Hello TeamLiquid, I’m JayPower and this is my 3rd guide I bring you. This is a build Grubby used during the IEM Season VII Global Challenge Cologne in all of his PvP matches. It is a very defensive, but economic aggressive robo expand build. The extremely fast observer makes this build also very safe and will allow you to be reactive.
Because that IEM was pretty much all about TvZ (IMMVP vs foreign zergs), this cool PvP robo expand build has gone a bit unnoticed I feel like. That is the reason why I wanted to write a guide about it
Build overview
You start with a very normal opening (9 pylon/13 gateway/14gas). Grubby doesn´t scouts with his probe with this build because he gets the observer so fast. He does however scout around his base to look for a potential proxy 2gate. You make your first zealot at 18 followed by a 21 gas (2nd gas).
Grubby uses an interesting way to balance his mineral/gas income. When his 2nd gas is done, he pulls a drone from his first gas and puts 2 in his other. This will make it so that 2 probes are mining in both gasses. He rallies his 3rd probe back in the gas once his mineral line contains 16 workers. This will be at 25 supply.
You make another zealot after your first one is done, this will be a 23 supply. Note that if everything lines up nicely, you will be able to make your 2nd zealot a few seconds before your cybercore is done. If your opponent has a probe in your base, you might want to delay your 2nd zealot by 1 or 2 seconds to trick your opponent into thinking you’re making a stalker or sentry rather than letting him know you’re going 100% for a 2nd zealot.
At 26 supply you make the robo, then make a sentry as soon as you can afford it. At 30 supply you should be able to afford a 2nd gate and after that you will make a 2nd sentry. You should be able to afford it right as your 1st one is done. These two sentries and the two zealots will be the core of your early game defense. When your robo is finished, you want to chronoboost out an observer and send it directly towards your opponent’s base.
Your opponent may pressure you with a zealot/stalker, 3 stalker rush or any other amount of units. This might mean that your opponent is rushing you. This could be 4gate, a 3gate pressure or maybe just a poke with a handful of units before warpgate research is done. When this happens, you absolutely have to get an immortal before your nexus. Use your sentries to hold your ramp and warp in units from your 2 gateways if you think your opponent is going to commit to an attack.
You can tell if he’s going to commit if he’s trying to push up your ramp even though you used a forcefield. If your opponent brought any probes, it is more likely that he is going to attack, it could be just for scouting however. Even if your opponent throws down a pylon near your ramp, it may not always mean he’s going to attack, don’t forget that he can always cancel the pylon(s).
The reason why this is so important is because don’t want warp in units from your 2 gateways before the nexus if your opponent is just pressuring you without committing to an attack. These units will be unnecessary and it will only delay your nexus, it might even set you behind if your opponent is going for a pressure while expanding type of build.
So if you see any pressure, make an immortal before the nexus. If you suspect an attack, make units right away out of your 2 warpgates. Else put down the nexus first and make an immortal right after it. Don’t forget that your observer is already heading towards your opponents base.
After you established your natural there is a lot of things you can do which all depend on what your opponent is doing. If your opponent didn´t expand and is going for a 1 base all-in, most of the time you will have to get a 3rd gateway really fast and chronoboost out immortals. Against a colossus all-in or a 4gate with phoenix, you want to be more stalker heavy than sentry. Against other attacks you really want to use sentries to hold the choke/ramp at your natural.
If your opponent did expand, there is a very strong follow-up push you can do. It is essentially a 6 gate blink attack with 2-4 immortals. The number of immortals depends on how fast your opponent is going for a colossus transition however. If you wait for 4 immortals and your opponent is going for double robo colossus after 1 immortal, you might have missed your timing and your opponent´s colossus count could be too high. So your goal here is to hit before your opponent hits the critical mid-game colossus count where you pretty much have no choice but to go colossus yourself.
Make sure you bring a warp prism even though you may have proxy pylons up. If you can´t break your opponent´s front, a drop in the main base can abuse the immobility of your opponent.
Video tutorial
For this guide I made a video tutorial where I explain the build and go over replays of Grubby and myself.
Part 1:
Part 2:
VODs and Replays (will keep this updated, short description of every replay/VOD will be added shortly)
Immortal expand is pretty cool, but i don't think it's a great build for all maps, so you might want to mention that. On blink obs maps like Antiga and CK, holding off the stalkers can be really tough, and there's plenty of professional games that show it: MC vs BeSt (or San not sure) on Antiga from OSL by.Rain vs PartinG on Cloud Kingdom from wcs asia finals.
