EPT[D] ZvP: 11pool vs 15hatch vs 14pool/16hatch - Page 2
| Forum Index > StarCraft 2 Strategy |
|
Defenestrator
400 Posts
| ||
|
usethis2
2164 Posts
The question is simple: Assuming NR10, is there difference between hatch-first and pool-first? Let's say you're going for 2 base saturation and only ling speed by 10 minute mark. Again, no attack or any army production for that period. | ||
|
Vermiiifuuge
Korea (South)112 Posts
On September 20 2012 16:07 usethis2 wrote: I happened to run into this thread (I don't play Z) but I know Z players are very calculated in early economy management, and I've had this question for a long time. It's not exactly on point to the thread topic, but it's somewhat related so if someone can answer it I'd appreciate it. The question is simple: Assuming NR10, is there difference between hatch-first and pool-first? Let's say you're going for 2 base saturation and only ling speed by 10 minute mark. Again, no attack or any army production for that period. Doesnt really make sense to look at it this way. Everything is always relative to what your opponent does, and the timing of your third would also be an important aspect to take into consideration. | ||
|
velvex
Germany226 Posts
On September 20 2012 08:31 Belial88 wrote: He does it as an aggressive opener, not as a macro opener. If he does not do damage because toss responds properly, then he would be way behind. As a macro opener, 12 pool is terrible. http://drop.sc/255829 I played this as a test against the computer a while ago. The details of what I am doing and the execution might skew the result at the 8:00 minute mark (I make roaches at the end), but it should show that you can get into the 70s quite handily with 12 pool. I also found the replay against nexus first I mentioned: http://drop.sc/255830. My macro is really bad in this game and Protoss is probably overreacting, but it should still show that 4 lings off of 12 pool with 16 drones behind them either get into the base or are able to kill the forge before the cannon kills all lings (assuming 17 nexus). That's probably why most Protosses go for forge first if they see an earlier pool. Also thanks to osiris for doing research on nexus first vs. forge first. I will take a closer look at that soon. And I am sorry if I have turned this into a "why 11/12 pool isn't that bad" thread, but I guess since almost all pro players do 14/15 pool as standard, it should be clear that that can't be bad. | ||
|
Mahtasooma
Germany475 Posts
On September 20 2012 07:27 Defenestrator wrote: In my experience I will get in if the toss goes nexus-first and is not conscientious about walling off, or if they go forge-first and forgo a cannon for some reason. At this point, it's very easy to either stop mining for a significant amount of time or pick off 3-4 probes. This early-on in the game that gives you a huge advantage, (...) Not entirely true. 4 Lings = 2 Drones anyway, not counted in the mining they would have done if they weren't lings. | ||
|
osiris17
United States165 Posts
Times your zerglings arrive at protoss base with an 11 pool vary with map. Given a bit of time to deal with the probe at your natural, it varies from ~4:00 on maps like ohana to ~4:10 on maps like daybreak. To have a wall complete @ 4:00, probe construction is delayed from 2:50 - ~4:05 for a typical build. Which is 1:15 total time delayed and 3.75 probes worth of delay. But I am also subtracting 0:10 from that, since a standard ladder opening with a scout has about 10 seconds worth of probe delay already in it to respond to a 14 pool. So 3.25 probes is the real delay. The protoss may be able to squeeze 1 more probe out on large maps where he can delay his gateway full wall by a tad, because your zergling arrival time is delayed by ~10 seconds. In these cases probe delay will be reduced from 1:15 to 0:55, and probes delayed @ ~2.75. But delaying the gateway, if opening nexus first, runs a real risk of the zerglings killing the gateway before the cannon completes. I'll have to actually try that on ladder. But I am not concerned about it that much; because either way, if they do get away with this, the zerg is still fine and breaks a tad above even. The only way the 11 pool is at a disadvantage is if the opponent does not scout and opens nexus 1st. In these scenarios there will only be 25 seconds of probe construction delay due to the extra minerals from not scouting being delegated to probe construction. So this is a compounded financial bonus from not scouting. Since almost every Protoss scouts on ladder I am not concerned about this very much. But in a tournament setting it may be relevant. 