On July 20 2012 18:21 trbot wrote: People were talking in the skytoss thread about playing with a macro nexus, and I wonder if it might be a reasonable response in this case. You could keep up on probe production easily, and you wouldn't have to make a third pylon, reducing the effective cost of the nexus to 300 minerals. It wouldn't be as critical an issue to get rid of the hatch as quickly as possible, since your probe count won't be delayed, and you can distance mine from your macro nexus to get part of the revenue you would otherwise, and eventually you can build your natural to regain full revenue, and be ahead of the game in probing up for your third. Meanwhile, you are safer versus baneling busts, since a nexus takes a ridiculous 26 banelings to bust (thanks for the 1300/600 resources bro), and covers so much space (5 hexes versus the 3 of a gateway) that you need fewer forcefields to cover the remaining space. It also seems safer versus speedling busts and roach busts.
Perhaps most importantly, it's pretty easy to saturate gas even with distance mining. You could prepare some pretty dangerous DT pressure, archons, fast colossi, huge immortal sentry allin, skytoss, etc.
Just theorycrafting, but if you kill the hatch before it finishes, assuming it was placed by 3:00, you should be able to place your third at about have your third (natural) nexus finished by about 7:00 (giving about a minute for creep to recede), and have 70 probes before 10 minutes, with ZERO chronoboost (so you can spend it all on upgrades or chronoing tech units), or by 9:00 with a reasonable amount of chrono.
Yep. If you build a macro nexus when you get hatch blocked, you can get the same kind of probe production and then just build a delayed natural Nexus with some long-distance mining for a minute. This means matching drone production and having tons and tons of chronoboost. Having all that chrono lets you do some pretty crazy things, like a 110 second Mothership build time or do 7-8 gate attacks where you boost all the gates at once and have 2 warp-ins back-to-back from the chrono.
another example of idra (hatch block before pool) (@15'50)
Though he loses the game (scout fail), i think he had a big advantage early on. I think Minigun responded well and applied pressure in order to force units instead of drones. Also note idra build some lings at the proxy hatch in order to scout.
I am not convinced that the fast 3rd of idra was a good decision though. At 6'24 its 26 (Z) drones agains 23 probes. Zerg is nearly on 3 bases, P has not even started the nat(!). Anyway Idra has not the drone count to make use of 3 bases. I think he would be better off when building the 3rd say 45sec to 1 minute later to get a drone lead (2cnd queen before 3rd). ofc i am very amateur, so well ..
Like I said in my post: It's not game-winning if you get a queen out
You need to successfully plant a tumor [in a spot where cannon wont' detect it]. A reasonable level protoss can generally build a zealot in time to kill the tumor a queen will plant. Zerglings help defend, but usually by that point they'll have 2 zealots, and with some probes they should be able to win the fight.
The situation is not straightforward for either side, and there are so many factors to deal with. Overall I'd say it's certainly more risky and/or not as worth it for zerg though, particularly at the highest level, as long as the opponent knows about this tactics (back when it was a surprise strategy some non-overreacting players could fall for it)
On July 20 2012 21:10 Schnullerbacke13 wrote: I am not convinced that the fast 3rd of idra was a good decision though. At 6'24 its 26 (Z) drones agains 23 probes. Zerg is nearly on 3 bases, P has not even started the nat(!). Anyway Idra has not the drone count to make use of 3 bases. I think he would be better off when building the 3rd say 45sec to 1 minute later to get a drone lead (2cnd queen before 3rd). ofc i am very amateur, so well ..
ive had pretty high sucsess agianst this by just going 4 gate without trying to expand against this. Especially if zerg makes a queen he will have invested a LOT into hatchery & queen & his own hatchery.
Roaches are limited because of the mineral intense early game, if he gets gas, he will delay his natural by a bit more, have less minerals
Zerglings, are also limited because of the 850 mineral investment for natural, hatch and queen (since he usually needs another queen at natural.)
On July 23 2012 00:07 weikor wrote: ive had pretty high sucsess agianst this by just going 4 gate without trying to expand against this. Especially if zerg makes a queen he will have invested a LOT into hatchery & queen & his own hatchery.