Also, vs one base colossus you want to be zealot heavy rather than stalker heavy. Stalkers without blink, forcefield support, or the ability to kite, are terrible combat units.
Actually it's quite the opposite. Immortal expand is very good against blink play. MC vs BeSt and Rain vs PartinG were bad examples imo. And the defending ones weren't splitting their armies well enough.
I didn't even try to raise points why am I sure immo expand is better, so just read this
On October 20 2012 02:32 Teoita wrote: Immortal expand is pretty cool, but i don't think it's a great build for all maps, so you might want to mention that. On blink obs maps like Antiga and CK, holding off the stalkers can be really tough, and there's plenty of professional games that show it: MC vs BeSt (or San not sure) on Antiga from OSL by.Rain vs PartinG on Cloud Kingdom from wcs asia finals.
Also, vs one base colossus you want to be zealot heavy rather than stalker heavy. Stalkers without blink, forcefield support, or the ability to kite, are terrible combat units.
^This
Also, I havn't watched the video yet nor watched the replays. But seems as if it would be more rough than you're alluding to deal with Stargate openers. At least that's the problem that I keep running into when opening zealot/zealot/sentry/sentry. Your opponent scouts the lack of an early stalker immediately, and knows that going Stargate is safe before they'd ever want to put down a Stargate because you can't do any early warp-gate shenanigans.
Edit: @Synd: It's not an insta-game over or anything vs Blink on spread maps with back doors... Just ridiculously in the Blink/obs player's favor because they are guaranteed damage if they play it well and time it well if you put down that nexus.
On October 20 2012 02:48 synd wrote: Actually it's quite the opposite. Immortal expand is very good against blink play. MC vs BeSt and Rain vs PartinG were bad examples imo. And the defending ones weren't splitting their armies well enough.
I didn't even try to raise points why am I sure immo expand is better, so just read this
I've read that thread, thank you very much.
How were they bad examples? Mc vs Best as well as Stork vs herO.join from proleague finals in particular are really good to point out how tough holding with immortal expand is. On Antiga even if you split your army perfectly, the blink/obs player can just add a warp prism and sentry reactively, split your army in half and kill both easily, just like MC did, and it's doable on other maps too (Stork did it on Ohana). Rain showed that blink/obs when executed well can just tear anyone apart if you don't snipe the obs asap. Notice how Rain attacked the nat and then went straight to the main the instant his blink finished, at which point no expand build can afford to split up their army as they will have only one immortal and a handful of gateway units.
Off the top of my head, the only recent high level pro game in which immortal expand held was some Creator came from wcs Korea. The higher up you go, the stronger blink play gets and the harder it is to hold on to the expasion.
I'm not saying it's bad, and in fact it's my go to opening on Daybreak and Entombed, and i've been able to hold my nexus myself vs even a GM player on Antiga, but it's not nearly as easy as you make it sound.
Thanks a lot, I will try to learn this style on certain maps as I play many allins in PvP. (Don't like the matchup at all...) Especially this Blink/Immortal follow-up sounds very, very interesting. I will give it a try soon!
I don't think you can compare strategies with unit composition. You mention the game between Rain and Parting where Parting went for a robo expand against a blink obs build from Rain. However Parting went for a 3 stalker rush, which is completely different from this build. The 3 stalker rush delays your tech by quite a bit. Parting got his robo at 5:30 in that game, however with this build you get your robo at 4:10. Parting just had his 2nd immortal finished up when Rain blinked into his main (where he completely was unprepared for). Anyways, you can already imagine how much different defending against blink obs is when your robo is almost a minute and a half earlier.
I think that's a pretty smart and safe build. I very much like the 2nd zealot made to look like a stalker. But that assumes your opponent is good. Sometimes your opponent isn't paying attention and keeps his probe too long in your base.. and if he sees a second zealot, that's free scouting that you can't get ride of..
The nice thing is that by going zealot zealot sentry sentry, you pile up a lot of minerals to afford a quick nexus, a robo and more gates. The fast obs helps to scout and you're safe from 4 gates in case he's agressive, with 3 force fields.
