The most helpful points you bring up: larva/structure timings, and spreading creep cycles. The problem I see is that usually the structures for which this is most important (spire, hydra den) come enough into the game (~9-10minutes) where it's hard to have all your injects synchronized, where as roach warren timings usually depend on when your queen is being built. I would say that a better way to time the tech structures is with gas/minerals.
The idea to set camera location to creep is interesting and worth trying; I also like that you actively move overlords as you spread creep. Combining the two tasks is helpful as they achieve a similar goal of map vision/control.
The problem I have with your cycle is that you don't include battle/harass, unless that's what you mean by step 4 (unclear in that case). I find myself missing injections, forgetting to spread creep and overlords when I'm microing mutas lings or roaches, and as a result I have to force myself to try and inject in the middle of battle. It's not hard to macro super well when you're left alone, the hard part is to do it while under pressure. That said having an outline for what your ideal macro should look like is always something good to fall back on.
But for something productive, this is very cool and a great resource! :D I think every race actually goes through this as well; it's like a mental checklist of things to look at just to improve macro.
I'm working on doing a few things like this, basically following the -refresh injects -build units -check creep spread(this needs major work!) -do other macro-y, scouty stuff(new tech buildings, new upgrades, new expansions, reposition ovies around base) -build overlords with the first "natural" larva spawned after inject bits.
I'm nowhere near good enough to sync up building or queen timers but everything else should work pretty well.
On October 12 2011 01:42 w3jjjj wrote: Good idea, but cumbersome execution, having to check that single hatchery takes a lot of multitasking especially in the late game with a lot to do.
The idea of cycling around larva spawn is good, I use it, but you don't need to check your hatchery. Remember that your hatcheries have natural spawn larva, if you wait for a full larva cycle then you have natural larva sitting around waiting for the inject larva to pop, which is not efficient. Instead, you should constantly be producing from your natural larva to keep the larva count down, and whenever you get a surge in larva count, that tells you your inject larva came out, which is a marker for you to inject again, spread creep, etc..
This way, you can maximize your larva efficiency and not miss inject at the same time.
Also, for overlords, whenever your inject larva comes out, you should always begin a production cycle with a couple of overlords, this should become a natural habit so that you don't need to keep track of it.
I disagree.
"having to check that single hatchery takes a lot of multitasking" - I don't think a single tap every ~8 seconds is a lot of multitaking.
"if you wait for a full larva cycle then you have natural larva sitting around waiting for the inject larva to pop" - Nope, those natural larave are the ones I'm using for the overlords
"you should always begin a production cycle with a couple of overlords" - Definitly wrong. If you create your overlords as your larave spawn you will have them before you need them. You just used all your larvae, your overlords build in 25 seconds. you now have your new overlords and a supply cap lead, but you only have as many larave as your hatchs, let's assume 2. it's better to delay the overlords to your first larave that spawns naturally after you inject as THOSE overlords will pop just as your larave does.
Time spent on thinking is a limited resource as it slows down your execution. Production cycling mechanics should be automatic with as little thinking time needed as possible, so you can save your thinking time for more strategic decisions. An example, anyone can have super high apm in the beginning by repeating useless tasks that require no thinking, its just clicking speed, but as the game progresses, one's true apm drops much lower because he has to think, and thinking takes up valuable time from execution. The more thinking time you spend on mechanics, such as a macro cycle, the less thinking time you have for making strategic decisions.
I'm not disagreeing with your idea of larva cycle, I said that's what I use, although my advice is to reduce/eliminate tasks in your cycle that requires thinking time.
The first point I raised is that you don't need to spend time to check the hatchery. The point of checking the hatchery is to see the progress of your larva inject, but I'm saying that if you just focus on keeping your larva count down, then anytime when you notice a surge in larva count, that automatically tells you your larva inject just popped. You might not think checking a hatchery involves much thinking but it does, every time you check your brain is measuring the progress of that larva spawn bar and making a decision of whether the bar is nearing its end. I guarantee it slows you down. Whenever you are producing, you brain naturally assess how many larva you are seeing, that is a task you have to do and are already doing, and it tells you the same information as checking a hatchery, making the hatchery idea a redundant task that yields no benefits.
For my second point regarding overlords, your said to use the in between larva cycle time to think about what units to make and how many overlords to make, so that you don't make too many overlords. I'm saying that making overlords is not a task that deserves any thinking time. In the early game you should have a bo ready that prevents you from making too many overlords, as resources are low at this point, but once into the mid-late game, it rarely ever hurts your game to have an extra overlord or two. I'm saying that if you mark your overlord production with a larva inject cycle, always make a couple when your inject pops, you can avoid spend any thinking time at all on overlords and use that time elsewhere, sure you might make an extra few overlords, but the benefit of not having to think about overlords ever in your game far outweighs having a couple of extra overlords flying around, plus it keeps you safe from a supply block if your opponent snipes a few overlords.
