|
Just wanted to add my two-cents after trying to use this build. I don't like it at all.
-Unsafe therefore, -Economically Unsound.
In a situation where you couldn't 14Hatch/xPooled anyway, you can't do 11Pool/18hatch and get the same economy you would have had 14Hatch/xPool.
Sure, the idea is "oh well you /can/ get defenses early if you need to cuz the pool is so early." This very often turns out to be the case. Thing is, the moment you purchase those defenses, you would have been economically better off going 14Pool in the first place.
In addition, there is a timing where, around when your hatch is finishing, you're more at least as vulnerable, if not /more vulnerable/ than you would be if you had gone 14 Hatch. A big reason, if not the main reason for this, is that your creep spread will be slower/later. This is absolutely key if you need to get Spine Crawlers into the correct place at the correct time, which is the current fashion when doing 14Hatch against P & T, and is also very good when facing Z if you do a double-queen ramp block.
So, despite the spirit of this build, I feel that for the above reasons, it is actually, -Less Safe -& Economically Inferior
|
I also like 14 hatch because of that fast creep but: what do you do if they block that? what do you do if they block your ramp? what do you do if they bunker rush your hatch? what about 6pool ,7pool? I know, Pros fight all that with drones... but i cannot so in any of the above cases i find myself better with 11Pool actually i do a 12Pool but that does not matter.
I mostly play with my friends so they know my stile and they find flows in any build if you just do that over and over. I lately like to use both 15hatch and 12pool... it seems i am 1 food late on both builds :-))
|
On December 09 2010 17:31 Shamaya wrote:Sure, the idea is "oh well you /can/ get defenses early if you need to cuz the pool is so early." This very often turns out to be the case. Thing is, the moment you purchase those defenses, you would have been economically better off going 14Pool in the first place. Eh? How would 14 Pool be better?
|
It's worthy of note that if you need to get spines quickly at your natural, you can build them in your main and walk them down to burrow as soon as the hatchery finishes. This should cut down the window when a 4 Rax or whatever can push you without resorting to mass Zerglings.
|
On December 09 2010 18:58 Hurkyl wrote:Show nested quote +On December 09 2010 17:31 Shamaya wrote:Sure, the idea is "oh well you /can/ get defenses early if you need to cuz the pool is so early." This very often turns out to be the case. Thing is, the moment you purchase those defenses, you would have been economically better off going 14Pool in the first place. Eh? How would 14 Pool be better?
i doubt hes even tried it if he believes something like that.. its worked wonders for my game
|
I can see that this is a good build but I don't see that many situations where I would use it
vs P: if i wanted to go economic i would go 14/15/16hatch then pool, which sets up a nice defense for 4 gate and is better than your build for the economy. If i somehow needed that pool early to defend against a 2 gate, then why would i use 11pool 18 hatch - since I won't be able to put my hatch down at that point anyway, its better to go 14pool and go roach etc
vs T: against 2 rax imo you need creep at natural to place a spinecrawler. expanding at 18 is not fast enough... so again, i would go 14/15/16 hatch first
vs Z: in zvz I think it is a good opener, since you can react very easily - if he does early pool you can defend well and if he hatches first you'll be able to 18 hatch.
at the moment i can't see why i would use this vs P or T
|
Ok I did some extensive testing of the build as I was sure the build did have some flaws which many of the followers were overseeing.
I tested the 11 pool build as described by the OP against a 14 pool 17 hatch build. I chose the 14 pool 17 hatch build as 14 pool builds are the most common for pool first builds (with gas or without gas) and 14 pool 17 hatch is a very good economic version.
