|
Sorry for the weird title, it's pretty hard to make it clear with only 50 characters :D
I want to create a discussion around 2 standards openings in ZvP against FFE : - 11 pool 18 hatch - 15 pool 16 hatch
Against an FFE, you want to take a fast third. But the "fast third" is not the same, depending of your opening.
With a 15p/16h, you must take your third at 4:00. With a 11p/18h, you must take your third at 6:00.
If you want to take a third at 4:00 with a 11/18, you will cut your drones productions and you'll feel a little "weird". It's not very natural.
I made 2 tests with theses openings. I'll just follow this :
- Make 4 zerglings - Take an "optimal" third (4:00 for 15/16, 6:00 for 11/18) - Take 2 gas at 6:00 - Take 1 gas at 7:00 - Start lair with 100 gas (+ put a roach warren, + put an evo chamber) - Start zergling speedling with 100 gas - 8:30 = end of the test
11 pool 18 hatch
* 6:00 - The third begin - 30 drones
* 7:00 - 47 drones - Third not completed yet
* 8:30 - 70 drones - Lair completed - Third completed - 3 queens - 900 mineral, 450 gas + many larvaes (i didn't start an upgrade with the evo chamber)
15 pool 16 hatch
* 6:00 - Third is nearly complete - 28 drones
* 7:00 - 42 drones - 3 queens
* 8:30 - 70 drones - Lair completed - Third completed - 3 queens - 900 mineral, 450 gas + many larvaes (i didn't start an upgrade with the evo chamber)
I tested also the 11p/18h with a third at 4:15
11 pool 18 hatch (third at 4:15)
* 6:00 - 25 drones - Third nearly completed
* 7:00 - 38 drones - Too much larvaes, you can't use them
* 8:30 - 66 drones - 300 minerals and 450 gas - No larvaes
As you can see, if you take your third at 6:00, 11p/18h is exactly the same as the 15p/16h around 7:00.
But theorycrafting is not the only thing in this game. We must check other things like : - 11p/18h is pretty better against an early cheese (canon rush & 2 gates) - With a 15p/16h, you can drone scout and go hatch first against a "one base" protoss (or 14g/14p) - With a 11p/18h, the third is complete around 7:40. So it's harder to defends it against an early pressure (like a zealot rush). But against a zealot rush, you can just cancel it (instead of just lose it). Against a double stargate after an FFE, you got the time to put somes spores crawler.
So i want to discuss with you : I use the 11p/18h in every ZvP. I feel it's pretty good against a "one base" protoss and it's competitive against an FFE style (even with a "fast third").
But, pros players don't use this opening. Even on Tal'darim (with rocks on the third). They use 15p/16h (or 14/16, or 15/18, etc ...).
Why ?
|
I think they use it sometimes, but I've always thought for the longest time that this build is underused.
|
I see pros players use this opening on the ladder (on their livestream). But in tournaments, they don't use it. I wonder why.
|
With my testing I swear I had more than 28 drones at 6:00, do you mind specifying your exact build including lings made for both test scenarios up to 6:00?
As in: 15 pool, 16 hatch, 15 overlord, 15 queen, 17 ling, 18 ling, 3 drones, 22 queen, 24 overlord, 24 hatch @third, etc. Or replays?
Replays in particular would be awesome, seeing how you work out the drone transfers (or not) in both scenarios.
|
Are you making 4 lings when the pool finishes in both cases? I'm curious how that affects the economy, specifically as 11p/18h only gets any advantage if it makes those early lings, but it also is required to hit very tight drone timings to keep up with 15p/15h. The hatchery getting blocked is one of the main reasons for trying 11p/18h, but as DRG showed it is easy to just run a drone to the third base location instead and the 15h after pool is still essentially guaranteed (though possibly vulnerable to cannon rushes as the distance for your workers to defend it is longer).
Personally I like going 11p/18h simply to scout and deny hidden probes around the map early on, as well as try to force players to go forge and cannon before nexus and gate (though if you only make 2-4 lings that isn't likely necessary, they don't always know if you will make 6 or do a full on all-in with drones etc). It's also quite possible to deny all knowledge of an expansion or gas timings in some cases, which can really throw off a protoss and may make them play overly defensive.
|
On March 07 2012 18:07 aebriol wrote: With my testing I swear I had more than 28 drones at 6:00, do you mind specifying your exact build including lings made for both test scenarios up to 6:00?
