get your expansion 1 second faster, you get your 2 x orbital command faster, which means faster mules.. It might not be the ½ second you gain in the start thats advantage, but its more of what follows up.. If you get mules a bit faster, then ur next Production facility will be a few seconds earlier.. With a bit faster minerals, you might be able to squeeze a marauder out, instead of a marine in your opener..
So don't think for a minute that this doesn't help.. The earlier in your game that you can improve your saturation, the more outcome you will get, later on in the game. It's simple math baby
I wrote a thread a little while ago that demonstrates the importance of having your first 8 workers on close patches. I try to get at least 6 on them every game and it does help refine openings.
On April 13 2011 18:59 Blurb wrote: Sockfolding returns. I predict this "bug" being removed within a month, for whatever silly reason Blizzard can come up with.
It's absolutely not a bug.. don't know if you're being serious or not though.
I think its foolish to say that any advantage is neglectable OR negligable. Especially in a mirror match as evidenced by that white.ra- MC series. Just watch the production tab as they are both building their cores. Its a BIG deal. Those split second advantages in the early game always snowball into huge lead when you are both racing to warptech.
I can produce a 1-2 second faster 12gate with a split and worker stacking than without. That is in no way a negligible advantage - I can already remember games where a 1s faster warpin would have won me the game.
I think stacking is more important than a nice split. It's a better habit to get into, and the faster you double stack those probes on the closer patches the more efficient your mining will be in the long run. Splitting looks nice, but doesn't do as much for you. Stacking for the win! Either way, nice little explanation.
The benefit of splitting is it makes it easier to stack things up faster. On XNC I can reliably have 4 patches stacked while my 9th probe is building when I split right - if the split gets fucked up, its much harder to do that.
I stack my probes too now and it allows you to get ur pylon and gateway down a bit faster. With a perfect 3-3 split u can get ur pylon at 50 and sometimes 49 sec. With splitting and stacking you can get pylon down at 46 or 47 if you stack your first probes immediatly at xel naga caverns. By doing this you can gain a big lead over lower league players because most of them won't build their pylon before 1 min and their gateway is also way later.
Not sure if my way is the most efficient but I do:
CTRL+F1 to select all workers (spamming it right before game starts, I find this works quicker than boxing) Send them to a key patch Make worker Split a few off to another key patch
After most games, I verify that I build my worker at 0:01 which seems good enough. I just need to work on splitting off the second half to the right patch
click nexus, hit hotkey, f1, click patch, f1, click patch, f1, click patch, f1, click patch, f1, click patch, f1, click patch, hope the observers were watching you spread like a badass, enjoy your early game lead.
Due to chronoboost, this makes a big difference for Protoss compared to other races. If you do it well, you can do chrono boosts as if you were 12scouting with a 9scout, as well being able to chrono your nexus quicker if your 12/13 scouting.
I'm gonna kind of be the negative one here. I do this every game, and when I am doing something like a 7 pool, I make sure that all my workers are mining efficiently. In that scenario, it seems like this technique would have the most effect, since there aren't as many workers, but to be totally honest I never notice a difference.
I have been slacking off recently at the beginning of games, and I never notice having less minerals. Going back to the 7 pool, I almost always have the exact same amount of minerals when the pool pops, regardless of whether I forced my drones to the close patches.
I would love to see some data on this though. From my personal, subjective experience, though, it seems negligible at best.
On April 16 2011 15:30 Zerg-Master wrote: I'm gonna kind of be the negative one here. I do this every game, and when I am doing something like a 7 pool, I make sure that all my workers are mining efficiently. In that scenario, it seems like this technique would have the most effect, since there aren't as many workers, but to be totally honest I never notice a difference.
I have been slacking off recently at the beginning of games, and I never notice having less minerals. Going back to the 7 pool, I almost always have the exact same amount of minerals when the pool pops, regardless of whether I forced my drones to the close patches.
I would love to see some data on this though. From my personal, subjective experience, though, it seems negligible at best.
I don't think it actually matters if you aren't Protoss TBH, Protoss gain a lot from this as they can use their Chronoboost much earlier than normally, thus there is a noticeable effect on economy and timings.