Methodology: + Show Spoiler +I just played the games on normal settings (which is fastest speed) on steppes of war. I made sure i had the bottom spawn in each case so that the positioning of the minerals compared to the hatchery (which I expect to have a decent influence) didn't matter. For the rest I basically played each build up till about the 4:00 min mark and then noted some key aspects of the build: - start time of the 2nd overlord - start time of the spawning pool - eventual drone delay at a hatchery - start time of the 2nd hatchery - total minerals mined at 2:00, 2:30, 3:00 and 3:30. The reason I chose these was because these characteristics are the only real differences between the build orders. The amount of drones wasn't measured as it easily follows from both the pool and hatchery starting times and the starting time of the spawning pool (which determines the starting time of the first and 2nd queen). The mineral times were chosen as I think those are the most crucial times were minerals differ between the build orders, beyond 3:30 you are always reacting to your opponent anyway making lings or getting gas or whatever. Beyond 3:30 measurement gets alot trickier as well as aspects like how well was the build executed, how many drones to maynard, what map was played etc. play a MUCH bigger role then. For measurement I used the match history screen to determine the start time of the overlord, pool and hatchery. For the minerals mined I used the 'Spending' tab and counted all the minerals spend + the current minerals. It is important to note by the way that this method is not completely flawless as the 'spending' tab doesn't always show everything, for example if something is queed or hasn't yet started building the money is actually already spent (ie. not showing as a player resource) but does NOT yet show in the 'Spending' tab. Example's of these can be queeing up the 2nd queen, or making a hatchery while the drone is still on his way. I carefully watched and checked my data for these kind of errors though. Finally the eventual drone delay i simply timed by watching the replay carefully. I performed both the 11 pool and 14 pool twice to see if there was a big discrepancy in measurement of the minerals. Between the 2 the measurements differed by 15 minerals max at any time, which makes me assess the deviation to be about +- 10 minerals.
The build orders I used are the following: 11 pool build as described by the OP: + Show Spoiler + 10 extractor trick 11 overlord 11 pool 16 queen 18 hatchery 18 overlord
14 pool 17 hatch as follows: + Show Spoiler + 9 overlord 14 pool 17 hatchery 16 queen 18 overlord
Data Replays used: + Show Spoiler +11 pool 14 pool
+ Show Spoiler + 11 pool: 2nd overlord started @ 1:08 pool started @ 1:33 hatch started @ 3:05 idle hatchery time: ~7 secs 14 pool 2nd overlord started @ 0:48 pool started @ 1:53 hatch started @ 2:50 no idle hatchery
minerals time 11 p 14 p effective difference 2:00 1464 1470 10 2:30 1704 1745 45 3:00 2019 2055 40 3:30 2334 2394 60
Interpretation + Show Spoiler +As can be seen 11 pool DOES indeed cost some minerals in the early game, roughly equal to 50 minerals. The spawning pool does start a fair bit sooner though (20 seconds) which results in both the first (and potentially the 2nd queen) being 20 seconds faster then a 14 pool build. The 2nd hatchery however is delayed by 15 seconds compared to a 14 pool 17 hatchery build which arguably already has a late hatchery. As for the larvae difference it basically comes down to idle hatchery time & late 2nd hatch VS. fast queen(s): 11 pool has the first queen 20 seconds faster. Given that a queen adds 1 larva every 10 seconds this is basically +2 larvae compared to a 14 pool. However there is also a downtime on the hatchery of roughly 7 seconds, given that a hatch produces 1 larva every 15 seconds this equals roughly 1/2 a larvae. Also the 2nd hatchery is 15 seconds later which equals another larva. As a result in total 11 pool has 2 - 0.5 - 1 = 0.5 larva more then a 14 pool build. Note that if the 2nd queen is made DIRECTLY after the the first queen that queen will also be 20 seconds earlier. In that case the larva advantage compared to 14 pool will be 2.5 larva. In practice directly queeing the queens after eachother is VERY hard with a 11 pool build though (as it requires no scouting, no lings, no gas and a very smooth performance) whereas it's easy to quee the queens with a 14 pool build. Overall I'd guess the larvae difference between the 2 builds is somewhere around 1 - 1.5
Discussion + Show Spoiler + The first thing to note is that my data deviates a bit from what the OP and some others have been saying in this thread. I can only answer those by saying that 1) the OP didn't test a proper 14 pool build and 2) the graph provided by the OP is very misleading, as mineral counts beyond the 4 minute mark are absolutely irrelevant. Of course a build that sacrifices early economy to get more drones will get theoretically infinitely ahead in minerals later on, this however is completely irrelevant when it comes to real play.