As in: 15 pool, 16 hatch, 15 overlord, 15 queen, 17 ling, 18 ling, 3 drones, 22 queen, 24 overlord, 24 hatch @third, etc. Or replays?
Replays in particular would be awesome, seeing how you work out the drone transfers (or not) in both scenarios.
Well maybe i'm not very good with the 15p/16h and i should got more drones at 6:00 ^^ But i don't think i can have more drones at 8:30.
I'll try to put the 3 replays as soon as i can ( = when i'm at home ^^ )
|
I think the main thing is number of larva. The real thing that differs is the fact that you get an extra spawn larva with an earlier hatch. By experience, i have a lot of trouble defending 8:30 pushes with 11/18, if I dont over prepare early. Im thinkin in particular the zealot void push. You wont have any spores up in time at your third. Also, i find that 7 gate blink +2 is harder to hold, but that could just be that the 15 pool build is more my style.
|
On March 07 2012 18:16 oOOoOphidian wrote: Are you making 4 lings when the pool finishes in both cases? I'm curious how that affects the economy, specifically as 11p/18h only gets any advantage if it makes those early lings, but it also is required to hit very tight drone timings to keep up with 15p/15h. The hatchery getting blocked is one of the main reasons for trying 11p/18h, but as DRG showed it is easy to just run a drone to the third base location instead and the 15h after pool is still essentially guaranteed (though possibly vulnerable to cannon rushes as the distance for your workers to defend it is longer).
Personally I like going 11p/18h simply to scout and deny hidden probes around the map early on, as well as try to force players to go forge and cannon before nexus and gate (though if you only make 2-4 lings that isn't likely necessary, they don't always know if you will make 6 or do a full on all-in with drones etc). It's also quite possible to deny all knowledge of an expansion or gas timings in some cases, which can really throw off a protoss and may make them play overly defensive.
Oh yes i forgot, i made 4 zerglings each times.
And yes with a 15p/16h you can just take your third if the protoss is trying to block your natural.
|
i think testing this with a pylon block at the natural would be most interesting. it´s more or less standard to face these pylon blocks imo. i could imagine that 11p 18h is even stronger there. on the other hand side a 15p 16h player can place the hatchery at the 3rd instead, that´s ok too. and it would be interesting to know what´s the better build for early ling speed (gas after hatch),i would guess 15p 16h is better here.
|
But theorycrafting is not the only thing in this game. We must check other things like : - 11p/18h is pretty better against an early cheese (canon rush & 2 gates) - With a 15p/16h, you can drone scout and go hatch first against a "one base" protoss (or 14g/14p) - With a 11p/18h, the third is complete around 7:40. So it's harder to defends it against an early pressure (like a zealot rush). But against a zealot rush, you can just cancel it (instead of just lose it). Against a double stargate after an FFE, you got the time to put somes spores crawler.
15p and 11p will both beat 2 gate just fine, even if proxied... Also, a cannon rush will win against an 11pool if you don't pull drones just the same as 15 pool, as lings will never come out in time.
You can drone scout either way against Toss. Yes, as a percentage, I understand what you are driving at, but either way it's still just 50 minerals lost, and 1 drone. Not a big deal.
Someone has tested this stuff before, and basically, the 'order' of how builds stack up economically, will remain the same if you do the same thing, ie drone scout at same time, take gas at same time, et cetera.
Also, double stargate will hit wayyy after third is complete no matter what build you do, or when you take third.
I used to 11p/18h every game in ZvP. It's a horrible build. I did it for a year until I realized how bad it was. The only merit in 11p is that it's 'better' if you remain on 1 base, because you have more larva, and less minerals, but on 1 base, you'll have enough econ for the larva production. But, with 14 pool, or 15 pool, or whatever, you build up excess minerals, that you can't spend.
Unless you were planning to spend that excess minerals on something other than larva, like a hatchery.