Like I said before, does the OP want to supply actual proof via replays with resource scores from doing a build up to like 6 minutes with both stacking and not stacking? I've been doing this and there are no differences when I try it.
Otherwise doing this has as much value as spamming solely to increase your game apm
Nice guide. I was actually thinking about writing similar to this a few days ago :D. This is exactly how i've been doing it for some time, but it is so hard to really pull it off, as workers randomly run away from double cloned patches quite often :/.
If you watch for example MCs PvPs, you will notice that he is very good at stack mining, which makes his build like 1-2 seconds faster (for example MC vs san (or anypro.. not sure) on tal darim)
It is absolutely true that these techniques are a small improvement in a phase, where there is no crucial need for apm anyways. But what would make your post even better would be a calculation like 10 tries without these techniques and 10 tries with them, so we have average numbers how much of an effect this really has.
I don't think there is a big benefit to 'stack mining' before you get 8 workers. The random pulling away from patches being double mined leads to more loss in mining time sometimes then gain from doubling up gets you at times.
The most stable way to micro your workers imo that is quite effective is imo: - do some worker split first, 3 by 3 or 5 by 1 whatever - make sure to mine from the close patches right away, then fill up the rest - double up on the close patches after the first 8 patches.
Doubling up too early hardly has any extra benefits over this method, is way more random and you shouldn't prioritize probe micro over some more important things, such as starting your pylon exactly at 100 minerals (way more important then doubling up).
I'd like to add a further (minimal) tip to your post...
When you split as you recommend, the first worker carrying minerals back will be the one mining from the furthest patch. As such, you can click that one from each set of 3 and rally him to the middle patches. After 2 carries you can stack him without income loss on the closest patches.
It's not efficient to force close stacking from your original six until after the 3rd carries of your far-mining drones, but if you mine one far-carry and two med-carries, it will be that much faster.
Granted this is not game-changing, but it is 100% an improvement over the standard of either not moving it at all, splitting differently (way worse), or forcing earlier close stacks. Also, the APM requirement during this time is at 6 supply, so you literally have absolutely nothing to do with your APM.
Thanks for posting this. I was going to write something like this but never had the desire to do so, and I think you did a better job than I could have done.
I have been trying to do this for awhile after I noticed pros do it I struggled to find much info though, so big kudos to OP!
One problem which I encountered is that "vigorously right-clicking" the mineral patch on some maps (Metalopolis 3rd and 6th patches e.g.) results in worker bouncing eventually. I use a different method described by Zelniq: http://www.teamliquid.net/forum/viewmessage.php?topic_id=188721#15 which i call stop-trick for those patches. I would like to ask those who figured out how to pair up workers: does it always work for you if you do it like OP suggests, or do you use some other tricks for some maps and patches? Any suggestions are appreciated!
Does anybody have some actual numbers to how much more this gives? I understand that it's something to do in the early game and it's a small edge over your opponent, but I'd be interested to know how much.
Side note, once you get to 16 workers, this doesn't matter, but this could get you to 16 workers faster.
On April 18 2011 00:13 azn_dude1 wrote: Does anybody have some actual numbers to how much more this gives? I understand that it's something to do in the early game and it's a small edge over your opponent, but I'd be interested to know how much.
Side note, once you get to 16 workers, this doesn't matter, but this could get you to 16 workers faster.
Once you get to 12 workers it doesn't matter (assuming you doubled up efficiently).
Otherwise you can get to 18 without doubling the all close patches if you leave it to the worker AI.
My advice is go to a map like XNC (bottom if you're zerg) which has the prototype layout, then run "no micro" versus "full perfect worker micro" and do some basic opener.
You'll find you can gain upwards of 8-10 seconds over split, and upwards of 15 seconds over box+rally and herpderpAPMspam.
This isn't 8-10 seconds that dissolves over time, it's 8-10 seconds that stays with you the entire game.
If you're doing a 6:00 push and you've done the build 100 times and it's perfect, but you always boxed your workers and rallied them to the middle patch and never micro'd them, you can easily do that exact same push at ~5:50.
The mineral advantage suffers exponential decay as you approach 16-18 workers, but the time advantage stays with you throughout the entire game.
