|
Warning: This is fairly lengthy, 4000 words
Introduction
This is an attempt to discuss a timing window in PvZ vs a 3 hatch spire into 5 hatch hydra build– from 7:00-8:00 depending on the opening builds (7 for overpool, later for 9 pool and earlier for 12 hatch), done using builds extrapolated from a few recent Violet pvz replays. Also, all times were derived from chaoslauncher’s plug-in apm-alert while watching replays (so they’re in in-game time). I’m calling this thread 7:00 timing push [edit : It should actually be called 7:00 timing attack, according to comments] for convenience, because that’s the approximate time a P should push out vs an Overpool opening, which is the most common opening on the applicable maps. I'm unsure whether to list this as a [G] or [I] - It definitely contains parts that are a guide, but I don't feel qualified to call it such. At it's core, it will be an examination of 2 different builds of Violet's that utilize this window.
I’m going to begin with a caveat, that I’m terrible at pvz and don’t have even close to the mechanics required to pull off the majority of what I will be discussing – just executing the first 7 minutes of Violet’s simplest build is a daunting task for me, to the extent that I had to spend around 3 hours of replay watching and and a few more practicing in single player to come close to the timings, but this has improved my skill in pvz greatly and showed me my weak points (it’s all macro). I'm writing this for a few reasons - I'm an english major, and have to practice writing papers again after not writing any for months (technical writing is my weak point too), because I feel that it's a topic deserving of intense scrutiny by far better players than I am, and lastly I feel that this would help with most lower ranked players' early pvz.
Despite my general terribleness at SC , I've watched every proleague match of r4 and r5, been following the scene pretty intensely since around april (right after Fantasy went 2-3 against JD), building on my general game knowledge. I’m slightly worried that this might just be common knowledge to people with a higher skill level than me, but they may not have it formulated quite this clearly – the first half of this document will be a discussion of PvZ metagames and the timing window, the reason’s for its existence, and how to take advantage of it. The latter half will be an examination of Violet’s build orders.
If I am wrong in anything, please correct me.
PvZ Metagame
The current metagame of protoss vs zerg, a matchup that was recently frequently commented to have been figured out by protosses, has swung back in favor of zerg, a fact established merely by examining the pvz ELO rankings – only 1 toss is above 2100, and it’s Bisu – and he’s below 2200, which he had been above for the entire time I’ve followed the SC scene. This is partially explained by maps, partially to an increase in the skill of Z players, but mostly it is due to the optimization of the neo sauron build, 3 hatch spire into 5 hatch hydra. The birth of this build had the 4 gate 2 archon timing push rapidly developed to counter it, however zerg’s figured out this timing and this build is almost useless except to defend now – the stargate, and getting 3 gates delays the attack by enough that the zerg players just switched to earlier hydra production and took the economic hit.
So many protoss were knocked out of the leagues early that almost no pvz has been played in the last 3 weeks - something like 14 games outside of heritage, and very few of those were good showcases of Protoss skill (Stork v Modesty is the obvious exception). This timing attack takes advantage of current 2 or 3 player maps, and the way they all encourage early aggression via mineral patches or temples. For example, on Medusa, the Z will attack the temples with his lings and force the Z to waste 250 minerals on a cannon. Heartbreak ridge’s mineral patch, outsider’s mineral hopping (same for holy world), etc. This reverses that by taking advantage of the slight delay in hydra production caused by making this many lings, and the messed up timing a very aggressive zerg will tend to have while they are busy microing their lings to kill that last probe.
Iccup and me
The growing prevalence of the 3 hatch spire into 5 hatch hydra build on iccup has had me personally desperate – I’m terrible at pvz, losing when the opponent has no idea of what a build is, let alone when they have a perfect one. Watching proleague was a dismay - toss were dying left and right, and no one besides Bisu seemed to have a handle on how to play vs Z anymore. Then I watched some of KT's games on a whim, a team I had hated because Flash bored me (sorry Flash fans). Violet's pvz style was very different from Bisu's, risky, and dark archon heavy. His pvz streak eventually died to ling all-ins and great (Who is one of the best zvp around), but I was still impressed.
So I looked around for some recent pvz pro replays, and was delighted to find that Violet had 2 floating around. (One vs Lzgamer, one vs Hyvaa [apparently, I didn’t check this with romad, that's what it said on yaoyaun i think]). The only other recent pvz I found from a good pvz’er is Bisu’s recent one, which is of no use to me- it’s an unorthodox game and I would never have his mechanics, nor would anyone else reading this, realistically. It’s also not recent enough – neo sauron wasn’t as commonly perfected at the time, even though it's only a month or two old. I will never have Violet’s mechanics either, however his build’s were built on timing and mind games, things that don’t rely on mechanics - they require mechanics to perform, yes, but they were builds that placed the vast burden of his play on thought rather than the sheer multitasking of Bisu. (He double fakes sair reaver vs hyvaa, it’s fascinating and a really good game, perfect example of s-class pvz- Hyvaa is never in control of the game). This is also a good read on why Violet died so many times to ling all-ins – his build encourages early lings, because it makes it stronger later.
The Window
Both of Violet’s games, his build revolved around a 7 minute timing push – he pushes out with a small force at 6:44 to kill the lings surrounding his base in order to deny scouting, while simultaneously harassing with corsairs. He then retreats, only to move out after the next two zealots are done – the pulling back is to trick the z into thinking that he is just clearing lings, a fairly standard action by most p’s, and one that forces the Z to remove his scouting units. The actual push comes at ~7 minutes both games – differing distances result in different arrival times.