Roaches are limited because of the mineral intense early game, if he gets gas, he will delay his natural by a bit more, have less minerals
Zerglings, are also limited because of the 850 mineral investment for natural, hatch and queen (since he usually needs another queen at natural.)
well, if the zerg expects it, he easily can defend (as long he did not build a queen at blocking hatch). From my experience ~60% of the time i can expect a 4 gate. Like that
I feel that whenever the protoss reacts properly it puts him ahead as the delay zerg takes (drone scouting time lost, minerals for hatch or evo) is worse then the delay protoss takes. Most setups protoss can easily respond by going the cannon a bit quicker and just clearing the evo or hatch with that while teching quicker in the meantime as well. The quicker tech lets toss pressure faster afterwards and zerg will have a slower eco as well. This is mostly because I feel drone scouting is a waste of money though, there are no cheeses toss can surprise zerg with and the early lings you need anyway can scout what toss is doing just fine. A drone scout never really finds information that is critical which a ling scout later on wouldn't in my opinion making such a drone scout a pitiful waste of money. If you feel however you do need a drone scout then the entire thing becomes a lot better to do and is generally a good move if you get pylon blocked. It's also quite good in situations where protoss chooses to wall quite far away from the wall and thus they can't cannon the hatch/evo effectively.
Going for the queen is not really a good move imo, it can delay immensely if the protoss screws up but that's depending on poor play from the opponent. Protoss can simply use 4-5 probes in a half circle around the spot where the queen will spawn (bottom left of the hatch) along with zealot/cannon backup to instantly surround the queen and kill it before it can do anything.
Overall any proxy hatch shenanigans (in base hatches, hatch blocks etc) are not really effective if protoss responds well as they hurt zerg more than P. Throwing them in occasionally in pro play can still be good though to not be predictable and surprise players not familiar with it. I suspect a lot of the Kespa players are not really familiar with all the strange tactics yet so they will be trying them out and facing more before they catch up with the rest of the pro scene.
I think if you Pylon block the Zerg's natural and third bases, then it kind of ends up evening out. As Protoss all you have to do is place your other buildings first, and essentially do a Gateway expand with a Forge. I've had this done to me, and more often than not, it ends up putting the Zerg at a disadvantage.
On July 23 2012 02:25 Animostas wrote: I think if you Pylon block the Zerg's natural and third bases, then it kind of ends up evening out. As Protoss all you have to do is place your other buildings first, and essentially do a Gateway expand with a Forge. I've had this done to me, and more often than not, it ends up putting the Zerg at a disadvantage.
oh that reminds me, one of the times I tried this I got 3-pylon walled and cannon rushed/contained. I couldn't manage the drone drill thing since I haven't really practiced it and I think I was doing it wrong. Pretty sure I ended up losing.
On July 23 2012 01:34 Markwerf wrote: Overall any proxy hatch shenanigans (in base hatches, hatch blocks etc) are not really effective if protoss responds well as they hurt zerg more than P. Throwing them in occasionally in pro play can still be good though to not be predictable and surprise players not familiar with it. I suspect a lot of the Kespa players are not really familiar with all the strange tactics yet so they will be trying them out and facing more before they catch up with the rest of the pro scene.
lol, I'm sure they practice on ladder, the cheesiest place on Earth...
Overall any proxy hatch shenanigans (in base hatches, hatch blocks etc) are not really effective if protoss responds well as they hurt zerg more than P. Throwing them in occasionally in pro play can still be good though to not be predictable and surprise players not familiar with it. I suspect a lot of the Kespa players are not really familiar with all the strange tactics yet so they will be trying them out and facing more before they catch up with the rest of the pro scene.
There is a world of difference between proxy hatch in the toss main and hatch blocking the natural. The former does nothing to slow FFE econ and means that you're investing in a way to bypass the cannons and wall at the front. Hatch blocking the natural means you're spending money to slow econ, when their build revolves around that fast nexus. Basically, you're doing a super pylon block of their natural.
You keep saying things but you also said you never drone scout, so is it correct to say that you've never tried this? Are you masters or higher weighing in or something? Because you're talking with zero personal experience and zero pro player observations, and I'm inclined to ignore everything you've said, given the inaccuracy of the statement I quoted.
I have not had the chance to use this against a masters toss yet this season. Diamond toss seem to roll over to it, either overproducing cannons (in which case I cancel the hatch, even if it let's them cancel the second cannon) or losing their single cannon off the broodling stuff I talked about before and thus having to do emergency stuff like chrono zealots, pylon wall the ramp, pull probes, etc. Most maps allow for easy scouting to see whether they drop the nexus immediately after creep recedes. If they don't, get 3-4 spines, grab gas, and check the forge. Gas goes to roaches if you see forge moving (+1 zealots), lair if you fear tech to dts or stargate (add gases as needed, of course). Those spines go a long way (remember to rebuild drones), especially since toss seems to have a gas surplus at that point and will often go sentry-heavy on the 4-5gate option. Spines ensure he can't do any forcefield or hug-the-wall shenanigans. I have also found that one tumor helps immensely in repeling the attack because you won't have roach or ling speed in time, so the speed boost can be the difference between stalkers escaping to the pylon and a full surround.