On April 16 2011 19:12 DarKFoRcE wrote: Nice guide. I was actually thinking about writing similar to this a few days ago :D. This is exactly how i've been doing it for some time, but it is so hard to really pull it off, as workers randomly run away from double cloned patches quite often :/.
If you watch for example MCs PvPs, you will notice that he is very good at stack mining, which makes his build like 1-2 seconds faster (for example MC vs san (or anypro.. not sure) on tal darim)
There's a way to use the rally points to hit timings where the spawned probe will double up on a patch and not bounce away. Though the rally point method only begins to be useful at probe number 9 iirc. I'll need to actually post here the rally orders and timings but it is specific to protoss as chronoboost allows the timings to let you hit auto mineral patch doubling on the closer patches. Wont have time until wednesday though to actually write down properly what I am trying to explain.
On April 13 2011 05:42 tehemperorer wrote: It really only matters in PvP in my opinion, and even then, it factors in to about .01% of the whole game. It has been discussed to death, however. Here's an example of my split that I believe is the best way to do it, but isn't optimal on ST, TDAltar, or SPlateau because of the mineral patch locations:
My opinion is that it is better to occupy all mineral patches before doubling up, since if you make a mistake you have drifting probes which equal no mining at all. I prefer in this case to play it safe and not lose mining time from the beginning of the game.
I find that it is equally useful in Terran MUs because of the need for workers to be away from mining to build, it gives a bit of an edge when you can squeeze a bit more out of your early game.
it's not just PvP; for Zerg (as someone mentioned) the faster you can hit 50 minerals in the early game, the faster you can make drones. with proper stacking i can make 2 drones at the same time after my 9 ovie spawns, with too much bouncing i have to wait a second or two between the drones. it's not a big deal, but as cecile mentioned, it's best to practice good habits!
Just curious but you should be able to do this anyway unless you're 9 drone scouting. Perhaps this allows you to get the 2 drones simultaneously with the 9 drone scout, but without it you should be getting this benefit anyway.
The time that this removes from my openings is amazing. All of my builds feel incredibly more fluid with no weird pauses where I throw down a building and can't QUITE afford a worker (most notably 15th probe after 13gate/14assimilator with chronoboosts at 11 and 13).
imo, the best way to split is like this: if you're familiar with the F1 key and it's functions, you know that with each click you move to a different probe that is idle, a probe that is doing nothing.
The actual method to split:
so, if you would like to split faster, as any race, you will hold the S/E/S+D key to make a worker, and after that you will click CTRL+F1, and what this will do is box all your probes, because they are doing nothing, and you will be free to split as explained in this thread.
The only reason I worker split is just to get my mind and my hands ready for the game. There's something about splitting them into 3 and 3, rather than boxing them, that makes you feel more 'prepared' in a sense?
It's purely mental for me. There's such a negligible difference to me.
Also @Mazer on previous page, I always tend to make my worker first, then split, because it's easier to click the hatchery (or whatever building you have) and press a button, than to do the slightly finer motions of selecting and splitting workers. To me, it's just FASTER to do it that way.
I think the problem I have with spitting and stacking is that if you mess it up badly enough it can actually cost you the game so I wouldn't suggest doing these on ladder until you have it down in custom games first. Also I'm not sure how much this actually helps your econ or if it is mostly just a psychological gain most of the time. However on some maps the mineral patch are set up horrible so spitting and stacking is almost required. Also I personally like to have a little calm before the storm at the start of a game to try and relax before the click fest begins so that is the main reason why I don't split or stack unless it is required by the map.
I think also knowing when to cut probe production and all that jazz could be discussed here. Because if you are 1 basing you don't need more workers once your fully saturated and a lot of the times I will watch a replay with someone who was clearly one base play that had like 10+ extra workers just bouncing in the mineral line.
I've been stacking my scv's like this for a few months now and it makes the standard terran opening (ie. 10 depot 12rax 13refinery) alot smoother for me. When I pull this off close to perfect I miss ZERO scv training time. More specifically my depot finishes before my 11th scv so no supply block occurs and I have enough minerals to throw down my rax and my refinery and make my scv's. When I'm too lazy to do this I usually am supply blocked for a few seconds and either have to delay my rax and refinery slightly or lose about 5 seconds of scv training time. Arguably this isn't much but it makes everything feel much nicer so I prefer it. Also I am not a spammer so this gives me something to do at the start of a game .