And that's why 14 pool is way better than 11pool. I'm not sure about 15p vs 14p. Also, I dislike your tests. Your third timing is so, damn, arbitrary. It would be better if you did a test where you took the third at some exact same time each time, like you took it at 4:00 (or whatever it is). Or maybe if you took it at 30 supply each time (ensuring same drone count). Maybe just try both those methods, see how they both compare.
Also, some more things:
1. Leave out the gas timings. People have already proven that, if you are comparing two builds, as long as you drone scout, or take gas, or whatever, at identical times in both builds, the relative order will stay the same. So the rankings of say, 6 pool, 8 pool, 10 pool, 14 pool, is still the exact same and by the same margin, if you do those same builds, but with gas taken at 1:00. You make it more complicated, what you just need to prove is which is better build.
2. No lair...
3. No evo chamber...
4. No speed...
5. End the test at 9:00, it will be more stable to then.
6. You need to have very, very specific overlord timings, and do it exactly the same in each game, and time them perfectly. As in super, super fucking perfect.
7. What map you choose is extremely relevant. I recommend you create a custom map with a single base, and 3 mineral fields in 3 different directions to it, and the 3 hatches 'stacked', or just super close. Maps make a huge difference in this.
8. You need to have a process for maynarding workers, because that's a BIG fucking deal. You need to perfectly split untis up. So you need to figure out, with perfect worker production and overlord production, how many drones you have, and where to send the drones. The other guy actually figured out how many drones to send where each time. Maynarding drones costs mining time, and if you don't maynard evenly, the results get skewed because 14+14+14 is like 20% more income than 21+21, even though it's same number of workers.
I really like what you are trying to do, but: 1. Someone has done this before, and better (not to say someone shouldnt make an update) 2. Your testing is all fucked up (at the very least, thirds at complete different timings, and taking gas????)
Just don't like the test method. Why don't you just do a test that tests which is better for 2 base. Then, do a separate test, on what time is best to take a third (because nothing definitive has been done on that).
Also, the things you say about 11/18 being better, is totally irrelelvant. I thought you were trying to test which is better economically. Why are you even mentioning stuff like cannon rushes and such? It almost seems like you are biased and just trying to tell everyone "hey, I think 11/18 is the best build in the world, why isn't everyone doing it, they are wrong, and here is some arbitrary test I did".
And, as I said, 11/18 is completely irrelevant to all the stuff you said anyways. A cannon rush from 13 forge at 17 supply, either 3 pylon ramp block or 2 pylon behind minerals, will kill you and win, if you wait on lings, and react any differently than if you went 15 pool. The lings will never be out in time to stop it (they only shorten the timing that the cannon rush can happen, but as soon as Toss throws 3 pylons to block ramp, or 2 pylons to block behind minerals, you lose). Same goes for proxy 2 gate, you can even hold that hatch first, as long as you respond appropriately (im pretty sure hatch first can beat proxy 9/9, not 100% sure on that, def. can beat like 11/13 or 11/12 proxy, you just cancel hatch, make 2 spines, make lings, and you still end up ahead because its so all in).
Yea your testing is just so goofy. Lings, gas timings, maynarding?, no replays, overlord timings?, map?, lair, speed, evo? wtf. If you want to compare the economics of 2 builds, just go pure drones and necessary queens. If you make 4 lings after pool, or gas, at same time in each build, it will still stack up roughly the same. It's not like 15/15 is better than 14/15 if you make 4 lings, but if you dont make 4 lings 15/15 is better, or gas, or whatever. That's just ridiculous. Everything about your testing method is suspicious, and goes against what other people have tested saying that 11/18 is a horrible build, and banks zero money for your expo (unless you make idle larva, which you do during 11/18), and that 14p is way better.
|
Btw that very early lair is not necessary IMO, id upgrade first if I were you. Also i find that if I go quick 3rd gas at 7:00 I will die to loads of stuff. Watch out if thats your go-to build.
|
What about the 12pool variant ?
Nestea popularized this opening a while back, I personnaly use it a lot, for the above stated reasons regarding the choice of 11p; earlier Ling to deny any hardcore proxy cheese, and early hatch for a pool build.