On July 23 2012 00:07 weikor wrote: ive had pretty high sucsess agianst this by just going 4 gate without trying to expand against this. Especially if zerg makes a queen he will have invested a LOT into hatchery & queen & his own hatchery.
Roaches are limited because of the mineral intense early game, if he gets gas, he will delay his natural by a bit more, have less minerals
Zerglings, are also limited because of the 850 mineral investment for natural, hatch and queen (since he usually needs another queen at natural.)
This only works if the zerg takes a blind 3rd hatch. Zerg can keep an overlord over your expo and see that you never take your natural. If you don't take your natural, there's no reason for the zerg to take a 3rd, and so zerg stays on 2-base, gets 1-2 spine crawlers and earlier tech. 2-base zerg can easily handle a delayed 4-gate unless there's some kind of huge micro error. OR, if you somehow manage to deny the 3rd as well as the natural until 4-gate.
All in all, I honestly think it's far easier for protoss to recover than for zerg. Zerg literally has no immediate advantages of hatch-blocking other than "gee, now my opponent can't expand quickly", while protoss, on the other hand, has a faster tech, has the ability to make cannons, and the ability to pretty much endlessly stop you from expanding a lot (at the expense of the protoss expansion). That being said, I'm totally fine with pylon blocking the natural, cannoning the 3rd, and then 4-gating as a response that zerg literally CANNOT stop.
But I prefer to play macro, so I just kill the hatch, block a little longer, and go on with life.
Just wanted to add some more numbers into this thread. This is directed towards killing the hatchery before it completes so that creep from the hatchery doesn't land just using forge tech, no chronoing zealots or going for a +1 4 gate.
Using Liquipedia numbers, a photon cannon does 16 dps, a hatchery builds at 13.5 health per second (just divide max health - starting health by build time), and probes do 3.3 dps to make up free time the hatchery had to build while the cannon completed. Note that buildings start building with 10% of their hitpoints, so that would be 150 for a hatchery.
Forge first FFE openings
To kill a hatchery with just 1 cannon, the cannon must complete before the hatchery reaches ~200 hp. The theory crafting says 250 hp is the golden number, but factoring in the lag it takes for a cannon to complete and start firing as well as the travel time of the cannon shots means you should send out at least 1 probe for numbers higher than 200 hp.
With plenty of testing, it seems the best option is to kill the hatchery before it spawns with 1 cannon and 3-4 probes. The cannon is necessary for an FFE anyways, and the 3-4 probes will cost you about 150 minerals worth of resources. The hatchery canceling + drone costs the zerg 125 resources as well as the delay of giving up 300 minerals that will generally delay their natural, third, pool, or at least drones, at an earlier point in the game when they are more valuable.
To convince you further, if 1 cannon starts firing on a hatchery as soon as it starts building, the cannon will take 70 ingame seconds to kill the hatchery. With the addition of 1 probe (such as the one that would have planted the nexus), the time is cut to ~30 seconds. With 2 probes, ~22 seconds, 3 ~15 seconds, 4 ~10 seconds, which will barely delay your nexus and tech. I also believe SaSe uses this to stop hatchery blocks.
Nexus first FFE openings
If you were doing a nexus first opening and got blocked, one option is to build the forge asap and get 2 cannons... you won't be able to keep the hatch from finishing, since it will take 85 seconds for your cannons to finish (giving them 15 seconds to damage the hatchery before it finishes), but the cannons will be able to kill the hatchery before a queen finishes 50 seconds after the hatchery completes (if the zerg builds immediately), or before zerglings can hatch and survive the cannon fire.
A better option, I think, is to build a forge and gateway quickly, then 1 gas, 1 cannon, and chrono out 2 zealots. You'll probably kill the hatchery before the queen pops, but if you don't, you can still kill the creep tumor before it can burrow. Zerglings can be a problem with this path, since only 1 cannon usually can't kill all the zerglings before they get away and block the nexus further or run up the ramp, but you can position your zealot(s) to kill the zerglings as soon as they pop and force them into the cannon. You can now pressure/scout with the 2 zealots, take your natural, get a cyber core and 2nd gas, and proceed as normal (usually with some kind of +1 gateway pressure).
