Recently i saw more and more Zergs blocking the Nat of the Protoss using a hatch.
I saw it in SPL, Vibe did it in MLG and Nerchio kind of made this to his standard.
Build is like:
rally 13th drone to scout (on 2 player maps) 14 pool, 16 try to get own nat, if it is pylon blocked => build hatch at protoss nat (only in case of Forge/nexus first expand). Queen, 4 lings get own nat ~4 (if later, consider taking 100 gas, because main is oversaturated anyways).
if protoss attacks the building hatch, cancel and replace with evo, else let it finish. The creep of a finished hatch will block the nat for ages ..
Advantages: * distorts P's build and timings massively, throws him off * each 12(chrono'd?)-15 seconds delay of P's expansion costs him a not-produced pobe. * you get your nat down before the Protoss even if he blocks your nat * P often attacks the hatch with probes (=lost mining time) * Hatch helps scouting
Disadvantage: * Delays your own eco ~45 ingame seconds. * susceptible to early allin
What do think, guys ?
a) Is its worth it ? b) should i let the hatch finish in case or cancel right before an put down evo OR cancel hatch immediately and replace by evo ?
no, toss can just send 4 probes attacking it (hatch hp will stay still) and get a cannon, i actually like when zergs do this to me cause i always end up ahead
On July 10 2012 04:01 RealRook wrote: no, toss can just send 4 probes attacking it (hatch hp will stay still) and get a cannon, i actually like when zergs do this to me cause i always end up ahead
But why do pros it with reasonable success then ?. 4 probes will not kill the building hatch, so it finishes and it takes >1 minute after the hatch kill until you can put down your nexus (creep). This means you have to oversaturate you main and got > 6 probes less because of the delayed nexus.
On July 10 2012 03:59 Toastie.NL wrote: cancel and turn into Evo, use the broodlings that spawn later to get scouting intel or kill something.
Hm, you get 225 back for the cancel and spend 75 for the evo, so overall the blocking cost 150+drone =200. If the hatch can finish it costs you 350 but creep delays the nat for ~1 more minute (for 150 minerals cost). However if the hatch is attacked and will die, doing the evo thing probably is the best solution. Nerchio seems to play this different depending on how fast he can kill the block pylon at his nat. In case its free he seems to do the evo replacement to get his nat down faster. I the protoss does a heavy block using an additional gate/pylon, it might be better to just let the hatch finish, but i dunno. Somebody doing this regulary ?
when I hatch block I: overpool -send drone after pool (early timing negates long rush distance/4p map, you can send it out later if 2p map) drone to 14 -my drone scouts them. on 4p maps, the overlord should be able to determine which base they are in, drone goes to them. drone confirms forge or nexus first. if forge first, IMMEDIATELY put hatch down. if nexus first, lings (nexus first can't really hold overpool). if gateway first, expand back home and play standard. -if the hatch gets up, queen -> creep tumor -if cannon/zealot gets up, cancel -> evo chamber lings (from overpool timing) will be able to run up the ramp, depending on when/where he put his cannon down, the lings may or may not be able to get the cannon. if you don't get the cannon, you can still run up into the main and force probes off the line.
in my experience, the advantage is toss being delayed around ~1 game time. you're only delayed ~30 seconds (hopefully droning/expanding behind this, gas for speed/roaches may be necessary if he all-ins). if you land the tumor, robo play is guaranteed to get the obs to expand, unless the tumor is in detection range of the cannon (just place it in back). with my particular overpool timing, since lings are guaranteed to run up into the main, we kill probes/delay mining, and see exactly what is going on. even if the hatch doesn't last, production facilities are delayed, and it is more likely to force another gateway unit if the probes didn't finish killing the lings. imo, it puts you ahead assuming you hit your injects and drone up back home.
a) definitely worth it, assuming you start the transition the moment the hatch is down and lings are out. it just may take practice to get a solid transition (obviously taking a third base is not likely, because toss will either all in immediately which you can't hold on 3 base no tech/gas, and you already invested 500+ minerals early for the block). b) i cancel my hatch if the cannon is in range or a zealot gets out that i can't kill with the 6 lings. on some maps, like antiga or entombed, depending on their wall and the cannon's positioning, it may not be in range. the hatch -> evo chamber is block enough, the queen is really a stretch. if you keep the lings alive, you can place them on the ramp just as the evo dies and use the broodlings/lings to focus something down. other thing is canceling the evo as it is about to finish and then go grab a gas in the main if it isn't already taken (forcing toss to put gases down at a natural that is delayed, and overlords by the main can peek in anyway)
overall good play to keep the toss honest. I'd recommend following it up with 2 base play -> third when safe (scout no all in), and 1 gas when the natural goes down. the gas is either for ling speed to continue with map control, or can be used for roaches if he all ins. hope this helps.