It is not necessary to win with this attack – it’s merely an attempt to secure an advantage by forcing them to construct something other than drones (and hopefully hydras). However, it is necessary to deal some damage, due to differences from the standard 4 gate build. The specific timing the P should be aiming for is for the attack to hit right after an OL has been killed and scourge are being produced, right as their hydra den finishes – you’re trying to force them to make lings instead of hydras. This timing window can be incorporated into pretty much any build order, however, I am personally just copying Violet’s as I’m not knowledgeable enough to create my own. The ideal outcome of this attack would be to deny the Z’s third while taking your own, however this is difficult and not necessary for utilization of this window to be successful. What must occur is that enough damage has to be dealt to survive a delay of gateways and tech.
Timing Differences from 4 gate 2 archon :
Rather than making 4 gateways, Violet only makes 2 (including his initial gate), and delays his templar archives in favor of leg speed and +1 attack, only throwing down archives after attacking, and additional production buildings after the attack has progressed some.
Vs Lz his push is very strong, 5 speedlots and 1 dragoon push into LZ with 2 more on the way, and +1 weapons a few seconds from finishing at 7:39. Violet’s micro is good enough to totally annihilate Lz, despite lz’s having 2 sunks and pulling drones. He cuts a total of 3 rounds of probe production from both nexus’, for a total of 6 fewer probes – both for supply and the additional minerals in order to get this out so early. So essentially all he had to do was deal the economic damage of losing 6 probes to lz to pay for this attack.
Against hyvaa, he only sent 4 unupgraded lots (1 additional rallied) – but they get to hyvaa’s third at 7:15, and combined with very effective sair harass forces hyvaa to make 14 lings to preserve his third. An additional 3 OL were required due to the delayed hydras – 10 larva wasted, with more likely lost due to supply block.
Requirements
This timing relies upon the protoss’ ability to harass and delay the zerg– a totally undelayed zerg will generally have defenses up in time. Both probe and sair harassment are necessary – the second to force scourge instead of hydra, and the first to delay the Z from starting hydra production at the optimum time. In addition, it punishes all early aggression – making extra lings delays hydras by that much, and the timing attack becomes progressively stronger the later hydras come. It seems to be strongest on two player maps, and is terrible on python (as with any FE opening). It can be more powerful if the early aggression actually does some damage, as this further encourages the Z to delay his build.
Weaknesses:
However, since you are focused on harassment with the sair, you must have excellent probe micro to scout potential all-ins and verify that they are indeed going 3 hatch spire – 5 hatch hydra (this is where I’m stuck currently, I continually die to ling all-ins/hydra busts). It is very poor against 2 or 3 hatch muta – but probes should be able to scout this.
Differences in Transition :
This transitions most easily into dt/sair – you only have 2 gateways available to you when your temp archives finishes, and storm is later. If they change their gameplan to mass air, a dark archons is advisable – one archon and 2 sairs won’t do anything to a well microed muta switch at this point. This transition is easiest because the Z should be, in theory, be playing more ling heavy and have issues defending his overlords. A dark archon is not a bad idea if the dt harass fails completely, in general – maelstrom is a much faster research than storm and has a comparable general viability.
Sair/reaver is also viable, but if this is done I would encourage doing Violet’s route of merely harassing with 1 reaver and using one to defend the first hydra attacks while converting to a zeal/ht army and ceasing reaver production - based upon my experience as to what is most viable at my level of play - at higher levels there are likely other transitions.
Specific Build Orders:
Here are Violet’s builds. (very long, I tried to be as detailed as possible, they’re in second person). Yes, they’re probably overly detailed – but it’s hard to know which steps are crucial and which are not. I've provided a small amount of commentary on my thoughts as to why he does certain things like miss pylons. They aren’t perfect – these are from games that had lots of initial aggression with several probes lost.
[spoiler]
vs hyvaa – 7 minute timing attack into fake sair/reaver
8 pylon
13 forge (overpool on destination)
13 nex
14 probe
14 cannon x 2
resume probes
16 gateway
17 assim
19 transfer probes
20 core
23 pylon
24 zealot (assumes scout died or will have to cut 1 probe)
28 stargate (4:27)
30-32 - non stop zeals from 1st gate
33-34 - natural assimilator
32 sair (he missed pylon and canceled probes, was being harassed by lings: 5:19)
32ish - pylon x2
34 robo (5:35ish)
The initial sair harass:
Continuous sairs - harass overlords with first 2, save at least one sair. The goal is to kill 2 overlords and supply block. Violet kills 3 by almost killing one, then changing targets, then killing the earlier wounded one. The fact that hyvaa had his OL somewhat distant from his hatches helped(they were on his nat + 3rd cliff). Violet lures the scourge away from the wounded OL with one sair while the other sair goes in for the kill and dies. This goes on from approx 6:20 - 6:50.
42 pylon
50 pylon
50 support bay (6:30 - from here on I'll use in game times instead of supply exclusively.)
6 : 45 : move out and pressure with lots, forcing lings back. The lots will attack in another 10 seconds, but it should seem like a feint intended to drive off lings at first. (Also checks ling numbers , can usually get a few due to the sair harass going on at the same time).