On July 20 2012 17:31 trbot wrote: You don't need two cannons to kill a hatch before it spawns. If the toss starts a single cannon within the first 15 seconds of you starting the hatch and then starts attacking the hatch with the probe, the hatch will die before it spawns, at a cost in lost mining time of approximately 60 minerals. Even if you skip the probe and let the hatch finish, if you focus fire it with the cannon, it will die before spawning a second larva, and you'll have made gates and units while it was building, anyway (100s for hatch, 93s for gate+chrono'd zealot).
This post is helpful, I just wanted to correct that the cannon must be COMPLETED with the probe attacking by the first 15 seconds of the hatch starting, so in this case you must already have the cannon started 25 seconds before the hatch started. This is obviously unfeasible, so I still think the more viable option is to build 1 cannon within about 10 seconds of the hatch starting and pulling 3-4 probes, which will kill the hatch before it completes and keep your tech from being delayed very far. If you start the cannon within about 4 seconds of the hatch starting and have two probes attacking the hatchery, you can kill it just about when the hatch finishes, before it can spawn creep, but it's obviously safer to pull the 3 or 4 depending on how late you are.
On August 14 2012 17:37 Doomball wrote: Just wanted to add some more numbers into this thread. This is directed towards killing the hatchery before it completes so that creep from the hatchery doesn't land just using forge tech, no chronoing zealots or going for a +1 4 gate. + Show Spoiler +
Using Liquipedia numbers, a photon cannon does 16 dps, a hatchery builds at 13.5 health per second (just divide max health - starting health by build time), and probes do 3.3 dps to make up free time the hatchery had to build while the cannon completed. Note that buildings start building with 10% of their hitpoints, so that would be 150 for a hatchery.
Forge first FFE openings
To kill a hatchery with just 1 cannon, the cannon must complete before the hatchery reaches ~200 hp. The theory crafting says 250 hp is the golden number, but factoring in the lag it takes for a cannon to complete and start firing as well as the travel time of the cannon shots means you should send out at least 1 probe for numbers higher than 200 hp.
With plenty of testing, it seems the best option is to kill the hatchery before it spawns with 1 cannon and 3-4 probes. The cannon is necessary for an FFE anyways, and the 3-4 probes will cost you about 150 minerals worth of resources. The hatchery canceling + drone costs the zerg 125 resources as well as the delay of giving up 300 minerals that will generally delay their natural, third, pool, or at least drones, at an earlier point in the game when they are more valuable.
To convince you further, if 1 cannon starts firing on a hatchery as soon as it starts building, the cannon will take 70 ingame seconds to kill the hatchery. With the addition of 1 probe (such as the one that would have planted the nexus), the time is cut to ~30 seconds. With 2 probes, ~22 seconds, 3 ~15 seconds, 4 ~10 seconds, which will barely delay your nexus and tech. I also believe SaSe uses this to stop hatchery blocks.
Nexus first FFE openings
If you were doing a nexus first opening and got blocked, one option is to build the forge asap and get 2 cannons... you won't be able to keep the hatch from finishing, since it will take 85 seconds for your cannons to finish (giving them 15 seconds to damage the hatchery before it finishes), but the cannons will be able to kill the hatchery before a queen finishes 50 seconds after the hatchery completes (if the zerg builds immediately), or before zerglings can hatch and survive the cannon fire.
A better option, I think, is to build a forge and gateway quickly, then 1 gas, 1 cannon, and chrono out 2 zealots. You'll probably kill the hatchery before the queen pops, but if you don't, you can still kill the creep tumor before it can burrow. Zerglings can be a problem with this path, since only 1 cannon usually can't kill all the zerglings before they get away and block the nexus further or run up the ramp, but you can position your zealot(s) to kill the zerglings as soon as they pop and force them into the cannon. You can now pressure/scout with the 2 zealots, take your natural, get a cyber core and 2nd gas, and proceed as normal (usually with some kind of +1 gateway pressure).
On July 20 2012 17:31 trbot wrote: You don't need two cannons to kill a hatch before it spawns. If the toss starts a single cannon within the first 15 seconds of you starting the hatch and then starts attacking the hatch with the probe, the hatch will die before it spawns, at a cost in lost mining time of approximately 60 minerals. Even if you skip the probe and let the hatch finish, if you focus fire it with the cannon, it will die before spawning a second larva, and you'll have made gates and units while it was building, anyway (100s for hatch, 93s for gate+chrono'd zealot).