On July 10 2012 04:15 Andromedan wrote: when I hatch block I: overpool -send drone after pool (early timing negates long rush distance/4p map, you can send it out later if 2p map) drone to 14 -my drone scouts them. on 4p maps, the overlord should be able to determine which base they are in, drone goes to them. drone confirms forge or nexus first. if forge first, IMMEDIATELY put hatch down. if nexus first, lings (nexus first can't really hold overpool). if gateway first, expand back home and play standard. -if the hatch gets up, queen -> creep tumor -if cannon/zealot gets up, cancel -> evo chamber lings (from overpool timing) will be able to run up the ramp, depending on when/where he put his cannon down, the lings may or may not be able to get the cannon. if you don't get the cannon, you can still run up into the main and force probes off the line.
in my experience, the advantage is toss being delayed around ~1 game time. you're only delayed ~30 seconds (hopefully droning/expanding behind this, gas for speed/roaches may be necessary if he all-ins). if you land the tumor, robo play is guaranteed to get the obs to expand, unless the tumor is in detection range of the cannon (just place it in back). with my particular overpool timing, since lings are guaranteed to run up into the main, we kill probes/delay mining, and see exactly what is going on. even if the hatch doesn't last, production facilities are delayed, and it is more likely to force another gateway unit if the probes didn't finish killing the lings. imo, it puts you ahead assuming you hit your injects and drone up back home.
a) definitely worth it, assuming you start the transition the moment the hatch is down and lings are out. it just may take practice to get a solid transition (obviously taking a third base is not likely, because toss will either all in immediately which you can't hold on 3 base no tech/gas, and you already invested 500+ minerals early for the block). b) i cancel my hatch if the cannon is in range or a zealot gets out that i can't kill with the 6 lings. on some maps, like antiga or entombed, depending on their wall and the cannon's positioning, it may not be in range. the hatch -> evo chamber is block enough, the queen is really a stretch. if you keep the lings alive, you can place them on the ramp just as the evo dies and use the broodlings/lings to focus something down. other thing is canceling the evo as it is about to finish and then go grab a gas in the main if it isn't already taken (forcing toss to put gases down at a natural that is delayed, and overlords by the main can peek in anyway)
overall good play to keep the toss honest. I'd recommend following it up with 2 base play -> third when safe (scout no all in), and 1 gas when the natural goes down. the gas is either for ling speed to continue with map control, or can be used for roaches if he all ins. hope this helps.
Thanx man, sounds like a solid game plan . Seems like a good alternative to standard: surfing the edge of losing to a 2 base all-in with 3 hatch no gas all drones. I'll play around with that
I've actually run into this more and more in the past few days, and as a Protoss player, I simply don't understand the logic behind it. My cannon is in range, so that stops the Hatchery from finishing (if that is the Zerg's plan). The worst case scenario is I don't get my Nexus down, so I either just get 2 gases, or my Gateway/Cyber sooner, and then mass chronoboost probes when I get my Nexus out to catch up on economy. Plus, you can't rely on it, but sometimes the Zerg doesn't cancel and just loses 300 minerals...
Honestly, either way, this seems like a loss of mining/expanding/resources for the Zerg.
I keep a probe down in my nat to build, so as soon as I see a drone I follow it with that probe and send another if it looks like it's trying to plant something. Works nearly all of the time, and it's an easy fix to this really annoying strat.
I have had this happen to me a few times and from the replays it seems that any benefit that can be gained from it is entirely dependent on a ratio between how well Zerg executes and how well Protoss reacts. The first time it was a little rough but only because I was unfamiliar with it. The subsequent times I just got pylons and forge/cannon/Nexus in different order and it didn't seem to make much difference when both players executed properly. I noted slightly later timings for both sides but it didn't really affect the flow of play; it more stalled it.