25 seconds leaves the zerg ~1 delayed probe behind the goal. But this effect will occur regardless of map size; 25 second delay is the floor of what is possible. I am going to assume he scouts since not scouting on ladder is an invitation to get 6 pooled. And if he doesn't scout your 11 pool at all, not even with a late scout, you will still surprise him and force him to pull some probes, and that will catch you up a bit. The ideal counter for the protoss is probably a late scout after nexus. Anyway I am going to overlook this for now, I may revisit it for a 12 scout / 16 scout later. Saving up chrono upon scouting the pool is not the proper reaction since it overstockpiles. So no need to consider chrono. The early gateway from the faster walloff reaction is layed down at 3:30 compared with a typical 4:00. At this point the 11 pool comes ~.25-1.25 workers ahead of a standard protoss build compared to what a 14 pool would by 10:00 in most circumstances. The rest of the analysis will be in calculating the effects this free early gateway has on the zerg, and if they are enough to make the 11 pool not financially worth it. Just thinking ahead, the earlier gateway will help stargate openings, warpgate rushes, and early gateway pressure the most. The immortal allin & blink allins come later, and they dont rely on a quick warpgate. They rely more on the forces being massed. The early gateway pressure will actually require him to cut probes or delay his core in order to take advantage of his earlier gateway; otherwise it may as well be a 4:00 gateway. Delaying the core is more intuitive... If he delays his core, you at least know there isnt a super fast warpgate timing coming. But a 2 zealot 1 stalker pressure can come 30 seconds earlier than normal. So this will require 30 seconds earlier 4 zerglings.. it's a negligible expense but it amounts to about a minute worth of mining time which is 40 minerals. But it doesn't permanently cut into the drone count since this pressure comes now or later. If he keeps up the pressure with another stalker it may get interesting, but the more he invests in this the more it indicates a warpgate timing.. Also the stalkers harassing you will give you a free ovi scout of his base. A warpgate timing with a normal core timing like this can be stopped easily so I am not worried about it. The earlier stargate opening may require a slightly earlier evo but that initself is a minor expense. The spores are reactive and will come earlier too. That altogether compounds into about 2:00 of mining time , about 80 minerals. Hydras will be 30 seconds slower. So the opponent can harass you more freely for a longer period. The advantage of this is larger if he opened 4 phoenix, not quite as important if he went 2 phoenix / 1 void, since the phoenix / void are usually driven off due to damage received by queens & not hydras. But the opponents earlier stargate means his gateways will be slower, and this means he will have trouble expanding if you churn out some lings upon scouting stargate. At this point I think its impossible to say for certain what comes out ahead. In this regard I'd do the build just to scramble the opponents timings up. The earlier potential warpgate rush is also a concern. I will have to test how much earlier an gateway timing comes. Protoss is delayed himself, gateways may be delayed. There are some 4 gateway rushes which will probably be the best example of the effect. But having the earlier queen from the 11 pool may give you more larva at that time, offsetting the problem. I will post back with more results later. At this point I am going to open 11 pool every game because of the economical advantage; and the safety from 1 pylon & 3 cannons delaying the natural. | ||
|
NOOBALOPSE
Canada802 Posts
| ||
|
osiris17
United States165 Posts
For just the finances it is possible the 12 pool is better, it doesnt seem to make sense and I didnt conduct the tests myself. The timing for pressuring toss is very critical, a 10 second delay will mean they squeeze out another probe. And that would make the whole build not worth it. So unless the 12 pool results in 200 more total resources accumulated @ 10:00 than an 11 pool (represents 1 probe worth of difference from later pressure), it isn't worth it. | ||
|
HoMM
Estonia635 Posts
On September 20 2012 08:31 Belial88 wrote: http://www.twitch.tv/quanticdecemvre/b/332051855 He does it as an aggressive opener, not as a macro opener. If he does not do damage because toss responds properly, then he would be way behind. As a macro opener, 12 pool is terrible. He does it as an aggressive opener, not as a macro opener. - no did you even watch the vods you linked? :| If he does not do damage because toss responds properly, then he would be way behind. - no As a macro opener, 12 pool is terrible. - no | ||