Something that isn't seen in the data but I did notice during testing was that the 11 pool build is extremely tight. ie it fits exactly which is not neccesarily a bad thing but CAN be when you do small deviations. For example if you work in a drone scout around 12 or so I foresee another bit of idle hatchery time at 18 pop as you're waiting for the 3rd overlord to finish and your 2nd queen will be delayed even more while the faster queens is a big reason to do this build at all. Also the 20 second delay on the 2nd overlord can be annoying at maps where you actually want to use that overlord to scout out a spawning spot / reach a zone to camp a cliff (LT and meta for example).
As a result I think the 11 pool build is simply bad on maps where you actually want to scout early (4 player maps) as then it's basically just worse in every aspect then a 14 pool build. Note that the larva advantage is HUGELY dependant on how fast you can actually make the 2nd queen which in turn does depend alot on having to make a scout, having to drone patrol the ramp etc. On 2 player maps however you could say that scouting with a 11 pool build is unneccesary and the delayed 2nd overlord is irrelevant as you are sure to scout their position with your first anyway. There is a fair point therefore to be made about this build on (small) 2 player maps. By making a small deviation to the build you can add in a early pair (or multiple pairs) of lings while still getting a super fast queen. Those early lings will then: - make sure you can get your 2nd hatch down quickly (whereas that's the big liability with a 14 pool 17 hatch build) - scout their base quickly - perhaps force them into a inefficient build (for example the Protoss dual stalker opening generally skips the first zealot, but that would be very risky if you arrive at their door early with 4 lings).
Conclusion Overall I think this build has potential in (only) these 2 scenario's: 1. (small) 2 player maps such as steppes of war, blistering sands , xel naga caverns and possibly scrap station. You don't really need to scout with a variation of this build on those maps as early lings combined with your overlords can do that instead. Early hatch builds are also the most risky on these maps as they are garanteed to block your hatchery IF they want to as it are 2 player maps. Faking aggresion on the small maps also has the most potential. 2. It's very likely your opponent will do a greedy expansion build, for example a Protoss forge first expansion. By doing a quick pool timing you can then completely throw off the timings of their build while still being relatively fine if you want to just macro up. LeenockFou vs Guineapig at Lost temple in GSL 3 is a fine example of this.
I would recommend doing a slightly adjusted build though: 10 extractor trick 11 overlord 11 pool stop drones at 15, then when pool finishes: 15 queen + 2 pair of lings using a extractor trick 19 overlord 19 hatchery (the lings should be at your natural exactly when you move the drone out, able to kill their scout and prevent a block).
|
I think the individual mineral/larvae testing with ideal build orders etc. is extremely irrelevant to making you a better gamer (which I assume all this work is going into), especially as Zerg it will give you a miniscule advantage in maybe 1% of games where you are able to do the build order perfectly.
I was at first extremely anxious to finally test out a pool first build that is economically even to hatch first builds. After trying it for a couple of games however (and some noobs on ladder even overreacting to it) I have ultimatively decided to going back to hatch first as my standard economy build. I don't have the numbers to support my statement, however, from a relatively neutral standpoint (unlike the people that have already invested a lot of time into defending/attacking this build order), I can tell you that it feels worse compared to a hatch first. Not only does it get blocked much more often than 14 hatch does, it also has much later creep spread (which is one of the most important things when holding off early rushes) and is more forced to use its queen energy on injects to stay on an even level of production compared to a 14 hatch. (drone production is irrelevant when you want to get things like scouting, gas geysers, gas miners, zergling speed, roach warrens and early defense)
ZvT there's absolutely 0 reason to do it. It loses to everything while beating nothing. (please don't give me the it does well vs 2rax crap) ZvZ it might be good, I haven't given this MU much attention though. ZvP it does worse against a 4gate than hatch first does and it tends to get blocked in my experience.