My opinion is that it's a risky idea since against a FFE you should get a fast third and if you do this hatch block you don't have as significant a worker lead over Protoss as you would have because your nat and third are later and you spent minerals on cancel costs.
If protoss underreacts and lets the hatch finish, you win the game if you get a queen out, pop down a creep tumor and the toss can never expand in their natural. If protoss overreacts and makes 2 cannons, he will have 150 minerals more in static defense + delayed nexus. If he pulls probes u can sometimes just cancel the hatch and its essentially 75 minerals lost for you vs all the mining time he lost.
Also, you throw tosses off of any buildorder they were planning to do, which is huge as protosses really rely on their thought out refined 2base/3base builds.
On July 10 2012 03:59 Toastie.NL wrote: cancel and turn into Evo, use the broodlings that spawn later to get scouting intel or kill something.
Hm, you get 225 back for the cancel and spend 75 for the evo, so overall the blocking cost 150+drone =200. If the hatch can finish it costs you 350 but creep delays the nat for ~1 more minute (for 150 minerals cost). However if the hatch is attacked and will die, doing the evo thing probably is the best solution. Nerchio seems to play this different depending on how fast he can kill the block pylon at his nat. In case its free he seems to do the evo replacement to get his nat down faster. I the protoss does a heavy block using an additional gate/pylon, it might be better to just let the hatch finish, but i dunno. Somebody doing this regulary ?
It's not that simple. You also have to count the resources that you could have mined at your natural but didn't because you put the hatch down at their natural instead of your own. More importantly, you also have to factor in the larva you lose. This is absolutely huge, you now only have one hatch worth of larva instead of two. You can't possibly make anywhere near as many drones as you could have, which means that even when you get your expo up it will be mining far less. Protoss can continue to make probes, so when he gets his expansion up, he'll have a lot more workers for it.
I think if you're going to do this, you should always do the immediate cancel and put an evo chamber down. That way you only lose 125 minerals and your expansion is not delayed that much. If you just put the hatch down with the intent of canceling it at the last second or when it's about to be destroyed, you sink 300 minerals into it until you actually cancel. As a result, your expansion will be severely delayed. As Zerg, when you delay your expo this way, you lose mining time and larva production, while the Protoss only loses mining time.
I do this whenever I can, even if not pylon blocked. i place proxy hatch at 16, after pool, as has been mentioned. I virtually always let the hatch finish, I will elaborate below.
My favorite variation is to go to 6 lings to kill the pylon at natural, then just wait outside the toss natural for them to kill the hatch. If there's just one cannon and no zealot, you can use broodlings and zerglings to kill the cannon and force a probe pull by targeting the pylon to stop a zeal from popping. In which case you add a few lings and take out everything on the low ground, or run around the main, not sure which is best but either is great for you.
If there're two cannons, I use broodling cover to sneak my 6 lings--plus lings from the larvae of the proxy hatch, properly timed--into the main.
If they're probes attacking the hatch, I send in the lings and try to runby if possible while the broodlings attack the probes and soak cannon shots. Even if he blocks the ramp and the cannon covers the nexus, 4-6 lings sitting out of cannon range threatening a runby will force him to keep probes at the ramp or drop an extra pylon to block. He will probably chrono a zealot out, so I don't like to add lings and go for a cannon kill from both sides of the wall, but if the gateway is idle (overlord nearby can spot this), it's a possibility.
I like to use the proxy hatch in natural because it gives you a lot of quick threats against two key parts of FFE--the cannon and the nexus (through hatch block). I think FFE openings are not as safe as everyone currently feels, and that zergs should be making toss worry more about their openings.
I will hopefully remember to add a few replays when I get home.
This is actually a very interesting topic. I don't think anyone has explored this field enough. At least, I have not seen a deep analysis anywhere yet. If anyone is reeeeeally interested in, it would probably take 20-30 hours to do extensive research on this.
I don't think anyone can say if it is worth it or not with 100% confidence at this point. We just don't have enough data and calculation. We can only make educated guesses.