6 : 54 : Shuttle speed (yes, before shuttle)
6 : 56 : reaver ( I assume in case of surprise hydra bust or ling-all-in)
7 : 00 ( 60 supply) - get 4 lots and attack z's 3rd , 5th lot should be on its way with 1 more rallied. After this, continue making lots but keep them in natural. Attack either the natural or main, whichever is closer (Desti vs hyvaa the 3rd was, vs lz nat was).
7 : 40 : Shuttle
8 : 17 :+1 weapons
8 : 20 : Should have 5 sairs and 1 reaver. Load 2 lots and 1 reaver into shuttle
8 : 25 : Citadel
8 : 30 : Move out with 6 sairs and shuttle, throw down 4 additional gates. Continue making sairs until you have 8, the stop production.
8 : 45 : Second reaver
~8 : 50 : Observatory (unexact time, too lazy to restart replay for this unlike the others. I've rewatched the first 9 minutes of this replay at least 15 times)
9 : 00 : Arrive at base with shuttle, harass
9 : 02 : Leg speed
10 : 00 : Add 6th gate
10 : 47 : Templar archives, begin taking third (attempt to)
11 : 30 : begin templar production, both light and dark - Essentially, you are switching from sair/reaver to sair/dt at this point.
12 : 40 : Harass with dts
13 : 30 : attack main base with most of your army, leave reavers to defend
Once sair numbers are reduced to less than 6, stop harassing and leave them to guard against drops.
After this, you transition to a ground force, zeal/ht or goon/ht if they went lurks.
[/spoiler]
Analysis of timing :
This timing attack hits right after they are supply blocked by the earlier OL kill and before any hydras have been produced with a 3 hat - 5 hat build. Forces the building of sunkens or lings. (Hyvaa had 37/35 supply). This timing attack forced hyvaa to make 6 OL instead of 2, and 7 larva on lings to fend off the 5 zealot attack (hydralisks wouldn't have hatched before significant damage was dealt). All in all, 4 + 7 = 11 wasted larva . The lings are even more useless in this case because Violet is going sair/reaver. It also likely prevented Hyvaa from throwing down his 5th hatchery.
Vs LZgamer 7:00 +1 Speedlot rush (Weak +1 Speedzeal Timing attack)
[spoiler]
8 pylon
13 forge (vs overpool on hb ridge)
13 nex (2:13)
13 cannon
14 resume probes
14 cannon – pull 3 probes
check for mining of back patch
15 gate (3:05)
16 assim (3:21)
19 cut probes
19 core (3:46)
19 zeal (3:57)
Resume probes
23 pylon (4:14)
25 stargate (~4:45)
26 goon (4:50)
+1 attack (5:09)
29 citadel (5:19)
continuos lot production after the dragoon.
5 : 30 : Add 1 gateway at natural.
Cut probe x2
5 : 38 : Corsair to deny OL scouting - use goon + sairs to kill ol by base - you must totally deny the zerg from scouting the number of units you've made.
5 : 46 : Pylon
6 : 05 : Leg speed, Natural gas
6 : 12 : Pump lots from 2 gates
6 : 17 : Pylon
6 : 37 : pylon
6 : 44 : Move out with initial force (4 lots 1 goon) to kill lings, feint as if it's just to drive off.
~ 10 seconds later :Wait for 2 more zeals, then proceed to move in for the attack. 6 lots + 1 goon with sair focus on OL kills to block supply, rally gates to opponent's main.
6 : 58 : Templar archives goes down now, pylon
7 : 12 : Cannon main mineral line
7 : 24 : Make 2 additional gates
7 : 37 : Pylon
7 : 40 : Cease zeals, make dts from initial 2 gates - possibly for dark archon.
7 : 57 : pylon
8 : 01 : Additional Sair
Note : 6 probes are cut in this build. It could probably be done more efficiently had he not had to pull probes multiple times.
6 lots and 1 goon attack z nat with lot speed almost done and +1 4/5 of the way done. Before hydras or mutas can possibly be out vs 3 hat/5hat build. -
rally initial 2 gates to opponents nat
Make 2 dark templar to abuse any advantage or possible dark archon.
[/spoiler]
Replay links :
Violet vs Lzgamer
http://www.repdepot.net/replay.php?id=18057
Violet vs hyvaa
http://repdepot.net/replay.php?id=18196
Comparison to 3 hat spire - 5 hat hydra build timings:
The example 12 hatch replay begins hydra production around 6:55, however they skip scourge entirely and are likely better in economy than the overpool openings in the other games. (At 7:08 they only have 4 hydras, however).
The 9 pool example only has 2 hydras at 7:13, and only at 7:31 does the zerg show enough force to fight off this timing attack - harass is crucial. However, this is with 0 pressure from the opponent. There is, unfortunately, no overpool example build.
Essentially, anything you can do to slow down the z right before this should be done - if you get 2 OL kills, you can force game right here, or do enough damage to eventually win.
A better comparison would be some real games : I just pulled the example from Liqiupedia, f91 vs JF : F91 overpools. In fact, for the first 5:30 or so, it's almost exactly the same as hyvaa's earlier build (they are on the same map), however, hyvaa's superior mechanics lead to a more efficient build. At 7:02, f91's hydra den is just finishing.
