Hello TeamLiquid, I´m JayPower and this is my 5th guide I bring you. This time it is a better sentry immortal all-in. Due to redistribution in economy you are able to get your sentries earlier and move out with 250 more energy at the same timing. This way you don’t have to slow down your push down to make additional sentries. Instead you can keep pushing and spend warpin rounds on units that actually deal damage instead of paying 100 gas for a new forcefield. Please take a look at Day9's daily on sentry immortal all-in or Remark's thread for a solid guide on execution and general tips. This is just an alteration in build order.
Build overview
You start with a 1gate core expand with double gas. You use your gateway to make 1 zealot followed by 4 sentries. While you expand you get your robo at 5:30. You can move out to your watchtower to try to fake a 5gate. I definitely recommend this to keep the zerg honest, be careful though since you’re only on 1 gateway. From here on the build is very straight forward. You make 3 immortals, 1 warp prism and an observer from the robo while getting 7 additional gateways.
Build order
9 pylon (at ramp) 11 chronoboost on probes 12 chronoboost on probes 13 gateway (at ramp) 14 gas / 15 pylon / 18 core (at ramp) 15 chronoboost on probes 19 zealot 21 gas 22 chronoboost on probes 23 pylon 24 sentry 27 chronoboost on probes 29 sentry 31 pylon (low ground) 32 sentry 35 nexus 36 sentry 36 robo 39 forge (low ground) 40 pylon (low ground) 42 warpgate done, warp in 1 sentry 42 immortal 42 chronoboost on robo 47 +1 attack from forge 48 chronoboost on natural nexus (after this just keep chronoboosting the robo) 52 immortal 60 3x gateway 60 immortal 64 pylon 64 sentry 66 4x gateway 66 warp prism 68 observer
Data / trade-off
As you can see in the graph above you start mining gas earlier and get a later nexus, this reflects upon the mineral income. But this evens out quickly when your natural is done and Parting has to get his 3rd/4th gasses and build probes to saturate those gasses. From the graph you can also see that Parting gets even in mined gas a little while before we both move out, here is where the big advantage kicks in. Parting needs a bigger gas income to reinforce with sentries while we reinforce right away with zealots and stalkers because our sentries have plenty of energy.
The move-out
The move-out you do with 5 sentries and a zealot is to fake a possible 5 gate to force unit from the zerg. It is important that you that you learn other builds that look similar in the beginning but play out different, especially in a boX format. While you move out it is important to bring a probe to scout for the 3rd base of the zerg. If the 3rd isn´t there, or very late, it could possibly be fast mutas or infestors. I recommend getting an observer after the 2nd immortal and scout what the zerg is up to. Some people prefer to still attack, kill the 3rd and don´t commit to an attack if there´s a massive spine wall. I´m not 100% sure which one I like better.
Video Guide
The video guide I made offers more in-depth analysis and visualization of situations mentioned above.
This game shows the purpose of the build very well. The zerg is droning comfortably off 4 speedlings and he would still be doing if I hadn’t move out. I forced enough units with my fake 5gate to keep the worker count even. After that the zerg couldn’t tech fast because he didn’t have the economy to support it and had to stay on low tech to get more drones out.
In this game my probe failed to make the pylon to finish my wall and the zerg had speedling killing 26 workers. For this reason my reinforcements weren´t very strong. Fortunately my opponent drones a bit too much and with a nice elevator into his main I was able to ignore most of his spines. Even with the fast infestor tech he wasn´t able to hold it.
In this game my opponent was going for a speedling heavy early/mid-game with fast infestors. I made a big mistake losing my 2nd forward pylon and the probe so fast. Fortunately My opponent didn’t have enough gas to make enough infestors so my sentries remained alive enough to help my zealot beat the zergling. Overall, more infestors wouldn’t have made a very big difference since he didn’t have spines.