Link to the 12p build
|
^ earlier pool/ling does NOT deny any cheese.
|
The only thing earlier pool does is deny nexus first on certain maps. Because really, Toss can always go nexus first vs 14 or 15p, except certain maps like metal
|
Also, I dislike your tests. Your third timing is so, damn, arbitrary. It would be better if you did a test where you took the third at some exact same time each time, like you took it at 4:00 (or whatever it is). Or maybe if you took it at 30 supply each time (ensuring same drone count). Maybe just try both those methods, see how they both compare.
I tried to take a third at 4:00 with a 11p/18h. It's very bad. You must delay the second queen or delay the third. In all case you must cut the drone production.
4:00 and 6:00 are not "arbitraty". If you spend all your larvaes, you'll have a 300 minerals excess. So you can take your third. Somes players use the 11p/18h to make double evo + 4 gas at 6:00 (instead of a third).
1. Leave out the gas timings. People have already proven that, if you are comparing two builds, as long as you drone scout, or take gas, or whatever, at identical times in both builds, the relative order will stay the same. So the rankings of say, 6 pool, 8 pool, 10 pool, 14 pool, is still the exact same and by the same margin, if you do those same builds, but with gas taken at 1:00. You make it more complicated, what you just need to prove is which is better build.
I just tried to test it as a "proper response to FFE". If i full drone (without gas), the 11p/18h will be behind. Because you'll reach 32 drones befores the third is complete (so more than 2 drones by mineral patch). And your income will not be optimal. With the gas, you reach 32 drones before the third is complete but it's later.
If you do only drones, you'll have way more minerals with the 15p/16h at 8:30. But if you take gas, you'll have the same ressources. Because of the bases saturation.
2. No lair...
3. No evo chamber...
4. No speed...
i agree with that, but i don't think it'll change something. Because in both opening, i can't have more larvae. So it'll only change resources at the end.
5. End the test at 9:00, it will be more stable to then.
Why ? Around 7:30, opening are at the same point. There's no reason it'll be different at 9:00 :s And against an FFE, a zerg should stop drones around 8:30 (depends of the protoss choices).
6. You need to have very, very specific overlord timings, and do it exactly the same in each game, and time them perfectly. As in super, super fucking perfect.
Well, i'm not a super perfect player as i'm only low master. But i think i can drone without supply blocking me ^^ But i see what you mean. If we want a super perfect thing, we should try it in evolution chamber program.
7. What map you choose is extremely relevant. I recommend you create a custom map with a single base, and 3 mineral fields in 3 different directions to it, and the 3 hatches 'stacked', or just super close. Maps make a huge difference in this.
All tests are on Shakuras Plateau.
8. You need to have a process for maynarding workers, because that's a BIG fucking deal. You need to perfectly split untis up. So you need to figure out, with perfect worker production and overlord production, how many drones you have, and where to send the drones. The other guy actually figured out how many drones to send where each time. Maynarding drones costs mining time, and if you don't maynard evenly, the results get skewed because 14+14+14 is like 20% more income than 21+21, even though it's same number of workers.
Same response as 6. I'll try in evolution chamber ^^
Also, the things you say about 11/18 being better, is totally irrelelvant. I thought you were trying to test which is better economically. Why are you even mentioning stuff like cannon rushes and such? It almost seems like you are biased and just trying to tell everyone "hey, I think 11/18 is the best build in the world, why isn't everyone doing it, they are wrong, and here is some arbitrary test I did".
I am biased. Because i use 11p/18h all the time so i feel better with this one. It's normal. I created this topic to have "counter arguments" on that.
I don't want to just say "hey 11p/18h is better, no matter what". I want to start a discussion on this and i want arguments for the 15p/16h (or 15/15, or 15/18, or whatever).
|
On March 07 2012 18:58 Magus.421 wrote: I am biased. Because i use 11p/18h all the time so i feel better with this one. It's normal. I created this topic to have "counter arguments" on that.
I don't want to just say "hey 11p/18h is better, no matter what". I want to start a discussion on this and i want arguments for the 15p/16h (or 15/15, or 15/18, or whatever). With 15p 16 hatch 24 pool, you will have 1 or 2 creep tumors out (depending on how you do the build) without sacrificing economy much it seems.
Do you have any spare creep when you do 11 / 18?