It is probably worth trying on maps like Entombed Valley, Antiga Shipyard if protoss walls at 2nd ramp, and perhaps Ohana too. On those maps, usual cannon placement does not reach the blocking hatchery, so at least protoss has to make a 2nd cannon just for the hatchery or sacrifice tons of probe mining time. If pylon placement is terrible at first, then protoss has to make a pylon before cannon as well, which delays killing hatchery furthermore. On maps like Shakuras Plateau, normal cannon placement covers the hatchery placement, so it is most likely not work and worth it.
On July 10 2012 03:59 Toastie.NL wrote: cancel and turn into Evo, use the broodlings that spawn later to get scouting intel or kill something.
Hm, you get 225 back for the cancel and spend 75 for the evo, so overall the blocking cost 150+drone =200. If the hatch can finish it costs you 350 but creep delays the nat for ~1 more minute (for 150 minerals cost). However if the hatch is attacked and will die, doing the evo thing probably is the best solution. Nerchio seems to play this different depending on how fast he can kill the block pylon at his nat. In case its free he seems to do the evo replacement to get his nat down faster. I the protoss does a heavy block using an additional gate/pylon, it might be better to just let the hatch finish, but i dunno. Somebody doing this regulary ?
It's not that simple. You also have to count the resources that you could have mined at your natural but didn't because you put the hatch down at their natural instead of your own. More importantly, you also have to factor in the larva you lose. This is absolutely huge, you now only have one hatch worth of larva instead of two. You can't possibly make anywhere near as many drones as you could have, which means that even when you get your expo up it will be mining far less. Protoss can continue to make probes, so when he gets his expansion up, he'll have a lot more workers for it.
I think if you're going to do this, you should always do the immediate cancel and put an evo chamber down. That way you only lose 125 minerals and your expansion is not delayed that much. If you just put the hatch down with the intent of canceling it at the last second or when it's about to be destroyed, you sink 300 minerals into it until you actually cancel. As a result, your expansion will be severely delayed. As Zerg, when you delay your expo this way, you lose mining time and larva production, while the Protoss only loses mining time.
agree partially, but protoss also loses probe production because of hatch block/late nexus. since P has no queen, he cannot compensate and massively loses probe production. if i play standard (protoss ffe, pylon block, i get my nat ~3'30..4'00) most of the time P gets nat down earlier than me (translates into eco/probe advantage for P). if i hatch block, i always get my nat earlier than P (so far), so it seems it is advantageous for me. If you are pylon blocked, you anyway need to wait for lings, and will get your nat late.
On July 10 2012 06:40 6xFPCs wrote: I do this whenever I can, even if not pylon blocked. i place proxy hatch at 16, after pool, as has been mentioned. I virtually always let the hatch finish, I will elaborate below.
My favorite variation is to go to 6 lings to kill the pylon at natural, then just wait outside the toss natural for them to kill the hatch. If there's just one cannon and no zealot, you can use broodlings and zerglings to kill the cannon and force a probe pull by targeting the pylon to stop a zeal from popping. In which case you add a few lings and take out everything on the low ground, or run around the main, not sure which is best but either is great for you.
If there're two cannons, I use broodling cover to sneak my 6 lings--plus lings from the larvae of the proxy hatch, properly timed--into the main.
If they're probes attacking the hatch, I send in the lings and try to runby if possible while the broodlings attack the probes and soak cannon shots. Even if he blocks the ramp and the cannon covers the nexus, 4-6 lings sitting out of cannon range threatening a runby will force him to keep probes at the ramp or drop an extra pylon to block. He will probably chrono a zealot out, so I don't like to add lings and go for a cannon kill from both sides of the wall, but if the gateway is idle (overlord nearby can spot this), it's a possibility.
I like to use the proxy hatch in natural because it gives you a lot of quick threats against two key parts of FFE--the cannon and the nexus (through hatch block). I think FFE openings are not as safe as everyone currently feels, and that zergs should be making toss worry more about their openings.