I do this every game (as posted in the OP)...but I notice that maybe 40% of the time, a probe will bounce off to a different patch after 3-5 mining trips paired up. ie, it looks as though I've got things paired up perfectly, but after several mining trips, one of the probes will just decide to wander. In the worst cases, the wandering probe will steal another 'occupied' patch, causing that worker to wander...within a few seconds, all of my precious mining pairs are gone. Anybody know why?
On April 19 2011 23:47 Chemist391 wrote: I do this every game (as posted in the OP)...but I notice that maybe 40% of the time, a probe will bounce off to a different patch after 3-5 mining trips paired up. ie, it looks as though I've got things paired up perfectly, but after several mining trips, one of the probes will just decide to wander. In the worst cases, the wandering probe will steal another 'occupied' patch, causing that worker to wander...within a few seconds, all of my precious mining pairs are gone. Anybody know why?
I think it's because it doesn't take exactly 1 trip's worth time to return the minerals so if there are patches open they still might break off (until you have at least 1 probe on all of your other patches). The probe thinks that a naked patch will be more efficient for this reason so it take more babysitting before all patches have a worker. I'm currently doing some testing on mineral stacking shaving off time on a 4gate and my results so far have been pretty crazy. stay tuned!
On April 13 2011 05:32 islandman wrote: Interesting.
Does splitting in SC2 actually affect the mining capability of the player? i.e would a player that doesn't split have the same amount of minerals at a certain point in time?
The second you gain 5 minerals .5 seconds faster then the other guy you are at LEAST that much ahead until something changes so yeah you can stay ahead the whole game by that amount assuming no mistakes from him or you. so if he doesnt stack his probes and you do... lets say it nets you 20 extra minerals over a 3 minute period thats 20 extra minerls you will have ahead... which effects your build order timings making them faster.. etc... its small.. but in a game as competive as sc2 gets in master league EVERYTHING matters.. I have had people just leave a game after 1 battle where they didnt break even and they knew there was no easy way to get back into the game.
On April 21 2011 07:42 Tallbus wrote: Actually splitting gives absolutely no advantage. Experiment after experiment has been done proving the lack of advantage gained from splitting.
Here are some benchmarks for the 4gate while I'm paying very close attention to mineral stacking (ie first probe out of the nexus gets stacked and the two that are mining slightly farther patches are corrected after a few trips to line up)
0:50 first pylon 0:56 10th probe (the one that get slightly cut after you put down pylon) 1:37 gateway (on 12, probe scouts) 2:44 core (17) 3:36 wg research starts 4:07, 4:09, 4:14 additional gateways added 5:22 warpgate finishes, warpins begin 5:27, 5:27, 5:27, 5:33 additional stalkers finish (you get a little capped on gas)
I will soon post benchmarks of a 4gate where there is no split and workers are rallied without any attention to mineral stacking (first 8 probes take unique patches, rally is not changed to assist stacking)
Becnhmarks for 4gate without ideal split (all 6 to one) nor attentional to mineral stacking
0:50 first pylon 0:58 10th probe 1:37 gateway, scout 2:49 core (17) 3:40 wg research starts 4:14,4:15,4:19 additional gateways added 5:27 warpgate finishes, warpins begin 5:32, 5:33, 5:34, 5:36 additional stalkers finish
So, you eek out aboutt 3-5 seconds on most of these timings. Some probes start to incidentally pair up after you have about 10 probes, so the advantage you get seems to really come from the first minute or so of the game. The benchmarks that are most significant are the times that wg starts (3:36 vs 3:40), and when the last gate is added (4:14 vs 4:19) as these timings are what give the 4wg its strength. I ran each test 4 or 5 times so this research is by no means flawless but I think it's pretty clear you are rewarded for babysitting your workers.
Been doing this myself. Although my APM is too low to worker stack past the 15 supply or so. However, worker splitting DOES work. I usually gauge my split on whether or not there is any delay from my 1st worker to my 2nd worker. A good split means you will have your 2nd worker already queued before the 1st one is finished being build. Whenever I look at replays, my workers are always a bit faster than my opponent.