In a lot of pro games you will see a player's supply limit continues to grow even when they are exchanging a lot of units, resulting in extra supply not needed. It might seem like a waste of resource, but this is exactly what I'm talking about, production cycles are fully automatic for the pros they just don't waste time thinking about things like "oh l lost some marines so I don't need to make a depot for a while". They'll notice "oh I have extra supply now from exchanging units and not stopping depot production", then they'll chill on the depot for a short while and go back to full automatic mechanics again.
That said, as long as you are not wasting time thinking about overlords, you can mark the timing of your overlords to anything you like, so if you like to make overlords AFTER you finish an inject production cycle, and using natural larva to do so, by all means. I simply mark my overlord production to the inject cycle because it's my habit. As long as it's automatic and you don't waste time on it, it's the right way to practice.
wow w3jjjj, That was fascinating and extremely well said! I got to admit I've never taken "Think-time" into account as a resource. Fascinating perspective!
We seem to be much closer in our opinions then what I thought you ment in your previous post. I have to admit that with time I found myself looking at that single-hatch display less and less, and using the cycle of overlord-larvae-overlord-larave more. But, this was written as a guide and that was how I started and I consider an important step in the learning process. It gives you a feeling of the urgency required by your macro, as well as a feel for the timing of the spawnns.
In regrads to going over-supply I would say this should be as streamlined as any other element in the game, if people do crazy worker splits and agrue BO diffrences that come to dozens of minerals, or avoid queueing marines, then going over-supply is just as bad, and we'll both agree that nothing is worse then being supply capped.
The way I minized the thinking with the overlords is to split my production to two types ( 1supply per larave versus 2supply per larave). I don't try to count my larave and do the math, I rely on using all my avilable larave and knowing that I can either get 4 or 8 supply from each larvae spawn.
I remapped 0 to spacebar, and always keep a single hatchery on it, so I can watch larva progress by tapping spacebar. this frees up 4 for more army control groups.
OMG this is so good to me thank you very much, I try to find this type of article about zerg a lot , I am a master protoss, recently switch to random, so I try to play zerg + terran a lot on other account, I have a hard time with Production cycle and hot keying with zerg because I did not know the timing well and the race is basically about macro queen hatch and larva, terran is OK for me.
On October 12 2011 05:02 IcemanAsi wrote: In regrads to going over-supply I would say this should be as streamlined as any other element in the game, if people do crazy worker splits and agrue BO diffrences that come to dozens of minerals, or avoid queueing marines, then going over-supply is just as bad, and we'll both agree that nothing is worse then being supply capped.
I think w3jjjj's point is that over-supply usually happens after or around the 10min mark, i.e. when you can't dedicate thinking time to decide how much supply you need for your production. The point of worker splitting and early BO optimization is that the starcraft economy grows exponentially, so that perturbations (i.e. differences in economy value due to different builds/splits) from early on grow larger with time; on the contrary perturbations due to oversupply at the 10 minute mark are not going to be as significant by 10-15 minutes.
If your thinking capacity is finite, perhaps you can actually achieve better economy by doing something other than oversupply.
Of course no one wants to be wasteful with oversupply but developing the skill of dealing with it is not a priority given how well people are playing around 10-15minutes right now.
This sounds familiar to the "tick" that I heard talked about on some Day9 daily I believe. Basically all pro-gamers have a tick going on in their head where they follow a set execution of things. Such as checking supply, then you check your unit production, and then you check your macro stuff (did I mule, inject, build drones), and finally scout.
Basically works like a state machine, where you are constantly moving through different states (you'll understand if you're a computer engineer :D). This is how I macro as zerg. It works really well with the notification that larvae inject is done. Obviously you need to prioritize things efficiently. I think inject is priority number 1, then scout, then decide what to make (drones, overlord, units). Constantly cycling these states is essentially this cycle you are talking about. Hmm, maybe I'll have to go make a finite state machine that simulates this .
On a side note, I found that constantly hitting larvae inject timer makes zerg VERY OP. If you are struggling with zerg, the first thing you should do is work on nailing this timer each and every time at least into the mid game.
On October 12 2011 01:42 w3jjjj wrote: Good idea, but cumbersome execution, having to check that single hatchery takes a lot of multitasking especially in the late game with a lot to do.
