|
Firstly, this is a problem, which I'm sure many zergs across the board have experienced and most platinum and diamond zerg players are starting to notice. It involves the correct way to spend your larva.
During a game, when you are pitted against a Terran, and they go for a crippling 3 rax timing push, to put you at a disadvantage for the rest of the game. (it is assumed you see more than one rax/late expansion, from darting lings up the ramp/scouting overlords)
(It is assumed that you have 2 hatcheries, operational by now)
If, both your larva pop off yout hatcheries at about the same time you expect the Terran to push out, what do you make? Zerglings or drones? Or do you hold your larva and wait and see? Too often I have seen about 9 larva total just pop off my hatcheries, and think that I can sneak in another early round of drones in before the inevitable push. Instantly, the Terran pushes out, and I'm left with a total of about 8 zerglings to hold off an entire Terran army. (You get to use the two larva that appear every 15 seconds, but this quite often isn't enough)
I have brainstormed this for a while and these are some solutions with pro's and con's.
Make drones:
Pros: Obvious economical advantage.
Cons: Lack of lings when Terran decides to push.
Make lings:
Pros: When Terran pushes, you can have between 20 and 30 lings. This lets you hold off the attack with ease.
Cons: You don't have full saturation on your bases, so even if you crush the attack, terran is still in a reasonable position to expand, tech, upgrade, and win the game.
Make nothing and wait:
Pros: Well, you have a giant bank of larva to spend whenever you want. And if the push comes, you can judge how many, lings and banelings you need and make the exact number to hold it off, and make the rest into drones.
Cons: However, in the meantime, you haven’t been mining much due to a lack of saturation.
Make your queens inject larva at different intervals. This is a subtle change and it involves changing your inject cycles such that you inject larva on one hatchery about every 8 seconds.
Pros: Well, it stops you from spending a massive amount on drones and then losing to a large army, and lets you have a constant re-enforcing flow of lings.
Cons: Possibly not strong enough, because essentially, you are only saving about 4 larva. Hard to remember this when you are paying attention to other things.
I can’t really think of a solution to this. The pro's always seem to just have enough zerglings to hold off the attack. Do they go pure drones, and then churn lings out when the push starts? Or do they save a couple larva (4 or so) just so the push can be held off easily?
Please fling your ideas at me, so I can try to develop a reasonable way to approach this.
Cheers and much appreciated,
Deathifier
(I appologise for any spelling mistakes in advance)
|
The Pros spend larvae based on their Strategy, Scouting and Timings.
I am still horrible at Zerg(High Platinum, beating Mid-Diamonds), but the best advice I have gotten from Masters players is:
Know when to Drone. Don't Drone blindly. Don't Drone too much.
This means that:
I need to know the rough timings on opposing pushes.
I need the ability to know what they are doing based on my scouting information.
I need a gameplan so that I know when to drone and when to make units.
I need branches to my gameplan according to what my opponent might do and if things doesn't go my way.
For example in my ZvP/ZvT, if I haven't scouted any ridiculous one-base attacks, I start massing units when my Lair starts. Why? Because at least Protoss 5-6gate push starts around the same time and against Terran I can prevent a fast 3rd as well as get heavy map control and contain him because my strategy doesn't involve Mutalisks. I also get a 3rd at that point so the units are to protect that as well.
|
there is a day9 vod about droning. i suggest you look into that
|
yes, it's hard to judge, but I try to decide upon 3 things:
-) how many drones do I have/need for what I want to do (70drones after 10mins gives you great income, but if you don't have expansive tech, it is plainly useless)
-) how big is my opponents army and can I hold with a single extra round of units (scouting the front)
-) am I ahead (overproduce units) or do I have to catch up (take a risk... e.g all-in or overproduce drones to catch up)
what a lot of players oft don't realize: people like NesTea and Idra who seem to purely produce drones are metagaming pretty hard (besides their units being more efficient than others, due to better control).
|
Check a number of replays where they do a 3rax push. Note down the time. Next game, if you scout a 3rax, drone hardcore until a few seconds before the push and start banking larva. Once you see them see their base, go into full ling/bling production and destroy their attack.
It can be alluring to build some lings "just to be safe", but due to the increasing returns nature of Zerg economy, it is really damaging to build anything but drones in the early game.
|
I suggest you to avoid massing zerglings because they put a big pressure on your early larva economy because they eat a lot of larva which you could use to make drones.