Was watching Artosis' stream earlier and couldn't help but notice the nonexistent worker micro....
On April 21 2011 07:42 Tallbus wrote: Actually splitting gives absolutely no advantage. Experiment after experiment has been done proving the lack of advantage gained from splitting.
Where is you're link? where is your evidence...
There was a thread on this, the general conclusion was that it doesn't matter too much.
Well to test the efficiency of this you could just play a game standard with
a) no order, no split b) splitting, no order c) splitting and order
let say saturate a base to 16 workers let them mine for 1min. and then count the Minerals to have discrete numbers of the effect. so we can evaluate the effects to Income to a specific timing.(+XX Minerals)
Don't know if the economy graph is accurate but you could get that information easily by counting how much minerals have been mined by counting the minerals patches
I do this after reading ths thread a couple of weeks ago.
Its not gamechanging, but its definitely something that keeps your APM high without spamming mindlessly. And I just love it when my 13th drone is able to be built right as the larva pops off ( normally its around 30 minerals, with efficient mining I usually get 45 or even 50 if Im on top of my timings).
Looking at the timings posted, on certain maps this could give you an insurmountable lead(mostly PvP only). On scrapstation and tal'darim, when perfectly executed if you can finish warpins when the other player is just starting to warp is huge. Just one or two free hits on warping units is all you need.
I think the OP also ommited a point to only pull probes off far patches to mine gas so that close patches are still stacked. This will lead to roughly a 100 extra minerals mined by the time your cybercore completes. This has the largest impact in the 3 stalker robo build where usually you will have to cut probes and chrono out 3 stalkers to reach 26/26 before building a pylon to unblock your supply. You will be able to prevent a supply block if you use my method mentioned above.
I dont believe that the OP's probe split is the most efficient too. I think the most efficient method is to send 4 probes to the close mineral patch nearest to the middle of the whole field and the other 2 probes to the 2 close mineral patches on the other end, then follow on the stack from there. I usually plop my pylon down at 0:48 using this method.
On April 19 2011 23:47 Chemist391 wrote: I do this every game (as posted in the OP)...but I notice that maybe 40% of the time, a probe will bounce off to a different patch after 3-5 mining trips paired up. ie, it looks as though I've got things paired up perfectly, but after several mining trips, one of the probes will just decide to wander. In the worst cases, the wandering probe will steal another 'occupied' patch, causing that worker to wander...within a few seconds, all of my precious mining pairs are gone. Anybody know why?
I've noticed that this happens only on some of the maps. So, it makes me think that most "close" mineral patches will maintain the stacked mining because the patch is at the optimal distance for the mining trip to be equal with mining time. However, some maps have patches that are too close. These patches have a distance where the mining trip is less than the mining time, and thus the worker bounces off. One map that comes to mind where this frequently happens is Metalopolis on the outermost mineral patches of each starting position.
You can cut around 5 seconds in being able to produce the crucial 3 roaches by splitting, stacking workers and using only far-patch workers in the gas transfers
for terran, if you do this perfectly, I've found that it actually smooths out the build times significantly in the early game. When I do it right, the completion of my supply depot and the start of the 12th scv is exact which gets rid of bumps in scv production in the early game. It's a minor thing that helps you optimize build orders and times so even though it won't have a huge difference in your gameplay. It feels good and it gets the game rolling smoothly! I'd recommend doing it
FWIW, in the OP it would be more accurate to say "antisynchronized" instead of "desynchronized". The latter confused me, since the goal is to align two probes mining the same patch.
I registered JUST to tell you this, I am a sad nerd.
Yup, good read. Started doing this after watching 12 Weeks with the Pros PVP with Socke. He said since you're not doing anything, might as well do this.
Already in beta we had a huge topic about splitting and I'm pretty sure it was found that the difference between splitting and letting the computer split is VERY small. We're talking less than 10 minerals.
Awesome OP, definitely going to start stacking after reading this, I always watched pro level games and assumed it was difficult to do for some reason so never bothered trying to learn it.