Violet's timing push reaches hyvaa's 3rd at 7:11 with 4 slow lots, and 1 on the way, and another rallied and building. Throughout Violet's entire attack, he does almost no damage besides killing some lings and 1 drone. The damage is the economic hit hyvaa takes from producing so much defense, and having to stop mining at his 3rd, and also forcing him to produce a terrible unit combination vs Violet's build (14 lings vs Sair/reaver). The timing hits lzgamer at 7:39, however with 2 additional units, leg speed done, and +1 attack close to finishing. It is important to note, however, that in both of these games Violet lost several probes and had to stop mining for significant periods of time, so would in theory have had a much earlier timing attack- but harass is crucial.
At 7:30, f91 has 2 hydras and 4 lings, and a sunk just starting at his third - this is when the timing attack is supposed to hit.
Experiences at utilizing:
I’ve been trying out the build Violet used vs Lzgamer - vs 3 hatch builds in general it's not quite as effective, but it does usually do some damage. I'm still messing around with the build order, don't have it down exactly, but it's helped my build order immensely to just understand where the timing window in pvz comes in. (I've gone from a 20% win rate vs Z overall to a 40% win rate today, with my first win against an equally skilled opponent I’ve ever felt good about in pvz). The initial move out is crucial, because every toss does this just so he doesn't see 6 lings sitting in a semi-circle right outside his base, to check ling numbers, etc. Moving back after doing this convinces the Z that they have no need to make additional defense. I find this build to be a bit safer than the mass gate builds that skip corsair and much safer than the 4 gate 2 archon build, as you produce enough units to hold off early hydra busts, and research leg speed. It's VERY safe vs 2 hatch hydra as the sair is always out in time to see and 5 lots are usually available. However, it is very weak vs 2 hatch mutas. It transitions very well to late game play, and it is usually possible to take your third after dealing damage – although not always advisable to do so. I could provide a replays or two if anyone wishes. (I fail to have good enough defense against the initial aggression in one game, however my timing attack succeeds in destroying their third, but I've lost all of my probes and have to concede - that would be the one posted). The most difficult thing I have practicing this build is finding d or d+ players who know the 3 hatch 5 hatch build order and don't go 3 hatch lings.
Concluding Thoughts
The other build Violet used is far, far, far more complex, and I'm still not sure I exactly understand it, because it's heavily based in mind games. Some of my thoughts on it, although I'm likely massively wrong on some things : he skips +1 air weapons because it's a dead giveaway that you're going sair/reaver. Hyvaa is only prepared for 1 corsair, and when 2 show up he loses 2 additional overlords he wasn't planning on losing at all. In addition, the lack of +1 may be purposeful - it's a sign that an early rush is coming, and with the shorter distance to the third on desti, neither +1 or speed was necessary for the timing attack. The timing attack shortly after this forces Hyvaa into either making 14 or so lings or losing his 3rd. Hyvaa makes the lings, but the Sair/corsair counters it perfectly, and Hyvaa only has a small group of hydras ready for the initial drop. Violet makes 2 reavers the entire game(iirc), though. He never gets air weapons - he masses just enough sairs to hunt down overlords easily and to convince Hyvaa that he's comitted to mass air, so hyvaa begins to mass hydras and research burrow, also taking an early 4th (to outexpand the sair/reaver toss) while denying Violet's third with his hydras. Hyvaa attempts a doom drop on Violet's main, as it's one of the correct counter to sair/reaver, only to discover that Violet has thrown down an additional 5 gateways and converted to a zeal/ht army which cleans up his drop with almost no loses, even though the sairs are across the map. Hyvaa hurriedly makes as many lurkers as he can, but since the map is destination, he can only block the front with the limited amount of lurkers available to him, and Violet just runs around the other way and rapidly takes his third and fourth. From here, Violet's lead is just too large to be overcome.
One last thing - thanks to lzgamer for posting the replay and giving me permission to use it.
Edit :
Wow... I just realized two things. I never bothered to research what time the 4 gate 2 archon push comes to compare the exact timings, and I have no idea how to use spoiler tags.
Example replay of build at D level during Kor hours (ignore after the first 8 minutes, I suck, and should've attacked his 3rd instead) :
http://repdepot.net/replay.php?id=18204
Also:
A note on execution from Louder :
Neither of these builds are new... I'd be surprised if any B level or better P haven't known of them for a while. They're just unpopular among foreigners because of the high risk of punishment due to minor errors in execution.
Something I didn't notice in the guide is that P needs to watch Violet's micro patterns here. Notice how he does lots of his base building in bulk - 4 gates at once, archive pylon observatory at once, etc. He does these things during calculated breaks from harassment micro. It's important to remember those times as opportunities to go build stuff. For example, when sair/shuttle first move out to harass, you have numerous opportunities to jump to your main and build quickly while watching the minimap to see what is spotting your force.
|
United States42689 Posts
Great post. I intend to spend a large part of tomorrow messing around with this on Destination, I'll post some reps and give some feedback then.
|
I was wondering, as a zerg player, if I can just stop the probe from delaying my hatchery and spawning pool, and scourge the first corsair, will it be enough to stop this build from doing significant damage to my economy? What I'm saying is, "Exactly how much harassment is needed to make this build work?" Sorry if you said it in the OP, I only read it once and skimmed a second time.
|
On August 04 2009 13:22 Kwark wrote:
Great post. I intend to spend a large part of tomorrow messing around with this on Destination, I'll post some reps and give some feedback then.
Thank you. I just realized that the distance to the third is shorter on desti than on heartbreak, so the timing push is actually more effective on it. I think that's part of why the build on that map skips +1 and speed before moving out, although the lack of +1 may be another mind game. (edited this into the OP). That game vs Hyvaa is so sick... I was in awe the first 15 minutes, just watched it over and over.