If he's teching to fast colossus, your obs should be there in time to scout it and go colossus at a not too delayed timing..
However, and here I'm theorycrafting, it sounds like it'd have trouble against pheonix openings.
Playing against phoenix can be quite difficult. The key here is to have your units split up perfectly when your opponent flies in with his first 4-5 phoenix. Against a phoenix -> expand build it's important to not lose a lot of probes, against a phoenix -> 1 base all-in it's important to not lose sentries. When the all-in comes you will need to use the ramp/choke at your natural in your advantage.
I'm not too sure how this would work out vs a phoenix into expand, economy wise. For example if your opponent opens phoenix and expands right after it, he will need sentries to defend it else he will lose you just attack with your immortals and 3 gates (this is assuming he spent the majority of his phoenix energy). This will delay his robo tech so much that I think the 6gate with blink will be hard to hold for the phoenix player. I'm not completely sure though. I will try to find/play games where my opponent when for a expand after phoenix.
Here is one replay where my opponent went phoenix and his expansion wasn't very far off in comparison to mine. He followed it up with a robo but was quite low on units in the early game (I guess with the phoenix out you can do that), so im not sure if I could've pushed. I did lose a lot of probes in that game, but most of that was poor defense on my part rather than being low on units due to the build. Please note however that I kept my observer at home to spot his army which was a big mistake. This resulted in me seeing the phoenix very late and kind of defeats the purpose of getting the fast robo. My opponent tried going for colossi after 3 immortals, but I ended the game with a 6gate blink push before he had any colossi out. Please leave your thoughts/replays about playing against phoenix, maybe im not considering every possibility.
So again, I'm not sure how this would work out against a phoenix into expand type of build. I am quite sure that you can hold a 4gate after phoenix as long as you don't lose the majority of your sentries. The 2nd part of the video tutorial was dedicated to phoenix openings that get followed up with a 1base all-in, because I thought myself that it would be difficult to play against, but it isn't nearly as scary as it looks.
I was more thinking of the pheonix into macro harass. Pheonix will scout the fast expand very quickly so his own expo won't be delayed. However he'll be able to harass mineral lines for a long time since blink is so far away.
The reason I mention that is because I'm one of those players who open pheonix frequently. And when my opponent goes robo macro, I usually win pretty easily. They have to get a forge and cannons, or a lot of stalkers, and even with a lot of stalkers you can always get some probe kills for free for a long time. He can't cover all locations, at least until a certain critical mass of blink stalkers. But if he does that, it slows down his tech anyway..
I have experimented with putting 2 guys in 2 gas quite a bit. and IMO there is no point to it. Pretty much any mineral/gas tradeoff you want can be achieved by building your gas geysers earlier or later. While yes, 2 guys on each gas and mineral patch is the most efficient use of your probes, most efficent does not mean best. You can get a better economy by only building one guyser (or none at all) to have the minerals to get your expansion up faster. You can hit tech timings better by getting them very early, possibly even delaying a probe. It is good to see Grubby experimenting with it, but he is new to the game. It is much like the extractor trick, a stylistic choice that offers no real benefit, and possible negatives.
On October 20 2012 05:49 GoldenH wrote: I have experimented with putting 2 guys in 2 gas quite a bit. and IMO there is no point to it. Pretty much any mineral/gas tradeoff you want can be achieved by building your gas geysers earlier or later. While yes, 2 guys on each gas and mineral patch is the most efficient use of your probes, most efficent does not mean best. You can get a better economy by only building one guyser (or none at all) to have the minerals to get your expansion up faster. You can hit tech timings better by getting them very early, possibly even delaying a probe. It is good to see Grubby experimenting with it, but he is new to the game. It is much like the extractor trick, a stylistic choice that offers no real benefit, and possible negatives.
What are you talking about? 2 probes in gas is very common both in PvT and PvP, and was pioneered by Sase, then used by all StartaleQ players, then spread to all Protoss. It does have a discernible advantage. I tested it and it came out to about 24ish minerals extra income iirc.
I don't think you can compare strategies with unit composition. You mention the game between Rain and Parting where Parting went for a robo expand against a blink obs build from Rain. However Parting went for a 3 stalker rush, which is completely different from this build. The 3 stalker rush delays your tech by quite a bit. Parting got his robo at 5:30 in that game, however with this build you get your robo at 4:10. Parting just had his 2nd immortal finished up when Rain blinked into his main (where he completely was unprepared for). Anyways, you can already imagine how much different defending against blink obs is when your robo is almost a minute and a half earlier.