1 larva from which you make zerglings uses only 50 resources. Instead make roaches which are more larva/cost effective 1 larva uses 100 resources but take 2 population or better make banelings which are more larva/cost effective 1 larva uses 150 resources. Or make 1 spine crawlers where 1 larva uses also 150 resources and frees one population.
Later when you'll have enought larva, massing zerglings will be no problem.
|
Cheers for all the great help guys, I especially appreciate the last one. I have always thought of roaches as bad in ZvT and im going to go try several games with them now. This makes a large about of sence and im always having trouble with hellion, so thanks! :D
|
Haha, this thread is really helpful, I can't say how many times I've done a complete drone pump with my newly spawned larvae only to see that a big terran push is coming.
I've started being more roach-based in my early game zerg and having lings at their choke to see the push timing. If terran isn't showing signs of expanding, I'll usually start banking larvae in preparation for some kind of push. I also sacrifice an overlord at ~5:00-6:00 to see their rax count, tech paths, etc.
|
Ok, thanks for the help. 
I was hoping other people had fallen into the same position.
|
I don't know if any zerg's do this but I kinda thought about it a bit and the idea is to spend all the larva on the one hatch while just leaving a bunch of larva on the other. That way you get the 1 larva regen constantly but you can also save up to a tun of larva if you happen to ever need it. I'm not sure if you could really just not spend hardly any larva on that one extra hatch and still produce enough in fact I'm sure you couldn't but it might be nice to have a hatch with 1000 extra larva just sitting around.
|
The way I do it is spend most of the larva on drones, if im going to NEED lings i'll get them when I need them, or early enough to be safe. just so I wont be wasting larva.
Otherwise i'll be slowly adding on static defense, just enough so that I won't die if he all-ins. Good scouting, knowing timings, and knowing when not to over-drone are good ways to not die to this sort of thing.
Maybe, after you are ~10 workers ahead go for something that will keep you safe like extra queen for creep spread if he doesnt push out, anti air or transfuse if needed. Or get evo chamber and early upgrade if you plan on needing a few spores if cloaked banshee is a possibility.
If you are already 10-15 workers ahead then is there any reason not to stop droning? Work on being safe vs other things, cuz once you get into mid-late game zerg gets a lot stronger (at least my playstyle :D)
1.3k master zerg/protoss
|
Like many people should generally do, I always focus on keeping larva count on each hatchery at 2 or lower as often as possible (if not, then temporarily just do it for 1 hatch, no point to have both above 2 when you can have 1 above 2 one under 2).
That said, my personal style is also not in favor of spine crawlers at all, hence I rely on an overlord (sometimes sacrificial) as well as some early lings and a gas steal to know what the opponent is up to, and build roach or ling if I need defense, or drones if I think I'm safe.
Generally even then though, unless confident of the opponent's actions, I may get a ratio of units to drones, just a higher drone ratio than a terran or protoss typically gets. It takes practice, but with experience, it's not hard to set up and/or switch between worker-army ratios, such as 3:1, 1:1 and 0:1, depending on the situations.
If I had 6 larva I could order 4 to be drones, then ctrl-click the remaining 2 larva and tell them to morph lings, then shift+#controlgroup (that way I don't need to reselect the larva, and add the lings to my army without having drones in it)
I personally just get a force that will prevent any sort of easy expand for terran, and if they push out with a bigger army, then I'll mass units. If they just push out carefully and/or with strong units, then I can keep droning and have my small force deal with harass like hellions. (good players might do a fast marine-tank push but that's another story)
When you keep larva count at 2 per hatch, that leaves you with a float of 4 if you have 2 hatcheries, plus the chance of 4-8 more popping in time from an inject. Having injects synced at the same time is generally bad (at least when ou have good macro/multitasking/injecting skills), since it makes it difficult to spend the larva in time. It's helpful for macro control, injecting all hatches at the same time, but will result in problems due to the difficulty of spending the larva.
|
On June 16 2011 17:05 xJaCEx wrote:
I don't know if any zerg's do this but I kinda thought about it a bit and the idea is to spend all the larva on the one hatch while just leaving a bunch of larva on the other. That way you get the 1 larva regen constantly but you can also save up to a tun of larva if you happen to ever need it. I'm not sure if you could really just not spend hardly any larva on that one extra hatch and still produce enough in fact I'm sure you couldn't but it might be nice to have a hatch with 1000 extra larva just sitting around.