|
osiris17
United States165 Posts
| ||
|
ulfryc
Germany115 Posts
| ||
|
osiris17
United States165 Posts
| ||
|
Belial88
United States5217 Posts
Anything the pros don't do, Belial doesn't approve. ; D Pros don't do it. You take one game a lesser-known pro has done, once, completely out of context. Here's a link on zergs going 14-15 pool ZvP: everyrecentstarcraft2tournament.com Btw, overpool is overlord before 11 pool. and yea, kind of. There are a ton of pros out there, hundreds, if not thousands, and they know better than me and you. They do what they do because they test it out, at a very high level, with a very detailed look at what is better than what. Everyone wants new strategies, but there's a reason pros don't open 11 pool as a macro opener every game. If you want to recommend it as an aggressive opener that's fine, but it's going to be behind 14-15 pool economically. He does it as an aggressive opener, not as a macro opener. - no did you even watch the vods you linked? :| If he does not do damage because toss responds properly, then he would be way behind. - no As a macro opener, 12 pool is terrible. - no No, in every game he takes his 4 lings and runs them in to toss' base. His third is also significantly delayed. Discussing zerg openers is fine, but this topic has already been discussed and hashed out... like 2 years ago. There's a reason none of the pros do anything besides 6/7/8 pool and 14/15 pool in zvp. Go into a YABOT and test it out. TLDR: 2.5 probe delay is indeed the goal, and the current analysis shows this is achieved; infact exceeded even against a nexus first, IF the opponent full walls (which he should, and you should make sure he does). The build looks extremely promising to me now but a more thorough analysis is still needed. You are down a lot more than 2.5 drones for doing this vs 15 pool, and I wouldn't say that's an accurate way to put it, since toss is just delaying workers to get something the need later on anyways. Whereas when you do 11 pool you aren't doing that, you are hurting yourself economically for time. Toss is just spending the same money, somewhere else. | ||
|
Belial88
United States5217 Posts
A while back someone posted a thread with graphed analysis of how the various pools performed relative to eachother. 11 pool outperformed 12 pool & 10 pool. 11 pool also allows earlier pressure against the toss, forcing more probe delay & earlier wall. For just the finances it is possible the 12 pool is better, it doesnt seem to make sense and I didnt conduct the tests myself. The timing for pressuring toss is very critical, a 10 second delay will mean they squeeze out another probe. And that would make the whole build not worth it. So unless the 12 pool results in 200 more total resources accumulated @ 10:00 than an 11 pool (represents 1 probe worth of difference from later pressure), it isn't worth it. That was 12 pool double extractor trick, and overpool, not a 9 overlord 12 pool and 11 pool before overlord. You can't justify a build because the opponent may overreact either, that's ridiculous. Why not just go 6 pool, ???, win every zvp because "forcing more probe delay"? Because if Toss reacts properly, the probe delay will be a minor hit compared to the hit you are taking for doing the aggression. This is generally true with all pressures/all-ins. Why not roach/ling all-in Toss every game? It surely delays their probes, tech, economy, right? Because if toss responds correctly, Zerg is going to take a much larger cost. Similarly, yea, toss may delay a probe or two against 11 pool, but you are going to be down 4-6 workers and a delayed third compared to Toss who might be down 2 probes (and he makes a quicker gateway, so it's not all bad either). The build is not worth it unless Toss doesn't react properly (or doesnt scout). Which happens all the time, I'm sure, and I'm sure many Toss won't think anything suspicious of a pool planted before 2:00 or is done before 3:00, and even if you get lings into Toss' base, I'm not so sure it would be worth it as long as Toss plays correctly. You gotta actually kill quite a few probes, if you make 6 lings than you need to kill like at least 4 probes to make it just even, which i dont see happening. i think the way you compare stuff is kind of weird too. Compare the total money mined in builds vs eachother, and drone counts. | ||
|
velvex
Germany226 Posts