For all the people making repetitive posts on how small the edge you gain by doing this is, I think the OP and anyone with any semblance of intelligence reading this thread knows this. He isn't saying you're going to win the game because you worker split, I don't know where in his post there is anything that you could glean that message from. He is saying it gives you a slight mineral edge very early in the game that in some cases with some builds will speed you up by seconds. If after reading the OP you feel the need to say '' in beta we had this discussion and it only gives you like 10 more minerals '' please don't.
Frankly, unless you constantly fuck it up, I don't see a good reason not to do this. You aren't doing anything anyway, not at least attempting this is lazy.
As it has already been stated, worker pairing at the close mineral patches helps a lot to improve your build, specially with strong timing attacks. However, I don't believe that is the more efficient way of pairing the workers at Xe'lNaga Caverns, here's a link to a youtube video demonstrating the way I believe is best to pair and split the workers at that map:
On April 13 2011 05:32 islandman wrote: Interesting.
Does splitting in SC2 actually affect the mining capability of the player? i.e would a player that doesn't split have the same amount of minerals at a certain point in time?
From my experience, splitting correctly allows you to start your second worker instantly. Not splitting does have about a 1 second delay which is huge in really tight builds.
I can't believe how many people don't think mineral stacking and splitting helps... Just try it in a game and check how many minerals you have in a key timing, ie when you throw down your additional 2 gateways in a 3 gate expo build. You will have more minerals...
On April 13 2011 07:28 Kornholi0 wrote: I may not have blue post or high masters rank or been in hundreds of leagues... But I can tell you this. Idra never lost a game cause he didn't split his drones. This basically means that sure, you get a "lead" but that lead is instantly lost once you hit the mid-game. Once you hit certain variables where 5 minerals no longer is an issue (maynarding probes, oversaturation, loss of probe etc) You having that lead means absolutely nothing. It is a good habit, and it actually was worth doing in broodwar because your units did not auto split, but because they do so in SCII there is absolutely no reason to worry about that kind of economic lead.
Purest form of this example, if 2 zergs 6p the one who splits his drones and the one who doesn't split his drones will have lings at the exact same time OR if this example holds true and you get an extra lead (by no means can you actually apply it but w/e) you should have 3 sets of lings FASTER than the other guy. But the point is you both 6pooled, which means you both have 6 lings anyways, which means that you didn't actually get ahead or behind by having those lings 0.05 seconds sooner.
I don't mean to downvote your post, I think it is a great idea to split guys, it gets into good habit and causes a certain flow to your builds. I would look more to how that helps your timing for your first buildings more than if it helps you get more workers... It doesn't create economic lead as you both have the same exact minerals regardless at the end of the day.
lol maybe the reason you don't have a blue post or high master rank is because of this attitude? In a game of starcraft, both players starts off the same (1 cc/hatch/nex and 6 workers) where then, does one start building an advantage?
In the game, you are looking to gain every advantage you can get. SO that it slowly builds up and eventually win you the game. Same as in most sports. You know that one point/basket may not necessarily win you the game immediately, but you still want those advantages. And if we go by the logic of your last sentence, then Terran shouldn't be using mules at all since they'll end up 'with the same exact minerals regardless" right?
The only plausible reason for not attempting this is that, your effort can be better spent elsewhere improving other aspect of play that will yield a greater advantage.
When I first started doing this I was probably messing myself up more than I helped, but its gotten more or less pretty smooth now.
It is also a good warmup for handspeed and keeping track of probes.
Some things it helps with for me as a protoss are: 1. It lets me get cybercore out using certain earlier scouting patterns (like for pvz and pvp) with less of a probe production break (or none at all in most cases). 2. it lets me go 13 gate, 14 assim, 15 pylon without any delay in probe production (and minimum delay if any on a 9 scout; I may not remember correctly)
However, I'm still having some trouble as my probe desynchronizes from time to time and I don't really catch it.
Also guys, its not the action of splitting that gives you the increased minerals, its the increased mining efficiency because of 2 main reasons:
1. probes mining faster from closer patches than further 2. not having probes walking around looking for empty patches early on
I am sure this has been linked already but if it hasn't, go read the In-depth Mining Analysis guide by Malloy. It's a great read and puts some numbers behind the affects of worker pairing and which patches to mine first.
