|
UPDATE: Added full 16 pool/15 hatch BO
Alright TL, I'm going to start this discussion off with an opening request. Please try and be open-minded about this post. Right now, everyone just "knows" that hatch first is more economical than pool first. But is this really true? Abandon your assumptions and preconceptions and let's analyze this objectively!
So, I've been using Lomilar's "Evolution Chamber" BO optimizer for zerg a lot and noticed that it almost never goes hatch before pool. No matter what kind of build I give it, (allin cheese, 1 base timing push, 2 base timing push, super macro 3 base saturation, and even a build where the only goal is to get 60 drones), 99% of the time it goes pool before hatch.
Now, I haven't tested it objectively in game yet (maybe some ambitious TLers can help me out here), but it certainly warrants some discussion. Here are some discussion questions:
What are the advantages of getting a hatch before pool?
1 hatch + 1 queen actually gives more larva than 2 hatch and there's a period of time where you can't support all the larva from 2 hatch + 1 queen, so you don't want to get 2 hatch too early.
One advantage is earlier hatchery creep spread for spine crawler placement. I could see hatch before pool builds being important on maps like delta quadrant or xel'naga for this reason.
As for economy, a 2nd base doesn't start being more efficient than 1 base until you have >20 drones on minerals. So by this logic, you don't really need to start your 2nd hatch until about 21 food.
One final thought is that the BO optimizer doesn't do creep tumors. Perhaps using creep tumor with the first 25 energy gets rid of that deadzone of time when you can't support all the larva produced by 2 hatch + 1 queen, because she doesn't spit the first time. This may be the main answer to this discussion, but maybe not. Only objective testing will tell.
Anyway, if pool before hatch turns out to be more economical than hatch before pool, this is actually great news for us zerg. We can have our cake and eat it too! We can do the economic build and have a faster pool for possible defensive lings. Here are two build orders that Evolution Chamber has suggested:
16 pool
17 hatch
16 overlord
16 queen
21 gas
16 pool
15 hatch
17 gas
16 overlord
16 queen
+ Show Spoiler +I tried to make this build as realistic/flexible as possible by requiring early zerglings for defense. Ideally, this build should be the most economical build possible while still being able to hold off early pressure. 9 Overlord M:104 G:0 16 SpawningPool M:202 G:0 15 Hatchery M:300 G:0 17 Extractor M:30 G:0 16 Overlord M:102 G:0 16 Queen M:155 G:0 18 Zergling M:108 G:0 20 +1 Drone on gas M:8 G:0 21 +1 Drone on gas M:103 G:8 21 +1 Drone on gas M:103 G:8 25 Overlord M:328 G  3 25 Queen M:228 G 3 27 Lair M:154 G:106 32 +1 Drone on minerals M:1 G:46 35 Overlord M:100 G:70 35 Zergling M:50 G:75 36 Zergling M:400 G:105 37 MetabolicBoost M:350 G:105 37 Zergling M:250 G:5 38 Overlord M:200 G:5 38 +1 Drone on gas M:100 G:5 39 Zergling M:50 G:5 40 Zergling M:54 G:10 44 Overlord M:143 G:40 48 Zergling M:134 G:69 49 Zergling M:84 G:69 50 Zergling M:240 G:88 51 Zergling M:190 G:88 53 Overlord M:113 G 0 53 Zergling M:102 G 7 54 Zergling M:52 G 7
The first build is a little more economical, while the second is a little more cautious (has extra larva for defensive lings if necessary, quicker gas for ling speed). In both builds, the 16 pool actually goes down sooner than the 14 pool in 15 hatch/14 pool builds! Pool first allows you to drone longer and not have to cut as many drones to make your hatch (you don't cut any drones for a 16 pool, and you cut far less waiting for the 300 minerals for your hatch). If you get pylon/cannoned rush or marine/bunker rushed, your pool is closer to finishing than it would be with 15 hatch/14 pool!
So, I'd love for some TLers out there to compare these builds to 15 hatch/14 pool and 14 hatch/13 pool. I'm not sure the best way to compare but here are some ideas. Play a mirror match with a friend with you doing one build and your friend doing the other and see who can saturate both bases first. Test which build is better at holding off rushes. And I'm sure you creative guys can think of other/better ways as well.
But in the meantime, how does this sound on paper? Any glaring flaws I'm missing or will we be seeing a lot more 16 pools in the future?
TL;DR: 16 pool/15 hatch is more economical and safer than 15 hatch/14 pool
|
As for economy, a 2nd base doesn't start being more efficient than 1 base until you have >20 drones on minerals. So by this logic, you don't really need to start your 2nd hatch until about 21 food
.
This logic is flawed. You want the second hatch to finish once you hit 21 drones, if efficient mining is your overall concern.
|
Pool first is 100% better than hatch first in every single scenario. Since there are pro gamers that hatch first, many people will mindlessly copy without question.
If you run a side-by-side comparison and race to 50 drones, you'll find that hatch first and pool first both reach it at about the same time. The only difference is, pool first grants you much more flexibility and is a safer opening build.