Although I haven't tried it out on Desti yet personally, been sticking to heartbreak.
|
amazing. Isn't that timing you're discussing the same timing window Movie exploited in his win over Jaedong in the MSL ro32?
|
On August 04 2009 13:28 Rainbow wrote:
I was wondering, as a zerg player, if I can just stop the probe from delaying my hatchery and spawning pool, and scourge the first corsair, will it be enough to stop this build from doing significant damage to my economy? What I'm saying is, "Exactly how much harassment is needed to make this build work?" Sorry if you said it in the OP, I only read it once and skimmed a second time.
The sair should get an OL kill unless you delay your hydras, which is playing right into this build's hands.
It depends on the map and how many probes/tech the p cuts. Also on how many lings you make to deny scouting/attempt runby
On destination, they only need to delay a Z something like 7-14 seconds with a 9 pool opening, less with overpool. I couldn't figure out the timing for the 13 nexus variant of this, unfortunately, or I could say for sure. (See the section towards the end : Comparison to 3 hat spire - 5 hat hydra build timings, it has the exact timings).
If the P plays perfectly and is not harassed by Z, I'm not entirely sure if they need to delay Z at all - but the timing attack grows progressively stronger the more the Z messes up their build, so it's to the p's benefit to encourage Z to harass his probes/buildings. In essence, this is a straight test of skill as to who can execute their build better while being harassed - Violet clearly thinks it's him, but his build is also a bit more flexible than a Z build, as he can lose 3 or 4 probes, and give up lots of mining time and still make the timing window by cutting probes- and come out ahead from the attack anyways. If your opponent can read your build well, it's really hard to stop.
|
On August 04 2009 13:32 DreaM)XeRO wrote:
amazing. Isn't that timing you're discussing the same timing window Movie exploited in his win over Jaedong in the MSL ro32?
Watching now. I missed this game somehow. Must have been an 18:30 game.
No, Movie's attack comes later, right after hydra production has started. They take advantage of different timing windows - Movie attacks at this point because range goons are > hydras, but not lings, which is the reverse of the timing attack Violet uses, whose whole goal is to force lings. Violet's build comes probably 15 seconds earlier. MSL uses real seconds rather than game ones - so Movie arrives at JD's base at exactly 7 minutes in real time. Violet moves out at 7 minutes in-game time, arrives between 7:15 and 7:40 in-game time. Violet's build also can transition to mid game fairly well, while Movie's can not.
|
I can't really comment on the sair/reaver thing because to me that just looked like an ordinary build but with 4 zealots to try and harass the Z. But the 2 gateway push with no +1 is something I've been seing progamers do a lot for the past few months. It isn't really all in or risky and does have a good chance to do damage. And because they always make the 2nd gate at their nat it is really hard to scout for as Z. Well not too hard, I guess the Z can just throw 2 lings into your nat and try and see the second gateway but these days a lot of Z's are so greedy they don't even bother to sac an overlord to see your post-sair followup. It will still put you a little behind if they respond well and make enough lings, but that is about it. Not sure about the build with +1, but I assume it can be treated just like a normal +1 speedlot rush that has a stronger followup.
I was honestly curious as to why I have never seen anyone mention this build before or seen anyone below the pro scene using it. Although to me it just always seemed like: hide your second gateway at your nat, pressure with 6 speedlots, nothing meta-game changing. I have never really tried it though, because I don't care enough with my offrace to bother to learn how to sim city my nats to allow for a second gateway to comfortably fit in.
|
United States10774 Posts
thanks for the excellent write up. i have been using this build quite often for a while now, but haven't practiced it sufficiently to write something like this. great job. i am gonna experiment with this a bit more and hopefully give some feedback as well
|
On August 04 2009 13:43 Nevuk wrote:Show nested quote +On August 04 2009 13:28 Rainbow wrote:
I was wondering, as a zerg player, if I can just stop the probe from delaying my hatchery and spawning pool, and scourge the first corsair, will it be enough to stop this build from doing significant damage to my economy? What I'm saying is, "Exactly how much harassment is needed to make this build work?" Sorry if you said it in the OP, I only read it once and skimmed a second time. The sair should get an OL kill unless you delay your hydras, which is playing right into this build's hands. It depends on the map and how many probes/tech the p cuts. Also on how many lings you make to deny scouting/attempt runby On destination, they only need to delay a Z something like 7-14 seconds with a 9 pool opening, less with overpool. I couldn't figure out the timing for the 13 nexus variant of this, unfortunately, or I could say for sure. (See the section towards the end : Comparison to 3 hat spire - 5 hat hydra build timings, it has the exact timings). If the P plays perfectly and is not harassed by Z, I'm not entirely sure if they need to delay Z at all - but the timing attack grows progressively stronger the more the Z messes up their build, so it's to the p's benefit to encourage Z to harass his probes/buildings. In essence, this is a straight test of skill as to who can execute their build better while being harassed - Violet clearly thinks it's him, but his build is also a bit more flexible than a Z build, as he can lose 3 or 4 probes, and give up lots of mining time and still make the timing window by cutting probes- and come out ahead from the attack anyways. If your opponent can read your build well, it's really hard to stop.