Anyway, I would love to see the replay when you get home, then I'll try to do add in a test of my own doing what I am biased for (10 drone scout,15p 16h 24h after overlord).
|
|
On March 07 2012 19:05 aebriol wrote:Show nested quote +On March 07 2012 18:58 Magus.421 wrote: I am biased. Because i use 11p/18h all the time so i feel better with this one. It's normal. I created this topic to have "counter arguments" on that.
I don't want to just say "hey 11p/18h is better, no matter what". I want to start a discussion on this and i want arguments for the 15p/16h (or 15/15, or 15/18, or whatever). With 15p 16 hatch 24 pool, you will have 1 or 2 creep tumors out (depending on how you do the build) without sacrificing economy much it seems. Do you have any spare creep when you do 11 / 18? Anyway, I would love to see the replay when you get home, then I'll try to do add in a test of my own doing what I am biased for (10 drone scout,15p 16h 24h after overlord).
With 11/18 you got one tumor (with a third at 6:00). If you want to take your third at 4:00, you can have two tumors.
But in both openings, you can make an extra queens for creep spread if you want.
|
noob question, why not hatch first?
|
4:00 and 6:00 are not "arbitraty". If you spend all your larvaes, you'll have a 300 minerals excess. So you can take your third. Somes players use the 11p/18h to make double evo + 4 gas at 6:00 (instead of a third).
You need to take the third at the same time. So compare 16/15 with 3rd at 7:00 to 11/18 with 3rd at 7:00.
I just tried to test it as a "proper response to FFE". If i full drone (without gas), the 11p/18h will be behind. Because you'll reach 32 drones befores the third is complete (so more than 2 drones by mineral patch). And your income will not be optimal. With the gas, you never reach 32 drones on minerals before the third.
try testing it instead of theorycrafting it. Because theoretically, you would think that 16p/15h would be better than 11/18, but if that's wrong according to tests, well, that would be interesting.
If you think 32 drones is critical, why not do a test where you do 16/15 where you grab third not based on time, but at certain drone count (ie 36 supply) and then 11/18 at identical supply.
It would interesting to see what is more important, or made for better results. I would think taking them at different times ultimately changes the build, but taking it at identical times keeps the 'ordering' of the builds same.
I also think testing for third is goofy. Ultimately you are getting at what's better, 11/18, or 16/15. Try testing 11/18 vs 16/15 first. Then, do testing to see when is the best time to take a third. Or, maybe, what's really important, is not what opening you do, but at what drone count you take your third.
i agree with that, but i don't think it'll change something. Because in both opening, i can't have more larvae. So it'll only change resources at the end.
No, it won't change anything, that's the point. If it's irrelevant, and all you are doing is testing what opening is better, leave it out. The time it takes for the drone to travel to plant the geyser or evo, what drone you select, et cetera, does make an impact. Leave out such variables, it just makes the test more complicated. It only makes the tests less accurate, not more.
We understand that in a real game, you need to make lings, gas, et cetera. But tests done before, have shown that, taking gas, or whatever, at identical times, keeps the orders of how builds stack up, the same. So just leave it out. You can always revisit the idea after you do more solid tests. I would think, that's it's better to take such things based on supply/drone count, rather than time though.
Why ? Around 7:30, opening are at the same point. There's no reason it'll be different at 9:00 :s And against an FFE, a zerg should stop drones around 8:30 (depends of the protoss choices).
A lot of the closer builds manifest differences around then, and often times certain builds don't manifest improvements until much later. It's odd. Someone did else this, check his results.
Well, i'm not a super perfect player as i'm only low master. But i think i can drone without supply blocking me ^^ But i see what you mean. If we want a super perfect thing, we should try it in evolution chamber program.
I've tried evo chamber, I'm actually the last person to post in the thread, trying to test this exact same thing (when is it best to take a third?). I'm going to work on it a bit, and maybe you can help me bring interest in the idea of using it to optimize economy rather than unit timing attacks, but so far I've had no luck there.
If you posted replays, it would be more helpful. That's all. Or maybe post when you make overlords - 26/28, et cetera. Look over your games and make sure you made them at the perfect times, and identical times. I know the 2nd and 3rd bases pop at different times, but for the most part, it should be same supply that you make overlord.