These and many of the ones listed above are, I think, the reasons we never see hatches beyond 17 supply at the professional level. While the pros aren't the number-crunchers and theory-crafters like some posters in this thread are, they certainly have a fuckton of experience with all different kinds of builds and naturally steer towards those that actually do well in a real game of Starcraft 2 (which you cannot emulate using pretty graphs and vsAI replays). Noone in this thread is at a level where he wouldn't benefit from copying the pros and actually focusing on things like decision-making, larva management, scouting and overall strategy more than from doing these calculations. All in all, the time used on comparing 20 mineral differences between build orders is largely wasted, you have probably already noticed that all of the successful Starcraft 2 pros sort of wing their builds after approximate ideas and past experiences and gain the advantage elsewhere.
|
Chickenlips is completely right on this but i was just giving my personal input just for the theory sake and fun of the argument. Also there are tons of circumstances where pro's (who indeed just tend to wing it) do horrible build orders even up to GSL level. With the advent of pro teams in korea and coaches dedicated to strategy fine tuned builds are seeing more play though as the level of play rises as well. Overall I agree though that 14 hatch is basically the best build in 80% of the cases as early creep spread and not getting blocked are HUGE for Z, but in some cases other builds can be alright. This particular build is only good for dealing with hatch blocks or (faking) early pressure both of which are often not really neccesary considering it does take a decent dent in economy to do so (50 minerals down is considerable).
|
He gets infestors, which is a very popular ZvZ strat these days, but I think they are honestly terrible. I haven't lost a game yet to any infestor play. I think the gas is much better spent in hydras or upgrades, but that is just my 2cents.
The way this guy from your replay used them - sure they r useless. Actually there is not many ppl which use them properly. Thing which he wrote in replay just shows he got no clue how to use them - u dont need critical mass of them - all u need is enough of them to cast fungal 5-6 times total. Thing is to use it properly: way he did it is stick infestors to roaches and wait in base for u - thats wrong and if u use it this way u can skip them coz they are not cost efective - even if u got 10. Way they are good: get 2 of them and met enemy army in middle of the map - fungal 2-3 times whole his army there and pull back to base - before he get into ur spine crawlers do another 2-3 fungals and pull back to spine crawlers with rest of roaches - u dont need any more fungals at this point - if he decide to rush in his army is ALREADY at half hp and 2 more fungals will wipe everything together with few roaches - if he had mostly hydras then probably army is already dead. This is proper way of infestors but if ppl just camp them and use them when enemy is already in their base then its just waste of gas. Thats also the reason why top pro zergs are using infestors - they know how to use them and if they wont - they will lose.
About build - its rly good - i especially like it vs zerg and terran. Vs zerg its just rly safe to any super early rush and super eco build. Against teran they mostly prepare for roach rush after super fast poll and/or bbust so they put shit loads of bunkers waste minerals on heavy wall and its pretty good vs 2rax push. Maybe what idra wrote is true that its easier to fight 2 rax with 14 hatch but well - im not idra and i got no such micro to just fend of stuff with drones, but my enemies are no foxer as well so its better then hatch first for me. Against protos i got mixed feelings - i see no big benefit from it compared to standard against protos which is 14 pool. Most ppl wont see difference between 11 and 14 pool on big maps and just play normal - 14 pool can already counter 2 gate pressure so i see no real benefit here - its not bad but nothing special. Anyway nice job and great opening.
|
On December 09 2010 18:58 Hurkyl wrote: Eh? How would 14 Pool be better?