Great writeup. I love the effort you put into it. For build timings, it's probably best to time it yourself in single player after you get the gist of it from watching a rep. You should always be able to get at least two sair kills (many times three if you get your tech up fast) as long as you produce more than one sair (his first two should be hanging somewhere around your base - always keep track of them as they start to leave and note the direction so you can hunt them down later), you just might have to fish around for them a bit. Cutting probes is a must for this build, because you need to get your sairs out ASAP so you can have as much time as possible before his spire finishes. You'll probably want to check his hatchery/gas timing with a probe or something before speed finishes to make sure he's not abusing lair tech (speed should finish roughly when you start your first sair). Normal hatchery timing is roughly when your sair finishes for an overpool, later for 9pool or speed builds. By the way, from experience, this sort of build works best vs 12hatch because you can delay your cannons a ton and start your gas/core really early, which speeds the build up a lot more than cutting a few probes.
Also, when you cut probes, you're not "losing" them, you're delaying them. It doesn't really hurt your economy and you get your tech timing when it's the most effective.
This is all from my wimpy offracing perspective, though.
|
On August 04 2009 13:52 Nevuk wrote:Show nested quote +On August 04 2009 13:32 DreaM)XeRO wrote:
amazing. Isn't that timing you're discussing the same timing window Movie exploited in his win over Jaedong in the MSL ro32? Watching now. I missed this game somehow. Must have been an 18:30 game. No, Movie's attack comes later, right after hydra production has started. They take advantage of different timing windows - Movie attacks at this point because range goons are > hydras, but not lings, which is the reverse of the timing attack Violet uses, whose whole goal is to force lings. Violet's build comes probably 15 seconds earlier. MSL uses real seconds rather than game ones - so Movie arrives at JD's base at exactly 7 minutes in real time. Violet moves out at 7 minutes in-game time, arrives between 7:15 and 7:40 in-game time. Violet's build also can transition to mid game fairly well, while Movie's can not.
quick question..what do you mean "real time" and "ingame time"
|
On August 04 2009 14:11 DreaM)XeRO wrote:Show nested quote +On August 04 2009 13:52 Nevuk wrote:On August 04 2009 13:32 DreaM)XeRO wrote:
amazing. Isn't that timing you're discussing the same timing window Movie exploited in his win over Jaedong in the MSL ro32? Watching now. I missed this game somehow. Must have been an 18:30 game. No, Movie's attack comes later, right after hydra production has started. They take advantage of different timing windows - Movie attacks at this point because range goons are > hydras, but not lings, which is the reverse of the timing attack Violet uses, whose whole goal is to force lings. Violet's build comes probably 15 seconds earlier. MSL uses real seconds rather than game ones - so Movie arrives at JD's base at exactly 7 minutes in real time. Violet moves out at 7 minutes in-game time, arrives between 7:15 and 7:40 in-game time. Violet's build also can transition to mid game fairly well, while Movie's can not. quick question..what do you mean "real time" and "ingame time"
Game seconds are considerably faster than normal seconds because everyone plays on the "fastest" setting as opposed to the "fast" (or was it "normal") setting (game second = normal second - real time). I always thought all the MSL/OSL in-game timers measured in game seconds, though...
|
On August 04 2009 14:13 Saracen wrote:Show nested quote +On August 04 2009 14:11 DreaM)XeRO wrote:On August 04 2009 13:52 Nevuk wrote:On August 04 2009 13:32 DreaM)XeRO wrote:
amazing. Isn't that timing you're discussing the same timing window Movie exploited in his win over Jaedong in the MSL ro32? Watching now. I missed this game somehow. Must have been an 18:30 game. No, Movie's attack comes later, right after hydra production has started. They take advantage of different timing windows - Movie attacks at this point because range goons are > hydras, but not lings, which is the reverse of the timing attack Violet uses, whose whole goal is to force lings. Violet's build comes probably 15 seconds earlier. MSL uses real seconds rather than game ones - so Movie arrives at JD's base at exactly 7 minutes in real time. Violet moves out at 7 minutes in-game time, arrives between 7:15 and 7:40 in-game time. Violet's build also can transition to mid game fairly well, while Movie's can not. quick question..what do you mean "real time" and "ingame time" Game seconds are considerably faster than normal seconds because everyone plays on the "fastest" setting as opposed to the "fast" (or was it "normal") setting (game second = normal second - real time). I always thought all the MSL/OSL in-game timers measured in game seconds, though...
AH This makes much more sense. Got it. So Movie's goon timing was geared for the hydra timing?
