|
the idea of 2 roach warrens to speed up the upgrade just screams all in. I understand that you're saying that this build isn't meant to be all in, but think about it; if you're investing so much extra just to get a marginal timing advantage, you really *should* be all in. All ins or heavy timing attacks are the only thing where huge added investment for little return is useful, because they can tip the balance. However, this 11 minute attack doesn't rely on any strong timing (except perhaps before colossus). By 11 minutes protoss can very easily have an observer and an immortal, along with several sentries and a handful of stalkers. I agree you can kill several sentries in this attack but unless you find a strong way to transition or to maintain tempo, I find it hard to rationalize the second roach warren. Most notably, the 150 minerals on the second roach warren would likely go towards your 3rd expo which seems to be your next step in this build.
Another thing is that I don't think this synergizes well with the ice fisher. Ice fisher gives you a big mineral economy, and relatively late tech. Again, whatever few extra roaches you might have because of this bigger economy is sacrificed in the late tech, and second roach warren to compensate.
One more thing is that you need to decide (for yourself and perhaps include it in your post) what you're going to do after this attack. Given that you can't reliably do critical damage every game with it, what's the next step? you'll probably be accumulating gas at this point, with heavily upgrades roaches that give you decent map control but eat up your supply, and once protoss stabilizes, won't give you much potential for aggression. What then?
Overall, yes it's a strategy where all the upgrades combined make for an attack that more or less makes sense, and sniping sentries around the midgame ZvP is commonly understood as a good idea. However, for a build to be good you really want every single thing in your gameplay to rationalize and call for everything else, and as it stands your arguments just don't support each other as much as I would like.
|
A tunneling claws + burrow timing hits as early as 9:30, potentially a minute and a half earlier than this push hits, so unfortunately I do not think the mineral/gas investment is not worth it in this case.
|
On April 20 2011 04:09 ShinyGerbil wrote:
the idea of 2 roach warrens to speed up the upgrade just screams all in. I understand that you're saying that this build isn't meant to be all in, but think about it; if you're investing so much extra just to get a marginal timing advantage, you really *should* be all in. All ins or heavy timing attacks are the only thing where huge added investment for little return is useful, because they can tip the balance. However, this 11 minute attack doesn't rely on any strong timing (except perhaps before colossus). By 11 minutes protoss can very easily have an observer and an immortal, along with several sentries and a handful of stalkers. I agree you can kill several sentries in this attack but unless you find a strong way to transition or to maintain tempo, I find it hard to rationalize the second roach warren. Most notably, the 150 minerals on the second roach warren would likely go towards your 3rd expo which seems to be your next step in this build.
Another thing is that I don't think this synergizes well with the ice fisher. Ice fisher gives you a big mineral economy, and relatively late tech. Again, whatever few extra roaches you might have because of this bigger economy is sacrificed in the late tech, and second roach warren to compensate.
One more thing is that you need to decide (for yourself and perhaps include it in your post) what you're going to do after this attack. Given that you can't reliably do critical damage every game with it, what's the next step? you'll probably be accumulating gas at this point, with heavily upgrades roaches that give you decent map control but eat up your supply, and once protoss stabilizes, won't give you much potential for aggression. What then?
Overall, yes it's a strategy where all the upgrades combined make for an attack that more or less makes sense, and sniping sentries around the midgame ZvP is commonly understood as a good idea. However, for a build to be good you really want every single thing in your gameplay to rationalize and call for everything else, and as it stands your arguments just don't support each other as much as I would like.
The idea is to lock the Protoss in his base while you expand, as should every Zerg attack work. The excess gas can be used in something else like +1 carapace or even overlord speed - it doesn't matter.
It is NOT a "huge" investment at all. Considering the drone, it's only 200 minerals for a roach warren, getting a much faster tech on a build which has "tech" as one of its downs(the Ice Fisher, I mean).
The Ice Fisher also gives you excess minerals after your first few roaches - assuming you are still getting extra tech, like OL speed or +2 attack/+1 armor, or even an Overseer - and those excess minerals perfectly line up with the fact that you want to grab a third base while pressuring your opponent, instead of allowing him to freely tech.
Also, regarding "by that time he has an obs and an immortal" - I've tested this build with YABOT to grab information about it, and despite barely missing an overlord(which had no impact in the build besides slowing down the third queen), at exactly 10:40 I had 23 Roaches and 8 more in production, with all three upgrades.
Notice that my build was not perfect: I didn't take my lair exactly at 100 gas and I had a few 5-second supply blocks, so you'd expect this build to have even more units when better executed.
I seriously doubt an obs and an immortal(plus the oh-so useless sentries) would be enough to hold about thirty roaches by that time.