this is completely wrong. larvae cap out at like 16 for each hatchery. Injects happen often enough for this to be completely unnecessary. not to mention it's completely wrong and wasteful.
|
On June 16 2011 17:26 Eknoid4 wrote:Show nested quote +On June 16 2011 17:05 xJaCEx wrote:
I don't know if any zerg's do this but I kinda thought about it a bit and the idea is to spend all the larva on the one hatch while just leaving a bunch of larva on the other. That way you get the 1 larva regen constantly but you can also save up to a tun of larva if you happen to ever need it. I'm not sure if you could really just not spend hardly any larva on that one extra hatch and still produce enough in fact I'm sure you couldn't but it might be nice to have a hatch with 1000 extra larva just sitting around. this is completely wrong. larvae cap out at like 16 for each hatchery. Injects happen often enough for this to be completely unnecessary. not to mention it's completely wrong and wasteful.
Hatcheries cap at 19 (which is a lot), and it's not 100% wasteful, but it is obviously more wasteful than keeping them both below 3.
For the same reasons that it's useful to stockpile larva on all hatcheries (not many reasons, and not overall good), it is useful to stockpile larva on one hatchery... it is just the balance a person wants,
The main thing about doing it though, is sometimes you might not be able to spend all your larva. In such situations, one should not just blindly start morphing units from all hatchery's larva. The hatcheries with the lowest amount of larva above 2 should be used first to go to 2 (or 1), and any others that can't afford to be spent should be resting there. That way it optimizes the natural hatchery generation of larva.
As Eknoid4 may have been implying:
For anyone that doesn't know the math, queens' inject spawns larva on average at a rate 150% of a hatchery. Both have their slow-downs though. If you're not timely with injects that number will go down. If you're not timely with spending larva AS SOON AS it hits 3 or greater (all larva inject completions), than hatchery regeneration will go down (relatively speaking making queens rate go above 150%)
|
I find it hard to be on board...as a Zerg player, that if you beat a 3-rax push and probably have/get speed, that you cannot deny for a while with your 10+ lings that lived, an expo for the Terran in this scenario. Granted you could focus on your second instead and Drone line mad for the next 3-5 minutes and probably tech to defend the next push, but it could be played either way. You already have map control, and the unit advantage, so you could tech and drone with map control (very smart) or tech a little and drone on a "mid" level while producing more slings or maybe even blings before Tank tech comes around to fully block the expo (situationally smart).
Larva is a resource, and each game can be different, but if I win a push without hatching first, it is smart to hatch and drone while keeping a small army until needed. If I hatched first, it is smart to back up my army to a potentially map-controlling level and then drone my ass off.
This being said, a 3-rax failure can be smacked around by Roaches if they come back quick. You can make A LOT of them when the have too few Marauders if they push again to deal with them effectively. I would say it is smart knowledge for Zergs that Roaches are the tanks, Speedlings are map control.
Off topic, there are 100+ Zerg threads about larva control, and that is the concept Zerg live by "riding the razor's edge" where you need to learn to drone or army (always tech). Scout, and read the attack vs. bases per person scenario and compare it with what you want to do and think is best.
Edit: I got asked by the OP to provide a link to any other threads that commented on Larva being a resource. Its a little old, but This Thread has a few good thoughts if you sift through the other information which is geared towards newer players, but it is a good post overall. Check the spoiler tabs attached to "Step 3 Game Sense" as well as Rules 1 and 3 under "Zerg Specific Learning" I know this edit is already long, but I figure these illustrate a little that not only are there a few basic concepts, there is still no 100% right way to learn game sense, which is what Zerg must rely heavily on as a reactive player. All you can do is win or lose and learn from it.
|
On June 16 2011 17:32 Xapti wrote:
As Eknoid4 may have been implying:
For anyone that doesn't know the math, queens' inject spawns larva on average at a rate 150% of a hatchery. Both have their slow-downs though. If you're not timely with injects that number will go down. If you're not timely with spending larva AS SOON AS it hits 3 or greater (all larva inject completions), than hatchery regeneration will go down (relatively speaking making queens rate go above 150%)
This is also not true. The reason it stacks to 19 is because the 3 hatchery larva regenerate regardless of injects. Hatcheries save 4 injects + the 3 natural larva. Everything over that dies when the inject spawns
|
Going back to the example in the op, I'd choose the hold off the attack with easy option. After you do that you can drone and take a third while terran can't even take his nat safely for a bit.
|
you're in diamond and terrans are still 3raxing you? 3rax is a really inefficient build against zerg, the timing comes out after the zerg has already droned too hard.