All tests are on Shakuras Plateau.
Try a custom map maybe? If you take a faster third on one build over another (an issue solved by my suggestions), maynarding a bunch of drones earlier can have a larger impact. You want to minimize those issues.
I am biased. Because i use 11p/18h all the time so i feel better with this one. It's normal. I created this topic to have "counter arguments" on that.
And I am biased because I used to do 11p/18h for a whole year, and then realized the build sucked ass. It all dawned on me when someone did a very normal, 13 forge, 17 cannon rush, pylon ramp block to me. The main reason I did it was because I thought, hey, I get more money because instead of going 14 pool and having to set one drone on the probe, and another patrolling the ramp, I can go 11 pool, and just keep both drones mining. The asshole did the ramp block, and the lings never came out in time. They actually popped when the cannon popped, and then they died before taking down the shields. This was on top of me using drones to attack the pylons too.
And I've never lost to a proxy 2 gate when going 14 pool. In fact, I've gone hatch first many times against proxy 2 gate and did just fine. I cancelled the hatch, the pool was on the way, and then I made 2 spines in base and then cancelled the one they attacked when they were 99% done, just like you do when you go 11 pool or 14 pool. Get gas, get speed, win.
If 11p/18h is better because of economics, great, I'd love to learn that. But the idea that 11p is safer against any sort of cheese, is absolutely wrong. Hatch first is safe against any proxy gate, I believe, and I can say with 100% confidence, that 15 pool is definitely safe against proxy gates. 11 pool, is just as vulnerable to cannon rushes, as 15 pool. If those 3 pylons ramp blocks go up, you lose. If toss puts 2 cannons behind the mineral line, and you don't react the exact same way as you would with 15 pool, you lose. The only reason hatch first is autoloss to cannon rush though, is because the lings come out so late, that Toss has a good 700+ banked up, and can throw down like 5 pylon/cannons that overwhelm the drones and the lings are just way too late to end the timing, but a 15 pool will get the lings out in time to deal with when toss has 700+ minerals banked to just throw shit down.
I originally started doing 11/18 because of aggressive reasons, against 3 gate expands. I then got out of low masters, and realized that any half decent toss, will easily hold off 11/18 speedling aggression. As for 1g expands, 14/14 is way better timed to deal with it than 11/18 and better for econ, and besides, you can't know that toss is going 1gate expand vs 3 gate expand until after you would have to make the lings to force cancels anyways. And, on top of that, even if you do force 1-2 cancels on the nexus on such 1/3 gate expands, you have to make 2 x larva injects of lings, or about 14 lings, or 7 drones worth, to stop it. It kills your economy so bad, that even if you force 2 cancels with speedling pressure like that, that you actually do more damage to yourself than to the toss who has a 1 minute later expo (i tested this).
So 11/18 is not good for any aggression. It's not any better than 15 pool for dealing with cheeses. So the only other issue is econ. And, the only way 11/18 is better, is if you don't expand, because 14 intentionally pools money so you can take an expo without cutting drones, while 11/18 cuts drones. It just hurts you economically if you plan to play on more than 1 base.
But then, all the 1 base roach all-ins in PvZ are 13 pool, or 12 pool, or some other pool, not 11/18, so even then, it's a shitty build.
If you can prove it's better economically, go for it. It is counter-intuitive, but that doesn't necessarily mean anything - after all, 16 hatch is worse than 15 hatch, and 15 hatch is worse than 14 hatch, when it comes to hatch first builds. But so far, you haven't sold me. It doesn't exactly work that you are basically just testing when is the best time to take a third, and changing the 11/18 third's timing until it's better than the arbitrary time you picked for the 16/15. What's to say that a better third timing wouldn't improve the 16/15 over the 11/18?
Why dont you just test 2 base of 11/18 vs 16/15 first, see what's better. It would stand to reason that, whichever is better as 2 base, woudl be better for 3. Then, do a test for when is the best time, or supply, to take a third. Or, compare 11/18 third vs 16/15 third, and take the third at the exact same drone counts, rather than times.
You should really check out the "A look at zerg openers thread". This has all been done. I think 11/18 was discussed in 'A second look...."
|
|
|
|