Getting 2 immortals that fast opens up another can of worms in the first place, namely phoenix play.
Monk's thread about how pvp builds interact was from a time when blink/obs was done off a 3gate robo opening, slowing the blink timing down a lot. Nowadays, blink is much faster and the blink/obs timing hits when blink is done and one observer is up. On some maps, this means it's really easy to attack in 2 places at once when you can't defend both (7 blink stalkers destroy an immortal ridicolously fast so getting 2 immortals asap doesn't change things that much), especially because the 2nd immortal means you will have a lower gateway unit count.
I'm not saying immortal expand is bad, but it's definitely not "strong" or "better" vs blink/obs. At best, it comes down to the execution of the blink/obs player. From experience, up to mid/top master it's definitely holdable. Go to the top level however, and the blink/obs play will be so good it's incredibly hard to hold it; indeed, that's the reason blink/obs is considered the "best" pvp build. It has no counters but at the same time it's really hard to be good enough to execute it perfectly.
I prefer to stay a bit higher on stalkers with this kind of build. Basing your defense on sentries is flimsy against blink obs and against phoenix and can just cause you a loss against them imo. I just open zealot-stalker-stalker (for the rest same build) and then I followup depending on what I scout with obs and stalkers. Stalkers have some advantages over fast sentry imo: - you can kill the probe and prevent scouting, with a sentry before stalker there is no real way to stop them scouting your robo if they don't move away the probe in preparation for it - don't need to make a second zealot which I think is pretty useless that early - get some map presence, you can basically scout a quick FE by them with it which is generally a soft counter to your 2 gate robo fe - much better against phoenix (I don't want any sentries against that) and i actually prefer no sentries against blink+obs either. Against other stuff I can either make the sentries later or just expand without sentries.
Pretty much I think sentries are pretty useless in pvp, you don't actually need them to stop a 4 gate if you go very fast robo and against most aggresive attacks they are useless too (blink obs, phoenix, colossus). With many builds of course you need 1 sentry to be safe, not so much with a very fast robo build.
I got a question (that I think) is relevant since the opening is somewhat similar - with LiquidHero's stargate opener, he also opens the game with 2 zealots with sentries, maybe I'm just not good enough, but I feel like its a pretty much an auto loss build against a proxied robo 4gate prism play. This is from my own experience, as you have no stalkers, you can't just go around scouting/gaining map control with just zealot sentries, which means you'll always be blind to it. Or am I missing something here?
Though the significant difference in this build which is better compared to the stargate opener is that the 2nd gas is much later, so opponents can't be certain that you are going for the tech/expand route. But you know, on ladder, there's always those players that do a predetermined blind all-in no matter what he sees -_-
Whao. I used to actually do this exact build, because I found it hard to produce off 3 gates and a robo while expanding so I cut one gateway, but I kept dying to 4 gate + warp prism, or 3 gate stalker, does work OK against a straight up 4 gate if you FF perfectly though.
Been using this build since 1 year ago (took a white-ra build and modified it), but with 3 stalkers instead of the zealots. The fast obs+nexus is the key.
Is, like most PvP builds, blind advantage vs some builds(blink, map dependant, any type of slow robo FE),blind almost death sentence vs others (prism 4g; phoenix 4g); and just veryrough vs some builds (phoenix+robo timing; 4gate+1collosi push; archon+charge push).
Funny, I've just found an old German show (Honor greift durch #028) that features this build (from minute 5:30 on). Honor is training this build vs Seiplo. He manages to stop an early blink Stalker attack in the second game, after dying to it the first time. I've tried this build from my own notes and liked it. Thank you for your work, I'll go further with this build.
On October 20 2012 08:26 Zealot Lord wrote: I got a question (that I think) is relevant since the opening is somewhat similar - with LiquidHero's stargate opener, he also opens the game with 2 zealots with sentries, maybe I'm just not good enough, but I feel like its a pretty much an auto loss build against a proxied robo 4gate prism play. This is from my own experience, as you have no stalkers, you can't just go around scouting/gaining map control with just zealot sentries, which means you'll always be blind to it. Or am I missing something here?