If you're really having trouble with 3rax, just do spanishiwa style, its an instant-build order win.
Otherwise, follow the other posters suggestions about when to spend larvae, you should almost always know when the terran is pushing out via pokes, just have larvae ready for production at that point, know the crucial timings, like when he'll have 3 tanks.
|
A few basic principles to follow:
1) Having more than 36 drones on 2 bases before taking your 3rd and 4th gas is a poor use of larva. Once you've got your 3rd up, or a macro hatch, you have plenty of larva to finish getting to maximum saturation. You should use the larva you don't make into drones to make units so that you can safely get your 3rd sooner.
2) Zerglings are a very inefficient on larva but are efficient on supply and money. Roaches are efficient on larva, but not on supply. Queens are effiecient on larva and supply, but not on money, meaning it takes more minerals worth of queens to kill a push than of other units. Spine crawlers are efficient on all three, but can't be used for offense. Banelings' efficiency depends on how good your opponent is. If he spreads his marines well, they can be worse than queens. If he runs on top of a cacoon and doesn't kill it, a single baneling could be all you need to hold a push.
3) You should go into the game knowing how you are going to defend early aggression . Will you make a very early 3rd queen and use queens+speedlings with a lot of creep? Will you throw down a safety roach warren and power drones until the push comes, then make a few roaches to defend? Will you rely on spine crawlers over units? The method you choose will depend on your midgame, aggression you want to put on, and if you scout an all in. Regardless, you want to make 40 drones (the 32-36+buildings/scouting) as soon as possible, so the less larva you use to defend early, the sooner you can enter a strong midgame.
|
Drone drone drone when nobody can attack you.
If you feel something fishy is happening, start 1-2 drones per hatch every 17 seconds and invest the rest in an army.
|
to defend against a timing push u need to know what kind of push its gonna be.
hellion marauder? > put spines and stay on lings
3rax? >baneling nest
reactor hellion?>roaches
early game is about effectiveness of your units. when all that is done, and you have more bases than your enemy, then its gonna be about holding bases while massing an army.
|
Airact has it right.
You need to scout your opponent constantly, until you know exactly what he's doing. This involves scouting with a drone early on to check for rushes & all-ins, and checking if he has 0, 1, or 2 gas. You should be constantly scouting the ramp with lings, seeing what units he has, and what buildings he has. This also involves sacrificing an overlord at the right times (usually between 5:00-5:30 is a good starting point).
Based on what you scout you need to know what your opponent is going to do (or guess  ). Example: in ZvP you scout 5 sentries + 1 zealot at his ramp, and a pylon at the bottom of the ramp. You pretty much know at this point he's going to 3gate expand. Confirm by constantly checking that he starts his nexus, make sure he doesn't cancel it, etc.
Based on what you know your opponent is going to do, you need to respond appropriately.
Examples:
3gate sentry expand -> make drones
1 base blink stalkers -> mass lings
1 base DT -> immediate evo chamber, 1 spore at each base
hellions -> make small amount of lings, block ramp with queens, make spine crawlers to cover nat, sim city your nat, then drone
1 port banshees -> get lair, make more queens, overseer when lair finishes
2 port banshees -> make at least 2-3 more queens, drone up
This process of scout, know, respond needs to be done multiple times, so that you know when to drone even in the mid-game.
Example: protoss does 3 gate sentry expand, but that's just his opening. He still has multiple mid-game options once his expansion is up, you need to scout his tech and see what he's getting:
-more gates -> 5/6 gate attack
-doesn't get both gases at nat -> most likely 5/6 gate attack
-twilight council -> 2 base blink stalkers (confirm by constantly checking his army and making sure he doesn't get stargate or robo)
-robotics facility -> GW + immortal, or GW + colossus. Confirm by checking if he makes a robotics support bay after the robo-fac.
-stargate -> some combination of phoenix/void ray harass
You need to scout, know all of these builds, and respond appropriately to them. Once you have this process down, you'll know exactly when you can drone, and when you can't.
|
I'm somewhat confused... it seems like you have some weird assumption that you can only make 1 or the other...
I understand that it's easier to build all of one unit... you just hold down the button. But you HAVE THE ABILITY to make SOME of your larvae be drones and SOME of them be lings.