I will hopefully remember to add a few replays when I get home.
would be very interested in replays . i haven't spend that much in-detail attention to this topic until now. in current meta, pylon block+ffe gives P a slight advantage up to ~10 minutes, that's why 2 base all ins are that succesful. it would be huge to be able to break/threaten standard ffe play
Ok, so I want to preface these replays by saying that I'm not trying to suggest this become some new standard of play. What i'm suggesting is that it is a new all-in/timing to keep FFE-only toss on their toes in the early game, which until nerchio started doing crazy stuff was widely perceived to be wholly locked down and safe. At best, I think this is something to add to your arsenal, should he forge first and you can put down the hatch. At worst, I'm grasping at straws and blues are going to eat me alive for this (in which case, let's playtest before temp ban?).
Let me be clear: I believe that proxy hatch at protoss natural only works against forge first. Should be obvious, nexus first means no place to put your hatch, and nexus first is a different story (though I still think zerg can pull 6 drones to kill that first cannon should he nexus first).
Also, I dropped to diamond at the start of this season so these games are against diamond. I was once low masters (i.e. last season), for whatever that's worth to you. Testing partners welcome (PM me). I don't present these replays as anything but food for thought.
close spawns entombed valley. he forge firsts, I put down a 16 proxy hatch at his natural. he places one cannon, i make lings from 15ish drones. as he kills hatch, i have lings ready to run in, and i shred his cannon plus some probes. from there he is in emergency mode, chronos zealot and pulls some probes, puts down two more cannons; ling reinforcements, plus lings from larvae of proxy hatch, ensure i break it all down. gg, though he could wall at top of ramp or something?
This is the only game I could find that really demonstrates what I meant. The Valley's wide ramp makes this abuse very hard to stop without them scouting and overcompensating with a giant early wall. Yes, I build a lot of lings and it makes it pretty all-in, I didn't realize this until I rewatched. But I still think that since broodling+8 lings took down the cannon even with 3-4 probes fighting, there's room for abuse on the right maps. http://drop.sc/217866
antiga close spawns. i proxy hatch at the usual 16, stopping my drone at my natural to do it. this gives him a pylon block. doesn't matter, I pump lings and shred the pylon, then the same kind of wait-outside-natural-until broodlings deal. i kill two cannons, game goes longer and we suck at macro so random stuff happens. pylon wall on high ground is interesting, me pumping lings after i took out power to gateway was a really really really bad choice. don't make bad choices.
Antiga. Wish I had a good game to show on antiga because the forward wall normally forces the first cannon to be built too far away to hit the hatch. This is a terrible game, super terrible actually. The only reason I'm including it is because it shows that you can take down two early and finished cannons (!, meh, and ! on those adjectives, respectively) by pumping lings and abusing broodlings on hatch death. Yes, probe scouting would tip them off and they could pull probes and be ahead; this is proof of concept to show that toss has to be on their toes or you'll steal a win from nowhere, even against two cannons. http://drop.sc/217868
I proxy hatch, he drops (eventually a total of) two extra cannons, weird choice. i get the queen out regardless but screw up the tumor placement. i don't scout and die to unscouted one-base. reminder: this one is mostly to see how the macro plays out for master toss who doesn't die to my shenanigans.
Just adding this one because it shows the protoss response you may need to be aware of, and I think this guy is masters, so provides a better example of toss macro when faced with this (probably better, but I didn't really check, having spent nearly an hour fishing out these three replays). He drops an extra two cannons (questionable or smart? no idea) and gets a quick zealot to be safe, then one-bases me (5gate +1? can't recall). I screw up royally even though I get a queen out by putting the creep tumor too close to a cannon. I die because I'm too cheap to sac the overlord near the natural. Don't be a cheap zerg. http://drop.sc/217867
If I may sum up my opinions: I believe that there are weak points in FFE that we haven't bothered to check for (busy 3hatching and being greedy, etc.), that proxy hatch before natural opens these up for exploitation against forge first, and that a drone pull opens up a timing against nexus first. I think that the two types of FFE openings are eminently punishable, and I have been saying for months that it's time to start making protoss worry a little more about their FFE.
EDIT: added game summaries to make it easier, and so you can yell at me if the replays are wrong.
what about 3hatch before pool at 18food (2overlords). 1 of those hatches in his natural
he can cannon rush your natural or his natural, either way his nexus is ultra delayed and you could even have queens/zerglings at his natural if he cannon rushes your natural. heck even start mining his natural
EPT