I had just seen and heard of this technique a few times before I tried it, and ever since then, I always have 10 minerals+ if I don´t mess up (have had times I just felt like doing it faster, and ended up putting half of them behind a patch, idle, not that big of a deal as my opponent was really terrible, but kinda stings on your ego )
perfectchaoss United States. August 03 2011 03:48. Posts 23 PM Profile Quote # Does this have a negative effect later in the game when 1/3 of our mineral patches are mined out much earlier than the rest?
Well, they wont be mined out that much earlier as long as you keep making drones/probes/scvs, but if they are, you probably got bad luck with saturation (i.e. having all of workers mine go to same patch when finished, resulting in sometimes they stack up there quickly).
But no worries, you just simply send some of to your expo, and you wont lose that much mining (if any at all, sometime you actually earn more minerals on that happening)
I made a custom map where each player immediately splits their workers according to one of the common strategies. You can find it on EU Battle.net under the name "Split Testing Map" by Indigo. All you have to do is start the game with six players or AI opponents and then watch the replay for analysis.
There are three methods that I think I implemented well: 6x1 - 1 worker to each of the close spots 2x3 - 3 workers to each of the distant spots (as recommended by l46kok) 1x6 - 6 workers to closest field
There are three more methods that are not set up quite right: 3x2 - 2 workers to three different close fields 4+2 - 4 workers to close field, 2 workers to edge field 5+1 - 5 workers to close field, 1 worker to edge field
I made a simple spreadsheet where I recorded mineral totals for each player for each of the first 30 seconds: Google Doc spreadsheet
I am really looking for suggestions on how to improve the remaining strategies, any new strategies that I should try, or how to get better timings for each mineral return (SC2 Gears?).
Worker splitting will never change the outcome of a game. Never. I'd suggest practicing other stuff than losing you time trying to find the split that will give you a 5 mineral advantage.
On August 07 2011 22:04 IndigoCZ wrote: I made a custom map where each player immediately splits their workers according to one of the common strategies. You can find it on EU Battle.net under the name "Split Testing Map" by Indigo. All you have to do is start the game with six players or AI opponents and then watch the replay for analysis.
There are three methods that I think I implemented well: 6x1 - 1 worker to each of the close spots 2x3 - 3 workers to each of the distant spots (as recommended by l46kok) 1x6 - 6 workers to closest field
There are three more methods that are not set up quite right: 3x2 - 2 workers to three different close fields 4+2 - 4 workers to close field, 2 workers to edge field 5+1 - 5 workers to close field, 1 worker to edge field
I made a simple spreadsheet where I recorded mineral totals for each player for each of the first 30 seconds: Google Doc spreadsheet
I am really looking for suggestions on how to improve the remaining strategies, any new strategies that I should try, or how to get better timings for each mineral return (SC2 Gears?).
On August 07 2011 22:04 IndigoCZ wrote: I made a custom map where each player immediately splits their workers according to one of the common strategies. You can find it on EU Battle.net under the name "Split Testing Map" by Indigo. All you have to do is start the game with six players or AI opponents and then watch the replay for analysis.
There are three methods that I think I implemented well: 6x1 - 1 worker to each of the close spots 2x3 - 3 workers to each of the distant spots (as recommended by l46kok) 1x6 - 6 workers to closest field
There are three more methods that are not set up quite right: 3x2 - 2 workers to three different close fields 4+2 - 4 workers to close field, 2 workers to edge field 5+1 - 5 workers to close field, 1 worker to edge field
I made a simple spreadsheet where I recorded mineral totals for each player for each of the first 30 seconds: Google Doc spreadsheet
I am really looking for suggestions on how to improve the remaining strategies, any new strategies that I should try, or how to get better timings for each mineral return (SC2 Gears?).
EDIT: Fixed up some grammar
Did you take work production into account? If you split 3x2 and then build a worker, it might be slower/faster than the other way around.
On August 08 2011 02:06 Hoon wrote: Did you take work production into account? If you split 3x2 and then build a worker, it might be slower/faster than the other way around.