About me
I'm JayPower. 19 years old and from the Netherlands. I play all races in sc2 at a master league, right now I'm playing protoss the most. I like to use unexplored/fun strategies to play since I gave up on competitive gaming a while ago. I make FPVODs of my games on my YouTube channel often and share my analysis.
Please give me feedback on the guide / video guide. I really do appreciate any feedback, so please give me suggestions on how I can improve. Any questions about the guide are welcome too, I will try to answer them as soon as possible.
This looks pretty interesting, I'll probably experiment with this build a lot in the future and see if there's any drawbacks to it that I can find. If there aren't, then this would presumably be even stronger than Parting's build because the Zerg can't play as greedy against a gateway expand as they could against a Nexus first.
A good zerg will notice that you don't take a 3rd or 4th gas and expect a 2-base all-in. I feel that a lot hangs on whether zerg overreacts to the the fake move out.
Have you faced someone going mutas? I ask this because fast muta used to be a common response against sentry-heavy gate expands (I don't know if it still is, I rarely do or see this BO), and the mutas should be up in time for the all-in. Take your third replay as an example, but imagine mutas instead of infestors.
With partings build you do the all-in after scouting a fast third, but if you are "forcing" the zerg to get fast mutas as a reaction to the 1-gate double gas expand, this could backfire.
I don't know if the muta builds takes into account a possible 5-gate pressure or what kind of overreaction you could cause by faking it.
On December 31 2012 09:47 bertu wrote: Have you faced someone going mutas? I ask this because fast muta used to be a common response against sentry-heavy gate expands (I don't know if it still is, I rarely do or see this BO), and the mutas should be up in time for the all-in. Take your third replay as an example, but imagine mutas instead of infestors.
With partings build you do the all-in after scouting a fast third, but if you are "forcing" the zerg to get fast mutas as a reaction to the 1-gate double gas expand, this could backfire.
I don't know if the muta builds takes into account a possible 5-gate pressure or what kind of overreaction you could cause by faking it.
With the small move-out that I do I scout the timing of the zergs 3rd. If there's no 3rd or a late 3rd, I get an observer after my 2nd immortal to scout my opponent. Maybe I should even cancel my 2nd immortal and go for the twillight council to be prepared to play a longer game. I'm not really sure at about it though, I honestly haven't put a lot of time to find the optimal solution for situations like these. Mostly because it's hard to tell the difference between fast mutas and fast infestors without an observer. I feel like I can kill a zerg going for fast infestors because I have more zealots and almost never more than 7 sentries than the regular build. It depends though on how many spines the zerg has and what map is being played. I don't think pushing up a ramp on shakuras with spines is a good idea. I think elavatoring should work fine vs infestors compared to vs mutas. Yes I have faced zergs going mutas and lost horribly after I kill their 3rd and trying to elavator my units past the spine wall . I'll try to put more thought into this.
On December 31 2012 09:51 WagonWheel wrote: Im very curious as to what program you used for you income graphs. Can you share that info?
I actually took Partings replay vs Sen from WCS and wrote down his income every 10sec, same with a replay from myself. Then just put them into a table in excel and made a graph in there. This is why both lines are a bit wobbly, I think my point is clear though.
Without a stalker it becomes very hard if not impossible to deny double overlord scouts. So your build becomes pretty transparent. Also, if you execute PartinG's build perfectly you can leave your base at ~8:40. I'm assuming yours leaves a lot later, which makes a big difference, especially if zerg has scouted what you are doing.
On December 31 2012 10:06 PandaTank wrote: Without a stalker it becomes very hard if not impossible to deny double overlord scouts. So your build becomes pretty transparent. Also, if you execute PartinG's build perfectly you can leave your base at ~8:40. I'm assuming yours leaves a lot later, which makes a big difference, especially if zerg has scouted what you are doing.
He actually moves out a bit earlier, you can see from the replays.
@Jay: Zerg can still go fast mutas while taking a third base at some point (obviously a later third than the standard no-gas third build), and as you said you can't know if it is mutas or infestors based only in watching his other units. So it is very hard to scout indirectly.