I don't understand how you think that offsetting your hatchery inject times so that you have to re-inject hatcheries more often makes any logical sense. You're trying to compensate for your tendency to be lazy/greedy by thinking that you need to hold down either the D or Z keys. You're getting the same amount of larvae in both scenarios, but in one you're increasing the amount of APM you have to macro with with no actual benefit.
If your 2 injects just popped, build 5 drones and 3 sets of lings. It really isn't hard. If they don't attack, you still made 5 drones! You'll have more larvae coming. Next you use those lings you made to scout around and actually see if things are coming. Pop one up the ramp and see what they've got.
Also, people are being really silly with their explanations of hatchery vs queen larvae injections:
Hatcheries spawn larvae at a rate of 1 every 15 seconds IF there are LESS THAN 3 larvae at the hatchery. If there are 3 or more larvae at the hatchery, the natural regeneration timer is suspended. If you inject, your hatchery will stop producing its own larvae when the inject pops. Once you use enough larvae to bring the number down to below 3, the hatchery starts making larvae again. Injects don't care about the number of larvae already there UNLESS you get to 19 larvae, in which case you are limited to that number.  http://wiki.teamliquid.net/starcraft2/Larvae
I actually don't know what the wiki means when it says that the larvae timer is "suspended." Maybe it's like this:
Hatchery is 10/15 seconds through its cycle (it's going to pump out another larvae in 5 seconds)
Inject pops right then, and the hatchery has 4 larvae while the timer is at 10/15
Player uses larvae so that there are only 2 larvae, when does the next larvae from the hatchery spawn?
- Is it in 5 seconds, because the timer resumed from 10/15
- Is it in 15 seconds, because the timer reset?
|
@ TheSambassador, granted this is an imperfect analogy as well as the fact that SC2 has few 100% truths, but the fact is, as a race car driver, you should either be accelerating, or braking. Driving in any way, outside of shifting gears (coasting if you will), is an inneficient use of your time in what is in essence, a game of inches. Yes you can make a drone or two to keep up with building creation every inject-pop, but I believe the true point is being able to either fully drone or fully army because one or the other benefits you the most based on the way Zerg play rather than dipping your toes into both.
Nobody ever said it was hard to do both, and sometimes it's the right thing to do, but I imagine you yourself either drone or army most of the time yourself.
Also, I find it a little sad there was this much lack of knowledge about how hatch-inject larva spawning works. The "Natural Regen Timer" is not suspended if there are more than 3 Larva at the hatchery, but the hatchery has two separate counts, the (up to) three larva the Hatch makes at a rate of one every 15 seconds and the sets of four added by an inject, reaching a max of 19 combined (3 natural and 16 for the injects). If you had the memory, luck, and insane clicking skills, you could constantly use the "1 every 15 seconds" larva and leave the 4 inject larva there and the 1 every 15 would continue to respawn. Why else do you occasionally have 0 Larva, and then the inject pops so you have 4, and you click over and then there are suddenly (!)...5 larva. Also, the 15 second count starts as soon as you use the first of the three "Hatch specific" larva as it triggers the hatch to start the 15 second spawn as you no longer are at the limit.
|
If he goes for a 1 base timing and you hold it off with some saturation at each base (read, after you have your main and natural partially saturated, go pure unit if he's all in unless you KNOW you can fit in more drones), he'll never win (because over saturation doesn't help), just stop him from expanding, sprinkle in drones and get an evo with a spore at each mineral line and surround his base with overlords, because the only way back in for him is banshees (DT's, Mutas, VoidRays, etc, etc, etc) or a big drop/nydus all in. He can eventually push out to his natural with seige tanks if he's terran, but you're in an overwhelmingly large lead at this point.
It's important to note that after you crush his push, it's tempting to go pure drone, because obviously, you're ahead, you want to get more ahead. If you know from experience you can pull this off, great, if not, just go pure unit occasionally sprinkling in a few drones.
You should listen to ...
http://www.teamliquid.net/staff/zatic/Day90004HOP1WinWithAdvantage.mp3
It's a podcast from brood war, but the same principles apply. Don't go crazy and expand a ton or make a sick amount of drones off 2 base, just continue with your plan, and keep scouting up (sometimes you'll encounter someone who will do some crazy shit like 3 port banshee/viking when they're pinned in a corner like that)
gl
|
|
|
|
|
|
EPT