On September 21 2012 00:49 Belial88 wrote: Similarly, yea, toss may delay a probe or two against 11 pool, but you are going to be down 4-6 workers and a delayed third compared to Toss who might be down 2 probes (and he makes a quicker gateway, so it's not all bad either). Tbh, you are totally making up that 4-6 workers thing. I'll give it one more try: Since both builds make all larvae into drones besides 2-4 lings and a fixed number of overlords, drone production equals larva production. So you should be able to show the supposed 4-6 workers difference by looking at hatchery and queen timings. For 15 pool, I'll take the first game of Symbol vs. Seed of today's GSL on Cloud Kingdom, and for 12 pool the replay I provided. Hatch timings (15 pool): ~2:45, ~4:00, Hatch timings (12 pool): ~3:30, ~4:15 Queen timings (15 pool): ~3:10, ~4:25, ~5:20 Queen timings (12 pool): ~2:50, ~3:46, ~5:06 So 12 pool in these sample games loses one minute of hatchery-based larva production, which equals 4. Queens, though, are much earlier, 1:13 in total, which equals an approximate plus of 6-7 larvae. Then there is a small loss through the fact that the second queen has to travel to the natural and that there are brief larva blocks in the main after putting down the third, but that isn't more than 2-3 larvae. Once your natural kicks in, you can make all your larvae into drones. Which mechanism then causes that 15 pool makes so many more drones than 12 pool openers (or 11 pool, for that matter, there isn't a big difference between those two)? | ||
|
Defenestrator
400 Posts
Ulfryc I think the differences between the builds you mention is very small (like 2-3 s). I use 11OL/11pool into 19hatch. I tested this vs 9OL/11pool once and there was basically no difference. I'm not sure but the ideas behind all these variations is the same (lings before hatch to secure hatch easily). As far as lings in their base, this is a complete game-changer. In customs vs mid/high masters (as well as ladder matches vs low-mid masters), if I get lings in their base I've won every single time. I could go into why this is, but suffice to say if you micro properly and they themselves don't have HuK-level micro, you will end up miles ahead. Toss should do everything in their power to prevent this (and it is preventable, as has been discussed above). One way to compare 11pool vs 14pool is similar to opening forge-first vs nexus-first: forge-first counters 15hatch because you can easily ocannon the Z player with him falling further behind than you. 11pool is similar vs nexus-first: it basically counters nexus-first because when you pressure with the 4 lings, you either draw even or he falls further behind. 11pool vs 6pool is much better because it's not a coinflip opening... worst result with proper play is you come out even against the toss, as Osiris has been so kind to analyze for us. If toss opens gateway-first, 6/7/8pool flat out loses. So why do pros do 14/15pool into 16hatch then? Well first, as has been pointed out, it is better economically if zerg doesn't pressure toss. Second, either A. Pros have a response to every type of cannon aggression with this opening that puts them even/ahead and/or B. they don't care about the 3-pylon block because this can't happen in tourneys anyway and they don't give a shit about ladder. Does anyone have any input on the proper response to common cannon openers from toss vs 14pool/16hatch? For example, are lings out in time to stop a double-pylon wall behind natural minerals with a cannon after hatch starts? And is there any way to counter a 3-pylon block without patrolling a drone? I haven't seen any pros do this, so I'm assuming they know how to deal with it (unless like I said they just don't care because it's not possible in tourneys). | ||
|
Schnullerbacke13
Germany1199 Posts
I think 11 pool is viable and PROS DO PLAY 11 POOL. E.g. Stephano does it sometimes, Lowely opens 11 pool frequently. 11 pool is an excellent allround ladder opening. If your opponent is likely to play greedy, 11 pool may be a slight waste, as you cannot do any real pressure, but this is by no means game deciding. A huge factor is your 3rd and tech timing. An earlier pool gives you more larvae early. In order to not idle larvae production you are forced to spend minerals for drone production, so you will barely bank minerals for tec/hatches. If you plan to 3 base at 4'10, an 11 pool might not the right opening. However a lot of zerg players tend to go for a 36..40 supply 3rd which is fine with an 11 pool. In this case an 11 pool is slightly ahead over 14+ pools in drone production. | ||
|
NOOBALOPSE
Canada802 Posts