This is what I've been saying all along. a 14pool is better if you get lings, because the 11pool build is _worse_ economically in the beginning, obviously, but catches up when the earlier larvae become drones. if you make lings instead you have had less drones mining and the new drones that you'll make when you can will have mined for a shorter time. the 11pool is good in the sense that its flexible and allows you to easily beat cheese that would give you a hard time if you went hatch first, while being a good economic build if it turns out villain isnt cheesing. However, I've had a lot of difficulty against fast hitting 4gates with it because of the worse creep spread; AND its not good at all against 2rax because it will leave you terribily behind economically and it's very easy for terran to set up a contain if he wishes to do so.
|
really good posts by shamanya, chickenlips and demetrio further explaining the problem this build has with early pressure. as strange as it may sound to the supporters of this build but the weakness of an 11 pool is actually early pressure because you will come out behind economically compared to the 14 hatch or 14 pool although it may be easier to fend it off (this is IMHO the reason many people like this build).
|
On December 09 2010 23:55 Markwerf wrote: ... considering it does take a decent dent in economy to do so (50 minerals down is considerable).
yes 50 minerals is huge :-)))) SURE! not to mention that when you do the extractor trick you end up with 4 minerals that you can never spend!
i also do the 15 hatch first because i like the statement, it feels very good that you expend with first money then build something else, but this build is ok and it has a lot of potential.
|
On December 10 2010 00:29 dementrio wrote: because the 11pool build is _worse_ economically in the beginning, obviously, but catches up when the earlier larvae become drones.
On December 10 2010 01:05 fleeze wrote:
because you will come out behind economically compared to the 14 hatch or 14 pool
Accepting data which conforms to your convictions and rejecting data which opposes your convictions is a very good life skill to have. The next step is to become involved in partisan politics, listen to pundits who reinforce your beliefs, and form a love/hate relationship with political candidates.
|
On December 09 2010 22:33 ChickenLips wrote: I think the individual mineral/larvae testing with ideal build orders etc. is extremely irrelevant to making you a better gamer (which I assume all this work is going into), especially as Zerg it will give you a miniscule advantage in maybe 1% of games where you are able to do the build order perfectly.
I was at first extremely anxious to finally test out a pool first build that is economically even to hatch first builds. After trying it for a couple of games however (and some noobs on ladder even overreacting to it) I have ultimatively decided to going back to hatch first as my standard economy build. I don't have the numbers to support my statement, however, from a relatively neutral standpoint (unlike the people that have already invested a lot of time into defending/attacking this build order), I can tell you that it feels worse compared to a hatch first. Not only does it get blocked much more often than 14 hatch does, it also has much later creep spread (which is one of the most important things when holding off early rushes) and is more forced to use its queen energy on injects to stay on an even level of production compared to a 14 hatch. (drone production is irrelevant when you want to get things like scouting, gas geysers, gas miners, zergling speed, roach warrens and early defense)
ZvT there's absolutely 0 reason to do it. It loses to everything while beating nothing. (please don't give me the it does well vs 2rax crap) ZvZ it might be good, I haven't given this MU much attention though. ZvP it does worse against a 4gate than hatch first does and it tends to get blocked in my experience.
These and many of the ones listed above are, I think, the reasons we never see hatches beyond 17 supply at the professional level. While the pros aren't the number-crunchers and theory-crafters like some posters in this thread are, they certainly have a fuckton of experience with all different kinds of builds and naturally steer towards those that actually do well in a real game of Starcraft 2 (which you cannot emulate using pretty graphs and vsAI replays). Noone in this thread is at a level where he wouldn't benefit from copying the pros and actually focusing on things like decision-making, larva management, scouting and overall strategy more than from doing these calculations. All in all, the time used on comparing 20 mineral differences between build orders is largely wasted, you have probably already noticed that all of the successful Starcraft 2 pros sort of wing their builds after approximate ideas and past experiences and gain the advantage elsewhere.
