EPTAnyway, this is a (hopefully) short guide to Liquid'Tyler/NonY's 2 gate sentry expand in PvZ, which if you guys have been following him through the last 5 weeks or so he does in a large majority of his games, including the one against Darkforce in the NASL and on his stream. The build is really quite fun, and this is the first time i have ever gone over 55% win in PvZ for more than a week or so, so thank you to him!
Update September 2012
+ Show Spoiler +
PvZ has actually become my strongest matchup for the first time ever, and i am now using a slight variant of the opening that i posted, which chronos the first zealot so that you can have two zealots out before 100 gas for the first sentry, and then doing a push with 4 sentries and 3 zealots, leaving two sentries at home. It hits much faster, which is the key. On a map like Ohana, where when you get into the natural you have a 2 FF sized ramp behind you and a 1 ff sized ramp ahead of you, if there aren't already two spines finished and quite a few speedlings you can do a lot of damage, and if there are, you can just walk home sticking to the walls and be cost efficient no matter how many he throws at you while still being even economically.
I still prefer to play robo/gateway styles as opposed to stargate ones, for exactly the reasons tyler was saying on his stream a few weeks ago: It provides an army that is ALWAYS strong, unlike the normal PvZ armies that are strong for brief periods that lean towards the 'timing based' PvZ that we are used to. Get ~3 immortals, use hallu/obs/pokes to see what composition he is going for and react accordingly. If he is going spire you already have a twilight for +2 and react as usual with blink and teching to storm while taking your third. You only have to survive the mass muta flock, then you are already prepared for any tech switch, and immortals are really good against heavy spine defences.
http://drop.sc/247871
Last game i played, despite terrible micro in the fight, not using my templars at all, i still won pretty handily. I decided to play super passive that game, but it isn't necessary, i should have sharked around much more, and poked, i just felt ahead from the start, and when i scouted the timing on his third i just wanted to wait for the fast +3 attack and go.
http://drop.sc/247913
Another completely normal game, i get to his base with my poke, see roaches, do 0 damage, walk home again, check he is still massing roaches with obs, proceed to take third at around 12:30, and go for a +3 timing attack and win. This is a super safe way of playing, which is much easier than the aggressive variant, but if you want free wins, at a timing before he gets a hardcore brood army with lots of infestors and spines, this is a really strong one. Behind it i am teching to collosus and should be also heading to mothership, except he was completely dead anyway.
DISCLAIMER:
+ Show Spoiler +
Tyler does multiple variations of this opening, but I have only focused on one, which does an early poke, and then goes straight into immortal production. I have seen him on his stream do no early poke, but get a faster obs and a faster 3rd gas, but I prefer the poke, because it is more fun, and denies thirds against the players I face on ladder, who do not know/remember how to open against gateway expands, and take very greedy third bases as if facing a forge expand.
Of course, it is entirely possible that I make mistakes here about Tyler's reasoning on some things, so maybe claiming that this is Tyler's build is not 100% true, but I can only inform based on my own experience, so please do not crucify me over this!
Of course, it is entirely possible that I make mistakes here about Tyler's reasoning on some things, so maybe claiming that this is Tyler's build is not 100% true, but I can only inform based on my own experience, so please do not crucify me over this!
Intro
+ Show Spoiler +
Recently we have seen a resurgence of people wondering whether gateway expands could possibly be viable again in PvZ, obviously, Tyler still thinks so, as he always did. Before I recent a lot of people hating on the build because it isn't FFE, I am just going to say that it doesn't overly matter to me if this build could be successful for Hero in the GSL, the point is that it has been shown to work at levels way above mine (average masters), so i can still greatly improve by doing it.
There will be others saying that this only works because people don't know what to do against it, and to some extent that might be the case, but i am actually looking forward to people reacting better, will be more fun for me.
The aim of this build is to expand of of 2 gates, immediately throw down a force to wall on the low ground, followed by a gate, and then while that finishes push with 3 zealots and 6 sentries while leaving 2 sentries at home to protect against counters. This push is meant for force spines/units if the zerg is on 2 base, and if they went a fast 3 base i can often deny the 3rd with decent forcefields.
There will be others saying that this only works because people don't know what to do against it, and to some extent that might be the case, but i am actually looking forward to people reacting better, will be more fun for me.
The aim of this build is to expand of of 2 gates, immediately throw down a force to wall on the low ground, followed by a gate, and then while that finishes push with 3 zealots and 6 sentries while leaving 2 sentries at home to protect against counters. This push is meant for force spines/units if the zerg is on 2 base, and if they went a fast 3 base i can often deny the 3rd with decent forcefields.
The Build
+ Show Spoiler +
9 Pylon [DON'T SCOUT]
10 Chrono Nexus
12 Chrono Nexus
13 Gateway
14 Gas
16 Pylon [SCOUT]
17 Core
18 Zealot
21 Assimilator
22 Pylon
23 Sentry
25 Gateway
26 Sentry
29 Pylon (low ground)
33 Sentry
35 Sentry
36 Nexus
37 Pylon
Then, add a forge, then a gateway, while never stopping sentry production until you have 8. Start +1 attack as soon as the forge finishes.