Nope. The splits in this test are instantaneous so it wouldn't make a difference.
On August 08 2011 01:20 Flying_Cake wrote: Worker splitting will never change the outcome of a game. Never. I'd suggest practicing other stuff than losing you time trying to find the split that will give you a 5 mineral advantage.
if you need to take any amount of time to actual practice this stuff you've got some other problems. 3-3 splitting is easy to do and gives you just a tiny edge
On August 08 2011 01:20 Flying_Cake wrote: Worker splitting will never change the outcome of a game. Never. I'd suggest practicing other stuff than losing you time trying to find the split that will give you a 5 mineral advantage.
This is false, if it makes a difference in your mineral count, it can, and does affect the game.
Never is a not correct, extremely rarely probably is.
5 minerals 1 second earlier can mean a unit is out 1 second earlier.
On August 08 2011 02:06 Hoon wrote: Did you take work production into account? If you split 3x2 and then build a worker, it might be slower/faster than the other way around.
Nope. The splits in this test are instantaneous so it wouldn't make a difference.
oh... well thats interesting, because splitting 1x6 is 6 actions, while 6x1 is only 1 action.
So the advantage that 1x6 brings you, is practically impossible to do. You would have to select and select a mineral node 6 times in less than 1 second.
On August 08 2011 02:06 Hoon wrote: Did you take work production into account? If you split 3x2 and then build a worker, it might be slower/faster than the other way around.
Nope. The splits in this test are instantaneous so it wouldn't make a difference.
oh... well thats interesting, because splitting 1x6 is 6 actions, while 6x1 is only 1 action.
So the advantage that 1x6 brings you, is practically impossible to do. You would have to select and select a mineral node 6 times in less than 1 second.
Not me...
Great stats to see non-the-less.
ty
You can F1>Mineral click 6 times in a row in 6 different mineral patches. Of course, it multiplies your chance to miss the split by roughly 6 times, so it just looks cute when playing with friends obsing.
Yeah, great stats, but can't be applied in the actual game, even tho splitting doesn't make such a difference.
I think that it is really important to realize that if you can send all six workers in 2 seconds(doable with 200+ APM), the average delay will be 1 second and your split may actually be better than 2x3 at some timings.
It's not like you have anything better to do those first few seconds, why not use this technique to get a few extra minerals? People hating on this idea are being ridiculous.
On August 08 2011 03:19 Hoon wrote: You can F1>Mineral click 6 times in a row in 6 different mineral patches. Of course, it multiplies your chance to miss the split by roughly 6 times, so it just looks cute when playing with friends obsing.
Yeah, great stats, but can't be applied in the actual game, even tho splitting doesn't make such a difference.
I use Select Idle Worker and 6x click on minerals for my split and after some practice it's not common to miss a mineral patch and I haven't noticed it being slower than boxing 3x2 spilt. Anyone who wishes to try this, don't be discouraged! :D It's a neat trick and the posted results show it's the most efficient worker split.
How I do it:
I've found that going in one direction gives you the best consistent accuracy and speed, while identifying and clicking the nearest patches also help.
The quickest way to practice; Start up a game in a build order tester/custom game, and slowly go through the motions of the worker split. Queue worker->hold Select Idle Worker->slowly click each mineral patch deliberately and accurately. Keep going through the motions until you feel solid with the movements and accuracy of your clicks. After ~20 games you might find yourself doing it quickly every time.
Hilariously, it outputs APM as ~600 for the worker split, so it's really cute when friends are obsing.
On August 08 2011 01:20 Flying_Cake wrote: Worker splitting will never change the outcome of a game. Never. I'd suggest practicing other stuff than losing you time trying to find the split that will give you a 5 mineral advantage.
This is false, if it makes a difference in your mineral count, it can, and does affect the game.
Never is a not correct, extremely rarely probably is.
5 minerals 1 second earlier can mean a unit is out 1 second earlier.
if you're getting 5 minerals a second then yes you have more important things to improve upon
in the real world people not only float money above what they actually need, but even if they didn't that 5 minerals will come fast enough as long as they're building workers. Thus any advantage gained by 5 minerals from splitting is only useful to people who haven't thought about what 5 minerals actually means in real game situation terms