Even if you get a fast obs for scouting, can you even transition into something that can defend mutas? You will have already invested all the gas in 5 sentries, robo and obs, probably a immortal, and you didn't even take the third assimilator. It looks tough, or at least that is what I remember from 2010 GSL games, lol.
The best bet is forcing roaches with the fake pressure, but I don't know if zerg even needs to incorporate a roach warren at that time with a reactionary muta build to your opening.
I will ask my zerg teammate to play some games against me and see how it goes.
A big problem is that a lot of zergs when facing gate expands go for 2base tech, and any variation of 2base tech (muta, infestor) typically destroys an immortal sentry build. You will have mutas flying into your base before you're able to even move out. Also, Infestors only require 2 fungals to rid of the sentries where they can then flood lings.
While I like the idea of opening gate/core expand I think the immortal all in is the worst follow up to it. Why? Because as someone already said the common response to gate/core expansions is 2 base muta which counters your build pretty well. Your fake 1 zealot 4 sentry push might force some lings but thats about it (I dont understand how are you supposed to scout zergs third with it without losing those units?), mutas should still be on time to defend and with just 2 gas you will not be able to overwhelm the zerg with stalkers. Not to be too negative though, I gotta agree that against roach/ling/infestor it should do great.
Its cool but the biggest problem that you don't talk about enough is that you only have 2 gas. The only reinforcements you get are zealots and stalkers which makes you more vulnerable to constant pokes to get rid of your sentry energy. The fact he knows you are on 2 gas for so long with an overlord over your naturals gas would encourage him to overmake units which is probably the best decision vs your low sentry reinforcement.
Your build also moves out at least 10 sec slower and some say even slower than that, you have no 2 gas at your natural. It has a gateway opening with a light pressure so it could be worth it, but I dunno if we can completely say its better.
Also, no 2 gas at your natural gives you very few options compared to having 4 total gas. That can influence the strength of your build as its very transparent.
I like how you explained the difference between the before and after, cus I kept recounting the units and was confused why the "After" had the same units but no stalker haha
On December 31 2012 10:50 Protossking wrote: A big problem is that a lot of zergs when facing gate expands go for 2base tech, and any variation of 2base tech (muta, infestor) typically destroys an immortal sentry build. You will have mutas flying into your base before you're able to even move out. Also, Infestors only require 2 fungals to rid of the sentries where they can then flood lings.
Although I agree that 2base tech is a lot more common vs a gate-core expand, it is absolutely untrue that mutas fly into your base before you can move out. If you do have an example of a game where the zerg gets 8+ mutas in the protosses base at 8min30 I'd love to see it since I'll be stealing that build As far as mutas destroying a sentry immortal all-in I would have to agree. If the protoss commits to an attack into a spine wall it probably won't end up going well. This is why I'm looking into a different responds to zerg that go for 2 hatch tech.
As far as infestors, I'm not convinced that 2/3 hatch fast infestors can hold the all-in. You will need to have the protoss player commit to an attack into spines while you fungal it to beat it or deny the elavator (which is quite reasonable on some maps, but not all).
You're probably going to refer to suppy vs parting in WCS to back your evidence but there was a lot more to that game than just strategy. What no one seems to mention is that parting hit a massive supply block when he tried to make his 3rd immortal (evidence: http://i.imgur.com/777uM.jpg ). This on top of his late robo caused his 3rd immortal to pop out a full minute later than usual. If you don't believe me please download the replays from WCS and compare parting vs suppy on antiga with parting vs sen on ohana, or even with the timings from remark's guide. Parting's timing was late by a full minute.
On December 31 2012 10:52 syriuszonito wrote: While I like the idea of opening gate/core expand I think the immortal all in is the worst follow up to it. Why? Because as someone already said the common response to gate/core expansions is 2 base muta which counters your build pretty well. Your fake 1 zealot 4 sentry push might force some lings but thats about it (I dont understand how are you supposed to scout zergs third with it without losing those units?), mutas should still be on time to defend and with just 2 gas you will not be able to overwhelm the zerg with stalkers. Not to be too negative though, I gotta agree that against roach/ling/infestor it should do great.