Though the significant difference in this build which is better compared to the stargate opener is that the 2nd gas is much later, so opponents can't be certain that you are going for the tech/expand route. But you know, on ladder, there's always those players that do a predetermined blind all-in no matter what he sees -_-
What I think is missing from this post that alleviates your issues is a probe scout and proxy pylon in your opponents side of the map. If you have a probe run up your opponents ramp at 5:00 you can see the aggression coming because he won't have a 2nd stalker building and will have 4 gates. This = warp prism usually. No one 4 gates without that second stalker unless it's warp prism.
Edit: you can also have this pylon in a common proxy robo location for an added scout. The key to this proxy pylon is that it's outside your base and off of common attack paths. If your opponent kills it he can't also be attacking you at the same time. If he does attack you, use this pylon to throw zealots to his mineral line while FF on your ramp for the free win.
Very cool variation. I do something similar but with subtle differences.
I like the idea of tricking the opponent into thinking you're getting a stalker, which forces their probe out (most of the time) and allows you to use extra gas for an earlier sentry and robo.
However, I do think that scouting on 13 is very helpful (although not necessary) because it can give you a clue to whether or not they are going to 4gate, which can clue you to get an earlier immortal or distribute chrono to other buildings. I also like to scout popular proxy robo locations on my way back home with the probe if I'm suspicious.
Against phoenix it does great because either your expansion is up earlier and so you can remake the probes you lose to the phoenix faster, or they go all-in and you have the economy to trump their attack.
On October 20 2012 08:26 Zealot Lord wrote: I got a question (that I think) is relevant since the opening is somewhat similar - with LiquidHero's stargate opener, he also opens the game with 2 zealots with sentries, maybe I'm just not good enough, but I feel like its a pretty much an auto loss build against a proxied robo 4gate prism play. This is from my own experience, as you have no stalkers, you can't just go around scouting/gaining map control with just zealot sentries, which means you'll always be blind to it. Or am I missing something here?
Though the significant difference in this build which is better compared to the stargate opener is that the 2nd gas is much later, so opponents can't be certain that you are going for the tech/expand route. But you know, on ladder, there's always those players that do a predetermined blind all-in no matter what he sees -_-
What I think is missing from this post that alleviates your issues is a probe scout and proxy pylon in your opponents side of the map. If you have a probe run up your opponents ramp at 5:00 you can see the aggression coming because he won't have a 2nd stalker building and will have 4 gates. This = warp prism usually. No one 4 gates without that second stalker unless it's warp prism.
Edit: you can also have this pylon in a common proxy robo location for an added scout. The key to this proxy pylon is that it's outside your base and off of common attack paths. If your opponent kills it he can't also be attacking you at the same time. If he does attack you, use this pylon to throw zealots to his mineral line while FF on your ramp for the free win.
On October 23 2012 04:59 Bahamuth wrote: What's the opinion on this build versus opening 3 Stalker rush into Sentry Sentry? I'm pretty sure I've seen other pro's do that too.
Teolita, don't you do this build on Ohana? I'd think the ramp and the fact that it's not amazing for Blink would make it pretty strong there.
I like to open phoenix on maps where blink/obs isn't an obvious choice.
On Ohana yeah it's 50/50. It's hard to blink in and out, but doable.
Regarding 3stalker intro 4 sentries, yeah that's a popular variation. It relies on being really good at reading your opponent just by poking around or using early scouting information, as going nexus before robo forces you to go immortal before obs if you want to hold any potential pure blink timing.
@Asmodeusx: since when exactly does immortal expand beat phoenix 9 out of 10 times exactly? Phoenix is actually really strong vs this, as your stalker count will be diminished and you are more or less always forced into building one immortal which is a completely useless investment. Wether it's phoenix/robo macro or phoenix 4gate, this build is behind a stargate opener.
EPT![[image loading]](http://i.imgur.com/g8BDL.png)
![[image loading]](http://i.imgur.com/GpBlc.png)
![[image loading]](http://i.imgur.com/tyjlH.png)
![[image loading]](http://i.imgur.com/YrMxu.png)
![[image loading]](http://i.imgur.com/glEaO.png)
![[image loading]](http://i.imgur.com/OdFUu.png)
![[image loading]](http://i.imgur.com/yMd9e.png)