There is absolutely no reason to hatch first in any matchup, on any map, ever.
|
On November 27 2010 08:50 Dragar wrote:Show nested quote +
As for economy, a 2nd base doesn't start being more efficient than 1 base until you have >20 drones on minerals. So by this logic, you don't really need to start your 2nd hatch until about 21 food
. This logic is flawed. You want the second hatch to finish once you hit 21 drones, if efficient mining is your overall concern.
I agree with you, but I said 21 food, not 21 drones. 21 food would be 1 queen, 2 zerglings and 18 drones. If you have an extractor, that leaves only 15 drones on minerals. So if you start your hatch at 21, it will finish around the time you're at 26 food (20 drones on minerals), right on time to be efficient.
|
maybe economically it's always better, but you want faster creep on your natural to defend it, don't forget that.
|
On November 27 2010 09:07 navara wrote:
maybe economically it's always better, but you want faster creep on your natural to defend it, don't forget that. This argument is self-defeating. If your opponent is doing some sort of build that is SO intensive that you NEED that creep early, it probably would have been better not to have expanded early anyway. If you hatch first, you forfeit the ability to make that strategic decision without wasting minerals and delaying your spawning pool.
|
On November 27 2010 08:56 Steven.Bonnell.II wrote:
Pool first is 100% better than hatch first in every single scenario. Since there are pro gamers that hatch first, many people will mindlessly copy without question.
If you run a side-by-side comparison and race to 50 drones, you'll find that hatch first and pool first both reach it at about the same time. The only difference is, pool first grants you much more flexibility and is a safer opening build.
There is absolutely no reason to hatch first in any matchup, on any map, ever.
This pretty much settles the discussion in my opinion. Hatch first leaves you open to way to many different cheese methods.
|
Instead of debating and rambling on why doesn't a zerg player or op just test out the 3 builds and give us some times/data so we can look at and actually KNOW which build is the best.
|
On November 27 2010 09:08 Steven.Bonnell.II wrote:Show nested quote +On November 27 2010 09:07 navara wrote:
maybe economically it's always better, but you want faster creep on your natural to defend it, don't forget that. This argument is self-defeating. If your opponent is doing some sort of build that is SO intensive that you NEED that creep early, it probably would have been better not to have expanded early anyway. If you hatch first, you forfeit the ability to make that strategic decision without wasting minerals and delaying your spawning pool.
What if it's ZvT and the T is doing a 2rax build? On maps with a small choke building 1 or 2 sunkens shortly after the hatch finishes leaves you free to drone up on 2 bases and for the most part shuts down your opponent's ability to rush. If you pool first you won't be able to have a sunk in time, so you won't be able to expo for a while, and there's no way that massing lings off one base to defend is better than having the early hatch and sunkens.
|
On November 27 2010 09:08 Steven.Bonnell.II wrote:Show nested quote +On November 27 2010 09:07 navara wrote:
maybe economically it's always better, but you want faster creep on your natural to defend it, don't forget that. This argument is self-defeating. If your opponent is doing some sort of build that is SO intensive that you NEED that creep early, it probably would have been better not to have expanded early anyway. If you hatch first, you forfeit the ability to make that strategic decision without wasting minerals and delaying your spawning pool.
yeah, was thinking more of the expand at 20 food, i really find it to be too late to be safe (creep wise and depending on your scouting information). but then you'r right in the fact that the spawning pool is done anyway by this time.
|
Another point is that you can spread Creep faster if you Hatch-first and can get away with double-Queening once your Spawning Pool finishes.
I also believe that there's a big benefit in getting a Hatch first if you can hold onto it, because it doesn't give the opponent an opportunity to prevent it going up later in the game.
|
On November 27 2010 08:50 Dragar wrote:Show nested quote +
As for economy, a 2nd base doesn't start being more efficient than 1 base until you have >20 drones on minerals. So by this logic, you don't really need to start your 2nd hatch until about 21 food
. This logic is flawed. You want the second hatch to finish once you hit 21 drones, if efficient mining is your overall concern.
Okay hatch takes 100 secs to build, drone takes 25. Larva spawn every 15. Lets say we time this perfectly and drone 21 pops when the hatch finishs. Before we start the hatch we drop 2 gas. 100 secs - 25 for the 21st drone = 75sec. 15 secs per larva = 5 larva when the hatch is building. so when the hatch first begins you need 16 drones, or 17 before the hatch begins. 17 hatch seems really late for a hatch first build.
|
I personally hate hatch first builds for the very vulnerability that was on display during this GSLs Ro64, so I'm all in favor of this post!
I'm also inclined to gas/pool first so as to discourage my opponents from going 2 rax+scv/cannoning my ramp. Upon seeing the gas, most opponents tend to assume standard play.
|
On November 27 2010 09:02 Cambam wrote:Show nested quote +On November 27 2010 08:50 Dragar wrote:
As for economy, a 2nd base doesn't start being more efficient than 1 base until you have >20 drones on minerals. So by this logic, you don't really need to start your 2nd hatch until about 21 food
. This logic is flawed. You want the second hatch to finish once you hit 21 drones, if efficient mining is your overall concern. I agree with you, but I said 21 food, not 21 drones. 21 food would be 1 queen, 2 zerglings and 18 drones. If you have an extractor, that leaves only 15 drones on minerals. So if you start your hatch at 21, it will finish around the time you're at 26 food (20 drones on minerals), right on time to be efficient.