The idea of cycling around larva spawn is good, I use it, but you don't need to check your hatchery. Remember that your hatcheries have natural spawn larva, if you wait for a full larva cycle then you have natural larva sitting around waiting for the inject larva to pop, which is not efficient. Instead, you should constantly be producing from your natural larva to keep the larva count down, and whenever you get a surge in larva count, that tells you your inject larva came out, which is a marker for you to inject again, spread creep, etc..
This way, you can maximize your larva efficiency and not miss inject at the same time.
Also, for overlords, whenever your inject larva comes out, you should always begin a production cycle with a couple of overlords, this should become a natural habit so that you don't need to keep track of it.
I disagree.
"having to check that single hatchery takes a lot of multitasking" - I don't think a single tap every ~8 seconds is a lot of multitaking.
"if you wait for a full larva cycle then you have natural larva sitting around waiting for the inject larva to pop" - Nope, those natural larave are the ones I'm using for the overlords
"you should always begin a production cycle with a couple of overlords" - Definitly wrong. If you create your overlords as your larave spawn you will have them before you need them. You just used all your larvae, your overlords build in 25 seconds. you now have your new overlords and a supply cap lead, but you only have as many larave as your hatchs, let's assume 2. it's better to delay the overlords to your first larave that spawns naturally after you inject as THOSE overlords will pop just as your larave does.
Time spent on thinking is a limited resource as it slows down your execution. Production cycling mechanics should be automatic with as little thinking time needed as possible, so you can save your thinking time for more strategic decisions. An example, anyone can have super high apm in the beginning by repeating useless tasks that require no thinking, its just clicking speed, but as the game progresses, one's true apm drops much lower because he has to think, and thinking takes up valuable time from execution. The more thinking time you spend on mechanics, such as a macro cycle, the less thinking time you have for making strategic decisions.
I'm not disagreeing with your idea of larva cycle, I said that's what I use, although my advice is to reduce/eliminate tasks in your cycle that requires thinking time.
The first point I raised is that you don't need to spend time to check the hatchery. The point of checking the hatchery is to see the progress of your larva inject, but I'm saying that if you just focus on keeping your larva count down, then anytime when you notice a surge in larva count, that automatically tells you your larva inject just popped. You might not think checking a hatchery involves much thinking but it does, every time you check your brain is measuring the progress of that larva spawn bar and making a decision of whether the bar is nearing its end. I guarantee it slows you down. Whenever you are producing, you brain naturally assess how many larva you are seeing, that is a task you have to do and are already doing, and it tells you the same information as checking a hatchery, making the hatchery idea a redundant task that yields no benefits.
For my second point regarding overlords, your said to use the in between larva cycle time to think about what units to make and how many overlords to make, so that you don't make too many overlords. I'm saying that making overlords is not a task that deserves any thinking time. In the early game you should have a bo ready that prevents you from making too many overlords, as resources are low at this point, but once into the mid-late game, it rarely ever hurts your game to have an extra overlord or two. I'm saying that if you mark your overlord production with a larva inject cycle, always make a couple when your inject pops, you can avoid spend any thinking time at all on overlords and use that time elsewhere, sure you might make an extra few overlords, but the benefit of not having to think about overlords ever in your game far outweighs having a couple of extra overlords flying around, plus it keeps you safe from a supply block if your opponent snipes a few overlords.
In a lot of pro games you will see a player's supply limit continues to grow even when they are exchanging a lot of units, resulting in extra supply not needed. It might seem like a waste of resource, but this is exactly what I'm talking about, production cycles are fully automatic for the pros they just don't waste time thinking about things like "oh l lost some marines so I don't need to make a depot for a while". They'll notice "oh I have extra supply now from exchanging units and not stopping depot production", then they'll chill on the depot for a short while and go back to full automatic mechanics again.
That said, as long as you are not wasting time thinking about overlords, you can mark the timing of your overlords to anything you like, so if you like to make overlords AFTER you finish an inject production cycle, and using natural larva to do so, by all means. I simply mark my overlord production to the inject cycle because it's my habit. As long as it's automatic and you don't waste time on it, it's the right way to practice.
The only problem I have with all of this is that 'tapping' to check the progress of certain timers is a common action that pros do. It's fundamental mechanic. So to call it a waste of "thinking time" is flat out wrong.
That said, you could tap your hatchery key (5, in this case) and simply look at the larva counts, or your method, constantly attempt to use your larva, so you notice when the larva swell occurs as soon as an inject is done. You're actually carrying out the same action as tapping 4, you're just using a different method.
Counting larva is a great method, and the one I personally use. But, one of the benefits of checking a single hatch (assuming your injects are synced*) is that you can see when an inject is ABOUT to pop, which could be helpful at some specific points in the game. For example, if your units are near the enemy base, or you know you're near your opponent's army, knowing the inject is in 5-10 seconds could allow you to back your units to a safe point, hit the inject, and go back about your business, whereas knowing the inject just FINISHED might happen just as you need to respond to your opponent's nearby army.