|
On August 04 2009 13:59 Saracen wrote:Show nested quote +On August 04 2009 13:43 Nevuk wrote:On August 04 2009 13:28 Rainbow wrote:
I was wondering, as a zerg player, if I can just stop the probe from delaying my hatchery and spawning pool, and scourge the first corsair, will it be enough to stop this build from doing significant damage to my economy? What I'm saying is, "Exactly how much harassment is needed to make this build work?" Sorry if you said it in the OP, I only read it once and skimmed a second time. The sair should get an OL kill unless you delay your hydras, which is playing right into this build's hands. It depends on the map and how many probes/tech the p cuts. Also on how many lings you make to deny scouting/attempt runby On destination, they only need to delay a Z something like 7-14 seconds with a 9 pool opening, less with overpool. I couldn't figure out the timing for the 13 nexus variant of this, unfortunately, or I could say for sure. (See the section towards the end : Comparison to 3 hat spire - 5 hat hydra build timings, it has the exact timings). If the P plays perfectly and is not harassed by Z, I'm not entirely sure if they need to delay Z at all - but the timing attack grows progressively stronger the more the Z messes up their build, so it's to the p's benefit to encourage Z to harass his probes/buildings. In essence, this is a straight test of skill as to who can execute their build better while being harassed - Violet clearly thinks it's him, but his build is also a bit more flexible than a Z build, as he can lose 3 or 4 probes, and give up lots of mining time and still make the timing window by cutting probes- and come out ahead from the attack anyways. If your opponent can read your build well, it's really hard to stop. Great writeup. I love the effort you put into it. For build timings, it's probably best to time it yourself in single player after you get the gist of it from watching a rep. You should always be able to get at least two sair kills (many times three if you get your tech up fast) as long as you produce more than one sair (his first two should be hanging somewhere around your base - always keep track of them as they start to leave and note the direction so you can hunt them down later), you just might have to fish around for them a bit. Cutting probes is a must for this build, because you need to get your sairs out ASAP so you can have as much time as possible before his spire finishes. You'll probably want to check his hatchery/gas timing with a probe or something before speed finishes to make sure he's not abusing lair tech (speed should finish roughly when you start your first sair). Normal hatchery timing is roughly when your sair finishes for an overpool, later for 9pool or speed builds. By the way, from experience, this sort of build works best vs 12hatch because you can delay your cannons a ton and start your gas/core really early, which speeds the build up a lot more than cutting a few probes. Also, when you cut probes, you're not "losing" them, you're delaying them. It doesn't really hurt your economy and you get your tech timing when it's the most effective. This is all from my wimpy offracing perspective, though.
When I tried the timing of it vs Comps, I wound up being slower than Violet even when I had an earlier nex. I'll go try it out a few times and see what times I get now that I have more experience with it. He also cuts pylons as well as probes, but it's harder to say where he does that because it's less blatant (He builds one at 23 and then builds the next one much later than expected, although that might be due to harass).
Just tested, I messed up the build a bit (forgot the citadel), but the same force hits the Z's natural (Leg speed hit tons later, however I had the minerals for it, but just forgot to build it, so it would've been done at the same time) at 6:32 and leaves at 6:10ish. So it's basically 40 seconds earlier, but I know it can be done better - I was at 500 minerals shortly before I moved out, and made a forge which sometimes isn't necessary vs 12 hatch. The build order would have to be tweaked for 13 nex - likely less pylon skipping and earlier 2nd gate, citadel and natural gas. Even untweaked, essentially, it would destroy the example 12 hatch build much harder than it did Lz's overpool build with 0 harassment from the P player - although I think that example build is probably outdated and gets hydras too late.
Also, about positioning - on the 4 o clock position on heartbreak ridge, the simicity violet uses makes the second gateway REALLY hard to scout short of ling suicide. (It goes in the middle of cannons), and he produces a dragoon to kill scouting overlords.
On August 04 2009 14:16 DreaM)XeRO wrote:Show nested quote +On August 04 2009 14:13 Saracen wrote:On August 04 2009 14:11 DreaM)XeRO wrote:On August 04 2009 13:52 Nevuk wrote:On August 04 2009 13:32 DreaM)XeRO wrote:
amazing. Isn't that timing you're discussing the same timing window Movie exploited in his win over Jaedong in the MSL ro32? Watching now. I missed this game somehow. Must have been an 18:30 game. No, Movie's attack comes later, right after hydra production has started. They take advantage of different timing windows - Movie attacks at this point because range goons are > hydras, but not lings, which is the reverse of the timing attack Violet uses, whose whole goal is to force lings. Violet's build comes probably 15 seconds earlier. MSL uses real seconds rather than game ones - so Movie arrives at JD's base at exactly 7 minutes in real time. Violet moves out at 7 minutes in-game time, arrives between 7:15 and 7:40 in-game time. Violet's build also can transition to mid game fairly well, while Movie's can not. quick question..what do you mean "real time" and "ingame time" Game seconds are considerably faster than normal seconds because everyone plays on the "fastest" setting as opposed to the "fast" (or was it "normal") setting (game second = normal second - real time). I always thought all the MSL/OSL in-game timers measured in game seconds, though... AH This makes much more sense. Got it. So Movie's goon timing was geared for the hydra timing?
Yep, they hit right after the first 4 hydras have hatched and it was too late for Jaedong to easily switch tech, especially since he had no ling speed.
|
Your tech timing should be based purely on your gas (and gate) timing (when you start your assim). Your nex timing doesn't affect your tech timing at all as long as you get it before your assimilator. But skipping cannons means your gas/gate will be that much faster. After you get your gas, it should take the exact same amount of time to get everything (you may have to cut probes) because the gas mining rate is pretty much constant and the tech timings are based on gas limitations.
|
On August 04 2009 14:47 Saracen wrote:
Your tech timing should be based purely on your gas (and gate) timing (when you start your assim). Your nex timing doesn't affect your tech timing at all as long as you get it before your assimilator. But skipping cannons means your gas/gate will be that much faster. After you get your gas, it should take the exact same amount of time to get everything (you may have to cut probes) because the gas mining rate is pretty much constant and the tech timings are based on gas limitations.