On the late scout:
This build, because it does not use an early stalker, actually isn't overly affected by scouting, and if you are wondering about 6 pools, I will link to one of Tyler's most recent stream VoDs where he completely beat down a 6 pool blind by chronoing out two zealots and pulling 8 probes to defend the pylon whenever it was attacked. Tyler does sometimes scout at 13, but I am not sure what makes him choose to do one than the other, I have however found that he build order is much tighter if you scout after the second pylon.
Stream Link: http://www.twitch.tv/liquidtyler/b/316815983 The game starts at 23 minutes. Remember, you don't have to micro as well as Tyler did, and you can pull a couple more probes, and still be far ahead.
If you scout earlier than the gateway then you cannot afford to get the second gate early enough, so the build simply won't work. Here seems like an appropriate time to say that Tyler in the early game almost always rallies probes to build buildings, even if it means they don't get placed perfectly on time, because for this build MINERALS are key. You simply cannot get the the second gate out as quickly as necessary unless you get the extra 10 minerals from every building that you get by rallying instead of taking workers off of mining.
A note on the wall:
+ Show Spoiler +
Tyler likes to use the classic gate/core wall off at the ramp to his main, but i personally dislike it a lot in comparison to a pylon-core or pylon-gate wall, as demonstrated here:
![[image loading]](http://i.imgur.com/SRUFE.jpg)
This DOES, 100% block zerglings from getting through, (and often causes zergs who doubt that fact to lose zerglings), as well as, most importantly, fits archons as well. On top of that, as the gap is in the middle, running between your main and your natural is much faster, which is obviously desirable. These two facts i feel heavily outweigh the slightly disadvantage of having a larger surface area on the zealot (allowing two zerglings to fight at once), due to the fact that no-one ever builds 8+ zerglings and has them at your ramp these days before your next unit is ready.
This DOES, 100% block zerglings from getting through, (and often causes zergs who doubt that fact to lose zerglings), as well as, most importantly, fits archons as well. On top of that, as the gap is in the middle, running between your main and your natural is much faster, which is obviously desirable. These two facts i feel heavily outweigh the slightly disadvantage of having a larger surface area on the zealot (allowing two zerglings to fight at once), due to the fact that no-one ever builds 8+ zerglings and has them at your ramp these days before your next unit is ready.
Early Game Poke
+ Show Spoiler +
So what is next? I will be talking about a mix of what I like to do, as well as what Tyler does. First there is the 6 sentry 3 zealot poke, what we are aiming for here, is to poke at the third, (which you should have tried to scout with a probe.) If there is no third, walk up to the natural, usually to the boundary of the creep, and see if they are building spines. Usually they will be, because you will have chrono boosted the core a bunch of times, so this could still be a nexus cancel into 5 gate zealot sentry all in. Depending on how close the spines are to done, and how scared I feel of zerglings, I will poke about and wait for the spines to finish, and then walk home having taken no damage.
A few things to remember!
Don't just run out! I have lost two games when the zerg has been going for an all in which I would have crushed if I hadn't just moved my sentries out of a small choke. Send 1 zealot out in front to see if there are a group of roaches/banelings getting ready to bust.
Have a pylon in your main to warp in the last two zealots as you leave you natural.
If you managed to time you assimilators perfectly you might want to wait an extra second or two and warp in 1 zealot and 1 stalker to join the poke, to make it 2 zealots 6 sentries and 1 stalker. I personally prefer this, but it only seems to happen on maps where the geysers are closer to the nexus. (I should probably make a list of these, but I tend to just notice them in the game).
It is important to be fast! As with warping in a stalker there, if you have to wait for it it isn't worth it.
Forcefields are good! This is all about experience and placing forcefields slowly. You only have 6 sentries, on about half energy, so you must not waste them.
Finally, err on the side of caution. This build has the advantage that you are building sentries back at home behind this, so that if you do get caught you aren't completely frikkin dead, but you are at least 1 foot in the grave if you lose your entire poke, so if you are worried, just go home.
Once more unto the mid game my brothers!
+ Show Spoiler +
Behind the poke, you should be warping in sentries until you have around 10 total, as well as adding a robotics, a 4th gateway, and starting +1 armour. The next 100 gas will go on hallucination (yay!) which I use about once every other minute to scout what the zerg is doing. If they have a lair, I immediately build two cannons at my front in case of burrowed roaches, because there isn't time to get an observer out while building immortals.
I cannot stress this part enough: The thing that I absolutely love about this build, is that we can have enough sentries to use hallucinated phoenixes prolifically. You can take zero (give or take a stalker or two) damage to mutas if you know exactly when they are coming.
The reason this is important, is because if they are going spire you will be warping in only stalkers, and teching to blink. If they are going roach/roach+infestor, you will be doing a mix of stalkers and up to 5 immortals for the standard immortal sentry mix, only with much more energy than you would have had off of an FFE. This is where you can do a push, try and take out a 4th, and expand yourself.