1) Your feelings are no basis for argument here. I feel like this build is better than hatch first. So what? 2) The purpose of the testing and data is not to make you a better gamer. It is to shut up the people who keep repeating like drones that this build is "behind economically." 3) Suggesting that a hatch block is more detrimental to a pool first build than a hatch first build is absurd. Getting your hatch blocked in hatch first is a HUGE hit. And I doubt that the few earlier seconds makes a difference in the frequency of hatch-blocking. 4) Saying it does well against 2rax isn't crap. It is documented fact and common sense. "But ret and Idra said....!" Yes, I know. Unlike some on TL I can think for myself and reach conclusions with common sense and data. I'm sorry, but it is mind-numbingly stupid to suggest that hatch-first is better against 2rax. Every ounce of data rejects this conclusion. I have replays of 11 Pool crushing 2rax. And we have countless games of professionals losing to 2rax by going hatch first. The best zerg in the world got kicked from GSL because he stubbornly insisted on going hatch first every game against a 2raxing opponent. Don't forget that the 2rax build was started as a COUNTER to hatch first play. That is the only reason it has been effective. So you can stubbornly follow the stubborn pros into defeat, or you can apply some common sense and think for yourself. 5) You say the build only gives an advantage 1% of the time when you can do the build perfectly? Um, I posted 12 games of this build beating diamond players, and not once was it executed "perfectly." There is no such thing as perfect anyways. 6) This build is all data and graphs against AI opponents, and doesn't do well in an actual game of starcraft? Again, look at point 5. Did you even read through the OP, or did you just look at the pictures? 7) Despite stating that you are from a "neutral standpoint," you end up sounding almost irrationally opposed to the build in every respect.
EDIT: I added another spoiler to the OP so that the "only look at the pictures" crowd won't see ANY replays at all. Hope that will prevent this issue in the future.
|
I didn't get a chance to read this entire thread, but from the perspective of a protoss player who goes 15 nex every single game except on Steppes/DQ, I met this build on ladder a few times and it's annoying because I have to put down my forge first. This build (11 pool) definitely comes out ahead of me versus 14 pool (where I can always nex first) or hatch first.
|
I've been doing 12ov pool, 15ov, 16queen + 1 pair of lings +gas, and hatch at 18-20. Your first 2 lings come out in time to drive the worker away and it won't get you supply blocked if the hatch is delayed. I'm not sure about how the economy goes in comparison to 11 ov pool, but the pool is only about 7 seconds later and you have an extra drone and a lot easier time with other stuff, so I don't think it should be too bad.
EDIT: Btw 12opool doesn't have even a second of larvae downtime where you sit around with 3 larvae while the pool is building, I think that the fact makes it pretty awesome.
|
On December 10 2010 03:17 jdseemoreglass wrote: 4) Saying it does well against 2rax isn't crap. It is documented fact and common sense. "But ret and Idra said....!" Yes, I know. Unlike some on TL I can think for myself and reach conclusions with common sense and data.
You seem like an articulate fellow but isn't it a little presumptuous to claim that two of the best Zergs in the world are flat out wrong, after weeks of testing against top competition? As opposed to practicing on...North American ladder and Build Order Gizmo #2394? Idra's practice habits appear to be unbelievably thorough -- he takes his job very seriously obviously. Ret is a beast as we all know. After dedicating all those hours to their craft, I have a hard time finding anything in this thread that overturns their opinion. As a decent Z (2200+) I can tell you I enjoy considerably more success with 15 hatch than this 11 pool build. Shrug.
Cheers.
|
Just wanted to pitch in and say that i find this build very interesting. I have used it for 2 days now, and i have had lots of success with it. Granted, I don't know if I would've had any less of success doing the standard build, but I am going to keep using this build for some time now, and see if it keeps giving me good results
|
I'd personally prefer to 13 pool, 16 hatch against protoss all the time, because that build is more robust under small deviations than the 11 pool 18 hatch.
However, there's always a pylon at my natural when I 13 pool. This delays the 16 hatch. There's also the danger of being walled in by pylons and a cannon.
So the real comparison should be between 11 pool, 19 hatch (extra pair of lings to kill pylon) and 13 pool, X hatch, when the hatch is so delayed by a pylon.
I suspect that the 11 pool is a lot better than a 13 pool, seriously delayed hatch, when dealing with this sort of problem. It also lets you handle an early nexus very easily.
|
|
|
|