Awesome guide, though I'm not agreeing with the specific hotkeys. Like you said though, that's more player-specific. I use my middle mouse button as my backspace key personally, with all queens on ` and hatcheries on f1. My spacebar is my single hatch. Mouse 3 for me is Show Last Message, so that I can quickly snap to "Our Forces are under attack" and other messages. Still, really cool guide.
On October 12 2011 05:56 Nemireck wrote: That said, you could tap your hatchery key (5, in this case) and simply look at the larva counts, or your method, constantly attempt to use your larva, so you notice when the larva swell occurs as soon as an inject is done. You're actually carrying out the same action as tapping 4, you're just using a different method.
Producing from your larva should be done constantly regardless of whether you use any timers. In contrast, checking a hatchery is an extra action on top of producing from larva, therefore it is not another method, it is simply redundant. There is no need to actively count your larva. When you produce constantly from natural larva, its only gonna be a few larva at a time, so whenever you have a larva surge from inject finishing, your brain will "feel" the difference. I use "feel" because I don't ever "count" the amount of larva, may be there's a better way to describe it, but producing from 3 larva "feels" different from producing from 12, I don't know if I had 12 or 13 or 14 larva, I just "feel" that I have made more stuff and therefore an inject must have finished, so I should begin a new cycle.
I use tapping to check on upgrades and buildings in construction, such as a spire, but not for larva inject because from producing stuff I already have the timer in my head, I don't need another tapping action to tell me that my inject has finished. You say pros use tapping, of course they do, but are you sure they use it for larva cycles? Like using a hatchery to check larva spawn? If they do I've never heard of it.
I think your valid point is that checking the hatchery allows you to know beforehand that an inject is about to finish, but I still believe that such knowledge is not worth the extra thinking time spent. As for your example, if you are about to attack an opponent, you are trying to hit a timing, (may be it's to hit before a tech unit comes out, or simply catching a window where the opponent's army is out of position), therefore you should never back off and delay the attack just to do a production cycle... your attack gets delayed a few seconds, some more enemy units finish, their main army gets closer to defend, etc... A major action like attacking should take priority over things like larva inject and creep spread, delaying a few injects is much better than to miss an attack timing and/or to lose a lot of units. That said, you don't really need to shift your attention from the battle to inject larva or produce units. It's a lot of multitasking but that's what separates the pros from the noobs. Pros are producing units even during an engagement. That's why I don't think knowing beforehand when an inject is about to finish is that valuable, certainly not worth the thinking time spent on checking a hatchery.
I have to either get a new keyboard or get use to using 5 and 6 as hatches with 6 being my single hatch, but my 6 feels..too far. Im too used to 4 being all of my queens, I been told from my friends that i have pretty good injections.
On October 12 2011 05:56 Nemireck wrote: That said, you could tap your hatchery key (5, in this case) and simply look at the larva counts, or your method, constantly attempt to use your larva, so you notice when the larva swell occurs as soon as an inject is done. You're actually carrying out the same action as tapping 4, you're just using a different method.
Producing from your larva should be done constantly regardless of whether you use any timers. In contrast, checking a hatchery is an extra action on top of producing from larva, therefore it is not another method, it is simply redundant. There is no need to actively count your larva. When you produce constantly from natural larva, its only gonna be a few larva at a time, so whenever you have a larva surge from inject finishing, your brain will "feel" the difference. I use "feel" because I don't ever "count" the amount of larva, may be there's a better way to describe it, but producing from 3 larva "feels" different from producing from 12, I don't know if I had 12 or 13 or 14 larva, I just "feel" that I have made more stuff and therefore an inject must have finished, so I should begin a new cycle.
It "feels" like more larva because your brain is capable of doing an instant count that tells you it was "more than 3 that time". Counting your larva and "feeling" that it's more larva is exactly the same thing, you're griping on a technicality of semantics.
And yes, pros use tapping to stay on top of their injects. Many pros don't hotkey their queens, they hotkey their individual hatcheries instead and tap them regularly, when an inject is ready, they 55, box queen, inject. Others work off larva counts like we do, but if, for instance, they have a maxed army, you still see them tapping their hatcheries even with no units to make, and a player like Idra, who uses the All Hatcheries on 4, individual queens on 5/6/7 (interestingly, he never makes more than 3 queens, or at least didn't at IEM), will check his hatch key regularly when his army is maxed, I assume to keep an eye on larva counts on his hatcheries and hit his injects. (by the way, his setup not allowing for individual hatchery tapping may be one of the reasons he often misses late-game injects).
EPT
.