On September 21 2012 00:40 Belial88 wrote: Pros don't do it. You take one game a lesser-known pro has done, once, completely out of context. Here's a link on zergs going 14-15 pool ZvP: everyrecentstarcraft2tournament.com Btw, overpool is overlord before 11 pool. and yea, kind of. There are a ton of pros out there, hundreds, if not thousands, and they know better than me and you. They do what they do because they test it out, at a very high level, with a very detailed look at what is better than what. Everyone wants new strategies, but there's a reason pros don't open 11 pool as a macro opener every game. If you want to recommend it as an aggressive opener that's fine, but it's going to be behind 14-15 pool economically. No, in every game he takes his 4 lings and runs them in to toss' base. His third is also significantly delayed. Discussing zerg openers is fine, but this topic has already been discussed and hashed out... like 2 years ago. There's a reason none of the pros do anything besides 6/7/8 pool and 14/15 pool in zvp. Go into a YABOT and test it out. You are down a lot more than 2.5 drones for doing this vs 15 pool, and I wouldn't say that's an accurate way to put it, since toss is just delaying workers to get something the need later on anyways. Whereas when you do 11 pool you aren't doing that, you are hurting yourself economically for time. Toss is just spending the same money, somewhere else. There are plenty of things that pros don't do that are good, and there are plenty of things that pros do that are BAD. For example: IdrA always puts 20 drones on any expansion before rallying to different minerals. It has been proved that 16 drones and then rallying your hatch is the way to get the most minerals. Another; Pros make the gasses at 6:30 in their main/nat which means that they have to get at least 22 drones on their main before rallying. This means that the rally is late and therefore they are not mining efficiently in the time that they have 22 drones on their main and are waiting for the gasses to finish. Making 2 at the third and one at the natural sounds most optimal to me because units sometimes target buildings instead of units which makes your army last longer and an easier defense. See, pros do things wrong sometimes. | ||
|
Schnullerbacke13
Germany1199 Posts
On September 21 2012 02:17 velvex wrote: Tbh, you are totally making up that 4-6 workers thing. I'll give it one more try: Since both builds make all larvae into drones besides 2-4 lings and a fixed number of overlords, drone production equals larva production. So you should be able to show the supposed 4-6 workers difference by looking at hatchery and queen timings. For 15 pool, I'll take the first game of Symbol vs. Seed of today's GSL on Cloud Kingdom, and for 12 pool the replay I provided. Hatch timings (15 pool): ~2:45, ~4:00, Hatch timings (12 pool): ~3:30, ~4:15 Queen timings (15 pool): ~3:10, ~4:25, ~5:20 Queen timings (12 pool): ~2:50, ~3:46, ~5:06 So 12 pool in these sample games loses one minute of hatchery-based larva production, which equals 4. Queens, though, are much earlier, 1:13 in total, which equals an approximate plus of 6-7 larvae. Then there is a small loss through the fact that the second queen has to travel to the natural and that there are brief larva blocks in the main after putting down the third, but that isn't more than 2-3 larvae. Once your natural kicks in, you can make all your larvae into drones. Which mechanism then causes that 15 pool makes so many more drones than 12 pool openers (or 11 pool, for that matter, there isn't a big difference between those two)? Anybody able to execute a build half way decent can veryfy, that 11/12 pool gets out more drones early on than a 14/15 pool because delaying queen lets you lose one larvae each 10 seconds, delaying hatch costs 1 larvae per 15 seconds. Also i'd like to mention, that expansion timing is frequently dictated by the timing of your first lings, as pylon block is frequent, so the early hatch timing of 14/15 pool is theoretical  | ||
|
usethis2
2164 Posts
On September 20 2012 16:16 Vermiiifuuge wrote: Doesnt really make sense to look at it this way. Everything is always relative to what your opponent does, and the timing of your third would also be an important aspect to take into consideration. Well I'm just theorycrafting. yes, it won't happen in real world. But there isn't much harm in theoricrafting for the sake of curiosity. So, if I may reiterate the question, Assuming there will be no attack or army production whatsoever (except getting ling speed), and assuming there is no 3rd base involved (you're just working off the main and the natural), which one gives you better economy at ~ 10 minute mark? Pool-first or Hatch-first? | ||
| ||