1:1 Ratio of Minerals to Drones is always the most efficient. There will be a slight lull in your economy when you still have the lack of minerals/queen but early on you can't use the larvae given by the queen anyways. And which sufficient micro two hatcheries is easy to utilize all that larvae and barely have excess minerals.
|
hatch first allows you to have more larva to give to units and drones. Thus, faster unit production, as well as the creep benefits for surround and the 1:1 ration of mins to drones.
That said, I do both, but hatch first is really weak in team games, due to not being able to help ally defend against cheese as quickly. (static defense isn't going to be helping your ally =( )
|
On November 27 2010 09:36 kineSiS- wrote:Show nested quote +On November 27 2010 09:02 Cambam wrote:On November 27 2010 08:50 Dragar wrote:
As for economy, a 2nd base doesn't start being more efficient than 1 base until you have >20 drones on minerals. So by this logic, you don't really need to start your 2nd hatch until about 21 food
. This logic is flawed. You want the second hatch to finish once you hit 21 drones, if efficient mining is your overall concern. I agree with you, but I said 21 food, not 21 drones. 21 food would be 1 queen, 2 zerglings and 18 drones. If you have an extractor, that leaves only 15 drones on minerals. So if you start your hatch at 21, it will finish around the time you're at 26 food (20 drones on minerals), right on time to be efficient. 1:1 Ratio of Minerals to Drones is always the most efficient. There will be a slight lull in your economy when you still have the lack of minerals/queen but early on you can't use the larvae given by the queen anyways. And which sufficient micro two hatcheries is easy to utilize all that larvae and barely have excess minerals.
Most efficient? 2 workers per mineral patch for close patches. 3 workers per mineral patch for far patches. If you're going to talk about most efficient don't be that far off... 1 worker per mineral patch over 2 bases is hardly better than 2 workers per mineral patch on 1 base.
The main question is, at what point or when does having better creep spread help you? It will answer when and maybe why we should or should not go hatchery first.
We then factor in creep tumour, or the value of queen energy.
|
On November 27 2010 09:08 Steven.Bonnell.II wrote:Show nested quote +On November 27 2010 09:07 navara wrote:
maybe economically it's always better, but you want faster creep on your natural to defend it, don't forget that. This argument is self-defeating. If your opponent is doing some sort of build that is SO intensive that you NEED that creep early, it probably would have been better not to have expanded early anyway. If you hatch first, you forfeit the ability to make that strategic decision without wasting minerals and delaying your spawning pool.
You're actually totally wrong. Look at 4-gate timing for protoss. You won't have had creep up for long enough to lay a tumor and expand from it if you expand at 20 food, but you will if you go 14/16 or hatch first. This is actually a huge deal against a 4-gate as if you lose your natural against a 4-gate you lose the game. Not saying a speedling expand can't hold, but its' much easier if you can spread your creep.
|
while i would say its better to get pool first against protoss im quite sure an early hatch is better against terrans
against protoss with pool first you avoid getting cannon rushed, force the toss to get a zealot and a sentry first and ofc you get rid of a pylon that might block your nat
its quite terrible against terrans imo even though its often necessary - against strong marines rushes you need spine crawlers or banelings and those rely hard on creep
|
On November 27 2010 09:51 MoreFasho wrote:Show nested quote +On November 27 2010 09:08 Steven.Bonnell.II wrote:On November 27 2010 09:07 navara wrote:
maybe economically it's always better, but you want faster creep on your natural to defend it, don't forget that. This argument is self-defeating. If your opponent is doing some sort of build that is SO intensive that you NEED that creep early, it probably would have been better not to have expanded early anyway. If you hatch first, you forfeit the ability to make that strategic decision without wasting minerals and delaying your spawning pool. You're actually totally wrong. Look at 4-gate timing for protoss. You won't have had creep up for long enough to lay a tumor and expand from it if you expand at 20 food, but you will if you go 14/16 or hatch first. This is actually a huge deal against a 4-gate as if you lose your natural against a 4-gate you lose the game. Not saying a speedling expand can't hold, but its' much easier if you can spread your creep.
I think this would be a very pertinent thing to test. Can some people try doing the two builds in the OP against different 4 gate timings and see how the creep spread at the nat allows for spine crawler placement? I bet (and hope) that at least the 16 pool/15 hatch build will be ok. I think I'll go try it now, but 2+ heads are better than one.
|
On November 27 2010 08:32 Cambam wrote:
TL;DR: 16 pool/15 hatch is more economical and safer than 15 hatch/14 pool
Is there a reason for such a late pool? You can do 14 pool 15 hatch without wasting any larva or cutting drones.
|
|
|
|
|
|
EPT