This post is helpful, I just wanted to correct that the cannon must be COMPLETED with the probe attacking by the first 15 seconds of the hatch starting, so in this case you must already have the cannon started 25 seconds before the hatch started. This is obviously unfeasible, so I still think the more viable option is to build 1 cannon within about 10 seconds of the hatch starting and pulling 3-4 probes, which will kill the hatch before it completes and keep your tech from being delayed very far. If you start the cannon within about 4 seconds of the hatch starting and have two probes attacking the hatchery, you can kill it just about when the hatch finishes, before it can spawn creep, but it's obviously safer to pull the 3 or 4 depending on how late you are.
Very very helpful research you did there. I wonder if you are taking about 2nd cannon when you just say "cannon." Because this hatch block is usually done or most effective on Antiga Shipyard when wall is at ramp instead of near nexus or Entombed Valley when first cannon placement doesn't reach hatchery. In these cases, it forces Protoss to make 2nd cannon. In other situations where only 1 cannon is required anyways, then hatch block doesn't make sense in the first place.
On August 14 2012 18:21 Orek wrote: Very very helpful research you did there. I wonder if you are taking about 2nd cannon when you just say "cannon." Because this hatch block is usually done or most effective on Antiga Shipyard when wall is at ramp instead of near nexus or Entombed Valley when first cannon placement doesn't reach hatchery. In these cases, it forces Protoss to make 2nd cannon. In other situations where only 1 cannon is required anyways, then hatch block doesn't make sense in the first place.
Technically the times when you need to start a cannon and/or attack with probes wouldn't change, but you are correct in that the hatch block would be much more effective when the wall is not near the nexus. Nevertheless, I still see it on all maps frequently enough to look into, so I figured I would post my results here. I should also mention that it's possible to wall on Antiga at the nexus, though there are problems with that as well, namely your base is more open to early aggression, later runbys, and it can be more cumbersome to defend against muta, since you can't run stalkers freely from your natural up your main base ramp, since there's a pylon and forge blocking.
So when are you supposed to put down the block and when do you send your drone scout? 13 scout, 15pool, 16 hatch blocking nexus if you natural is blocked or something like 10 scout into 14-15 hatch first blocking nexus, then pool? What are the pros/cons of these (or other versions) and should I get lings to go with the broodlings?
I'm looking to do it to mess with their timings without sacrificing too much economy or attention. I tried the first version a few times but they could often make a cannon+zealot to force a cancel. If they already have a forge or gw building, should I just cancel the hatch right away and make an evo chamber instead? Looking for a rough gameplan. Thanks!
1 zealot has too little dps to force a cancel. If he's going 1 gate expand, the hatch block probably could do a reasonable amount of damage. Of course, if his gateway is already done, I probably wouldn't do it, because chronoing zealots at that point could make things dicey. Lings are still pretty efficient against unupgraded zealots, you just don't want 3 zealots standing around your 6 lings when they pop.
I don't know the zerg side of things, can't give input on when zerg scouts or when they take natural with this build other than what the OP said.
On August 14 2012 17:37 Doomball wrote: Just wanted to add some more numbers into this thread. This is directed towards killing the hatchery before it completes so that creep from the hatchery doesn't land just using forge tech, no chronoing zealots or going for a +1 4 gate.
Using Liquipedia numbers, [lots of numbers]
Hatch still has 1 armor while building, right? So that reduces probe dps by 20% (5-1=4 per attack), cannon dps by 5%.
I feel like a zerg who lets the hatch complete will end up playing, for at least a little while, a zvp of two bases against one, which is significantly better than three bases against two. FFE is focused upon getting that fast expansion, and because it leans so heavily on that, I think hatch block disruption is at least worth using, if not advantageous.
There's a lot of theorycraft around that purports to "prove" that the hatch block isn't worth it, but as a player who hatch blocks preferentially to putting down an expo hatch, I really really like the position it puts me in. Pylon blocks have died to my initial lings by the time I go for the expo, and afterwards, toss is either clearly worried about a queen or lings and I can drone, or is not worried about lings so I can just build all lings, including from the proxy hatch, run in alongside the broodlings on hatch death, and overpower the cannons. The threat of that all-in should haunt every FFE player who sees the hatch block, and of course would add some variety to the first 6 minutes of ZvP matches.
EPT
photon cannon