I agree that 2 base muta would be the counter to this build, however like I said before I'm working on an alternative vs 2 basing zergs (the ones that go for mutas). If you don't commit to the attack I think you can still recover if you scout the spire in time.
As far as my small move out with 4 sentry 1 zealot. Although I agree that this force is extremely small and easy to kill, I find it very unusual that the zerg would have enough units out at the watchtower at the point in the game. He would have had an all-in planned for that to be the case which would be very unwise if he scouted the constant sentry production and the forge on the lowground. But it is possible and definitly game-ending. I'm not sure if you misread but you don't scout the 3rd with those units, you use a probe for that which you escort to the watchtower to increase the chance of probe reaching the 3rd. After that you send your units back home.
On December 31 2012 11:03 TheGreenMachine wrote: Its cool but the biggest problem that you don't talk about enough is that you only have 2 gas. The only reinforcements you get are zealots and stalkers which makes you more vulnerable to constant pokes to get rid of your sentry energy. The fact he knows you are on 2 gas for so long with an overlord over your naturals gas would encourage him to overmake units which is probably the best decision vs your low sentry reinforcement.
Your build also moves out at least 10 sec slower and some say even slower than that, you have no 2 gas at your natural. It has a gateway opening with a light pressure so it could be worth it, but I dunno if we can completely say its better.
Also, no 2 gas at your natural gives you very few options compared to having 4 total gas. That can influence the strength of your build as its very transparent.
I have no idea where you get the 10 sec slower from or even slower than that. Maybe from a replay I posted where I got the robo late or didn't chronoboost it perfectly? It does hit at the same time as parting though, I've compared it more than a dozen times with multiple replays.
As far as the gas count/income. I stay on 7 sentries while other go up to 10/11. This gives me 400 more gas to spend on top of my sentries having 250 more energy. My 1st and 2nd gas start very early and give me a gas lead roughly up until the 3rd immortal is done (8:30~). So at 8:30~ is where parting caught up in gas but has 400 gas more spend in sentries which takes a little less than 2 minutes. So at roughly 10:20 is where parting would jump ahead in stalker count assuming we both spend our available gas on stalkers. By that time the game is usually decided.
Looks solid if you can get a 1gate FE off without harass. Maybe this is a dumb question though, but vs 1gate FE as a Z player I will usually try to pressure with lings since I consider it to be a pretty greedy build with insufficient defense. Do you struggle vs this at all?
On December 31 2012 12:12 Defenestrator wrote: Looks solid if you can get a 1gate FE off without harass. Maybe this is a dumb question though, but vs 1gate FE as a Z player I will usually try to pressure with lings since I consider it to be a pretty greedy build with insufficient defense. Do you struggle vs this at all?
Unless you go for gas before pool for super fast speedlings, i think a one gate expo is very safe. At least on all maps with a reasonable choke point at the natural. So if you dont go blindly for gas first against protoss, i would strongly recommend you just forget about ling pressure but focus on overlord scouting and ofc the standard 2-4 lings to see if he is poking with zealot stalker or going straight for sentries. If he goes sentry before he gets a stalker you should have a good 3 min window for massive droning. Stalker zealot poke you will probably need something like 8 lings + 1-2 queens to hold it away without any discomfort.
Pretty interesting build. It appears to be similar to NonY's 2 Gate Sentry expand, but leans more towards going for the Immortal/Sentry push as opposed to going for the late game. Neat though, would love to try it out.
EPT![[image loading]](http://i.imgur.com/eAAYP.png)
![[image loading]](http://i.imgur.com/A7eXw.png)
![[image loading]](http://i.imgur.com/MbxmM.png?1)
. I'll try to put more thought into this.