The problem here is that I cannot give you a straight-forward order to exactly what you should do, because it depends on the map and what the zerg is doing. If they are going roaches, wait until you have 2 or 3 immortals, and then take your third quickly, but carefully. Remember, Hallucination is your friend, you shouldn't ever have to do anything until you know it is safe to do so. At this point you should also have an observer out in case of a burrowed roach attack. I usually build the observer at the earliest after the first immortal, but usually after the second or third.
I cannot stress this part enough: The thing that I absolutely love about this build, is that we can have enough sentries to use hallucinated phoenixes prolifically. You can take zero (give or take a stalker or two) damage to mutas if you know exactly when they are coming.
The reason this is important, is because if they are going spire you will be warping in only stalkers, and teching to blink. If they are going roach/roach+infestor, you will be doing a mix of stalkers and up to 5 immortals for the standard immortal sentry mix, only with much more energy than you would have had off of an FFE. This is where you can do a push, try and take out a 4th, and expand yourself.
The problem here is that I cannot give you a straight-forward order to exactly what you should do, because it depends on the map and what the zerg is doing. If they are going roaches, wait until you have 2 or 3 immortals, and then take your third quickly, but carefully. Remember, Hallucination is your friend, you shouldn't ever have to do anything until you know it is safe to do so. At this point you should also have an observer out in case of a burrowed roach attack. I usually build the observer at the earliest after the first immortal, but usually after the second or third.
Conclusion
+ Show Spoiler +
This build isn't meant to claim that FFE is dead in anyway, in fact, FFE is still essential on many maps, as I would never do this build without a ramp to the main, and I advise against it on GIANT maps like daybreak. However, this is to say that Gateway expands ARE viable, and this is a strong way to do them. I am bound to add more to this soon, but I thought I would gauge the initial reaction first.
Replays
+ Show Spoiler +
I am uploading the most standard games, I have a lot of stupid ones, like where the zerg massively over reacts thinking of 4 gate, and so i win, a lot of games of me holding zerg all ins, or zergs playing stupidly greedy which leads them to just die to the second push. Lastly there are some others where i do not execute the build properly at all. I will also link to some Tyler vods which should be much more informative.
Tyler vs Darkforce Game 1
The casters miscall this as a 1 gate expand i think, but this is exactly the build i am talking about.
Tyler vs Darkforce Game 2
This is one of the variants that i spoke about, which he probably did because he wanted to do something slightly different. I thought about doing this version instead, but i chose to just practice one of them really hard instead of trying to do both, as i only have limited time before my exams.
Still uploading more but here are a couple:
Game 1: http://drop.sc/171119
This was one of my earlier ones, because i only recently started using hallucinate very recently, and this is why. Lost 4 sentries in a warp prism to mutas. :o
Game 2: http://drop.sc/171118
This one was better, but again, need HALLUCINATE. When i saw roaches i simply ran away, but the early units triggered me to do a slightly earlier attack as i thought he must have droned less and therefore his 3rd would have been more vulnerable.
Game 3: http://drop.sc/171120
This is probably the best execution of the opening, but it is a good example of how if you lose your sentries, the only safe transition is into losing the game.
Game 4: http://drop.sc/171123
In this game i get greedy at the third i think i should have pulled back when i first saw roaches, so i lose a couple of sentries. We proceed into a decent game, where my macro is terribad as usual, and eventually i lose to roach hydra broodlord. Possibly could have won with better macro but hey. I'm no liquid'NonY.
Game 5: http://drop.sc/171679
Game on daybreak, the build works absolutely fine, i pressure the third, and transition into blinkstalker + immortal, once i scout only 4 gases and no spire i take my 4rd. He pressures, i lose my third but kill his army, and counter while retaking my third. Probably the best executed game of the reps so far.
Tyler vs Darkforce Game 1
The casters miscall this as a 1 gate expand i think, but this is exactly the build i am talking about.
Tyler vs Darkforce Game 2
This is one of the variants that i spoke about, which he probably did because he wanted to do something slightly different. I thought about doing this version instead, but i chose to just practice one of them really hard instead of trying to do both, as i only have limited time before my exams.
Still uploading more but here are a couple:
Game 1: http://drop.sc/171119
This was one of my earlier ones, because i only recently started using hallucinate very recently, and this is why. Lost 4 sentries in a warp prism to mutas. :o
Game 2: http://drop.sc/171118
This one was better, but again, need HALLUCINATE. When i saw roaches i simply ran away, but the early units triggered me to do a slightly earlier attack as i thought he must have droned less and therefore his 3rd would have been more vulnerable.
Game 3: http://drop.sc/171120
This is probably the best execution of the opening, but it is a good example of how if you lose your sentries, the only safe transition is into losing the game.
Game 4: http://drop.sc/171123
In this game i get greedy at the third i think i should have pulled back when i first saw roaches, so i lose a couple of sentries. We proceed into a decent game, where my macro is terribad as usual, and eventually i lose to roach hydra broodlord. Possibly could have won with better macro but hey. I'm no liquid'NonY.
Game 5: http://drop.sc/171679
Game on daybreak, the build works absolutely fine, i pressure the third, and transition into blinkstalker + immortal, once i scout only 4 gases and no spire i take my 4rd. He pressures, i lose my third but kill his army, and counter while retaking my third. Probably the best executed game of the reps so far.