Yeah I had pretty much come to this realization after seeing the beauty of the 15 gate 16 assim,19 cut probes, 19 core ,19 zeal, but I'm still processing it a bit (I find this easily comparable to the 10/15 pvt build in ingenuity). I'm almost positive that's the right timing on the attack, though - I followed the tech pattern on that to the letter. (4 zeal + 1 goon with 2 rallied)
The issue essentially was that while I used all the gas, going 13 Nex gives me more minerals than Violet had, especially without pulling probes. and I'm not really sure what they should be spent on.
|
I'm practicing this on iccup now, if anyone wants to play me. I've just realized that doing the level of harass necessary to this build frequently causes zerg to go 3 hatch ling. So basically, time to learn better sim cities.
|
On August 04 2009 15:05 Nevuk wrote:
I'm practicing this on iccup now, if anyone wants to play me. I've just realized that doing the level of harass necessary to this build frequently causes zerg to go 3 hatch ling. So basically, time to learn better sim cities.
beautiful. I just played a D+ zerg on HBR with this. I scouted a 3 hatch spire and then 5 hatch hydra, i pushed out with 4 zlots and a goon, around the 9 minute mark. I got supply blocked while micro-ing a bit, but i forced him to build lings; i then transitioned into a mid-game ht/goon/zeal army while taking my third. I was at a loss for what to do after my initial attack so i macro'ed up.. Quick question. when is storm to be upgraded?
|
On August 04 2009 15:16 DreaM)XeRO wrote:Show nested quote +On August 04 2009 15:05 Nevuk wrote:
I'm practicing this on iccup now, if anyone wants to play me. I've just realized that doing the level of harass necessary to this build frequently causes zerg to go 3 hatch ling. So basically, time to learn better sim cities. beautiful. I just played a D+ zerg on HBR with this. I scouted a 3 hatch spire and then 5 hatch hydra, i pushed out with 4 zlots and a goon, around the 9 minute mark. I got supply blocked while micro-ing a bit, but i forced him to build lings; i then transitioned into a mid-game ht/goon/zeal army while taking my third. I was at a loss for what to do after my initial attack so i macro'ed up.. Quick question. when is storm to be upgraded?
I just got off a really nice game against a kor zerg, I got really far ahead but wound up losing due to my inferior macro and multitasking later on and because I didn't pull out soon enough - I killed tons of lings and forced something like 4 sunkens but didn't retreat and lost my initial force needlessly, and got contained to two bases : http://repdepot.net/replay.php?id=18204 . Rewatching this, had I gone for his 3rd, my attack would've been much more successful - attacking over the bridges into the nat in desti is brutal. Oh yeah, I'll be on under LaughingSphinx
As for storm - whenever you want, basically- this is just an opening that punishes 3 hat spire - 5 hat hydra, it can branch out however you wish. It really depends on their unit composition and if you wish to go sair/dts instead- I typically go Sair/dt and attempt to harass expos because the initial 2 dts are more useful against an immediate counter due to the length it takes storm to research and temps to get energy. If my harass with dts fails and both are left alive, I make a dark archon, research maelstrom and start ht production /storm after that point.
You should have the gas right after the attack to research storm and start 2 temps, but you can delay storm if you force lots of lings because they won't have the hydras necessary to break your speedlots soon. I prefer going heavy sair builds or getting archons (dark or light) because of a likely muta switch, and because storm won't finish in time if he comes directly at you with hydras.
|
On August 04 2009 15:53 Nevuk wrote:Show nested quote +On August 04 2009 15:16 DreaM)XeRO wrote:On August 04 2009 15:05 Nevuk wrote:
I'm practicing this on iccup now, if anyone wants to play me. I've just realized that doing the level of harass necessary to this build frequently causes zerg to go 3 hatch ling. So basically, time to learn better sim cities. beautiful. I just played a D+ zerg on HBR with this. I scouted a 3 hatch spire and then 5 hatch hydra, i pushed out with 4 zlots and a goon, around the 9 minute mark. I got supply blocked while micro-ing a bit, but i forced him to build lings; i then transitioned into a mid-game ht/goon/zeal army while taking my third. I was at a loss for what to do after my initial attack so i macro'ed up.. Quick question. when is storm to be upgraded? I just got off a really nice game against a kor zerg, I got really far ahead but wound up losing due to my inferior macro and multitasking later on and because I didn't pull out soon enough - I killed tons of lings and forced something like 4 sunkens but didn't retreat and lost my initial force needlessly, and got contained to two bases : http://repdepot.net/replay.php?id=18204 . Rewatching this, had I gone for his 3rd, my attack would've been much more successful - attacking over the bridges into the nat in desti is brutal. Oh yeah, I'll be on under LaughingSphinx As for storm - whenever you want, basically- this is just an opening that punishes 3 hat spire - 5 hat muta, it can branch out however you wish. It really depends on their unit composition and if you wish to go sair/dts instead- I typically go Sair/dt and attempt to harass expos because the initial 2 dts are more useful against an immediate counter due to the length it takes storm to research and temps to get energy. If my harass with dts fails and both are left alive, I make a dark archon, research maelstrom and start ht production /storm after that point. You should have the gas right after the attack to research storm and start 2 temps, but you can delay storm if you force lots of lings because they won't have the hydras necessary to break your speedlots soon. I prefer going heavy sair builds or getting archons (dark or light) because of a likely muta switch, and because storm won't finish in time if he comes directly at you with hydras.
icic..i love how this forces lings and sunks. And i do agree with you. Attacking the third is a great idea. 4/5 lots can take down a hatch somewhat quickly. I'm assuming you take your third at around storm timing?
|
|
|
|
|
|
EPT