Gonna update the OP with more information btw, thanks for pointing that out.
|
On April 20 2011 04:09 ShinyGerbil wrote:
the idea of 2 roach warrens to speed up the upgrade just screams all in.
so does the idea of 2 evos scream all in too?
|
I like the idea of this because roach warrens are so cheap. However you would definately need to transition out of this into something else. It is way too easy to max on roaches and if the protoss does hold and you don't transition he could end up with an advantage. I think aggressively trading roaches with the toss is essential, especially for gas heavy units as you mention. If you just try to contain so you can expand and you don't actually do damage to the toss, the build may suffer somewhat. I am definately a fan of reducing sentry count and energy early though with a push like this.
|
OP updated with important information, like numbers and the problem of excess gas - which I didn't notice until I tried to actually polish the build and lower down the time it is ready. Not a big deal tho.
|
On April 20 2011 06:16 SaviorSelf wrote:Show nested quote +On April 20 2011 04:09 ShinyGerbil wrote:
the idea of 2 roach warrens to speed up the upgrade just screams all in. so does the idea of 2 evos scream all in too?
The second evo is only useless once you max out upgrades. The second warren is too, but obviously you max out roach upgrades much faster.
The whole "all-in/not all-in" dichotomy is a false one in general -- many builds sacrifice future strength for present strength, some more extreme than others, with the idea that the present strength can be used to force an equal or greater longterm disadvantage on the opponent. The builds that people call "all in" are just on one end of that continuum.
Building an extra roach warren is unequivocally this. It's sacrificing 150 minerals and a drone to get roach speed faster. This is sacrificing future strength for short-term power, in the hopes that by having roach speed for your first push you'll get more than 200 minerals of utility out of it.
What it *does* scream is "I'm going to hit you in a very particular timing window with speed tunneling roaches, and I am going to hit you with a lot of them, because I'm expecting to get my investment back." It's a scouting telegraph, certainly.
So, don't build your two warrens next to each other, unless you're using them to simcity and don't mind being obvious.
|
On April 20 2011 06:49 entropius wrote:Show nested quote +On April 20 2011 06:16 SaviorSelf wrote:On April 20 2011 04:09 ShinyGerbil wrote:
the idea of 2 roach warrens to speed up the upgrade just screams all in. so does the idea of 2 evos scream all in too? The second evo is only useless once you max out upgrades. The second warren is too, but obviously you max out roach upgrades much faster. The whole "all-in/not all-in" dichotomy is a false one in general -- many builds sacrifice future strength for present strength, some more extreme than others, with the idea that the present strength can be used to force an equal or greater longterm disadvantage on the opponent. The builds that people call "all in" are just on one end of that continuum. Building an extra roach warren is unequivocally this. It's sacrificing 150 minerals and a drone to get roach speed faster. This is sacrificing future strength for short-term power, in the hopes that by having roach speed for your first push you'll get more than 200 minerals of utility out of it. What it *does* scream is "I'm going to hit you in a very particular timing window with speed tunneling roaches, and I am going to hit you with a lot of them, because I'm expecting to get my investment back." It's a scouting telegraph, certainly. So, don't build your two warrens next to each other, unless you're using them to simcity and don't mind being obvious.
Well, regarding this...
200 minerals is not a big deal. At all. Maybe it would be a good idea to try and hide it(heck, you have 4 queens, spread the creep everywhere), but it isn't the end of the world if its scouted. It is still very strong.
|
United States15275 Posts
On April 20 2011 06:54 Zephirdd wrote:Show nested quote +On April 20 2011 06:49 entropius wrote:On April 20 2011 06:16 SaviorSelf wrote:On April 20 2011 04:09 ShinyGerbil wrote:
the idea of 2 roach warrens to speed up the upgrade just screams all in. so does the idea of 2 evos scream all in too? The second evo is only useless once you max out upgrades. The second warren is too, but obviously you max out roach upgrades much faster. The whole "all-in/not all-in" dichotomy is a false one in general -- many builds sacrifice future strength for present strength, some more extreme than others, with the idea that the present strength can be used to force an equal or greater longterm disadvantage on the opponent. The builds that people call "all in" are just on one end of that continuum. Building an extra roach warren is unequivocally this. It's sacrificing 150 minerals and a drone to get roach speed faster. This is sacrificing future strength for short-term power, in the hopes that by having roach speed for your first push you'll get more than 200 minerals of utility out of it. What it *does* scream is "I'm going to hit you in a very particular timing window with speed tunneling roaches, and I am going to hit you with a lot of them, because I'm expecting to get my investment back." It's a scouting telegraph, certainly. So, don't build your two warrens next to each other, unless you're using them to simcity and don't mind being obvious. Well, regarding this... 200 minerals is not a big deal. At all. Maybe it would be a good idea to try and hide it(heck, you have 4 queens, spread the creep everywhere), but it isn't the end of the world if its scouted. It is still very strong.
But it still would be a good idea to build the warrens far apart.
And have you thought about skipping ling speed (and giving up 2 roaches) to get +1 carapace?
|
On April 20 2011 07:44 CosmicSpiral wrote:Show nested quote +On April 20 2011 06:54 Zephirdd wrote:On April 20 2011 06:49 entropius wrote:On April 20 2011 06:16 SaviorSelf wrote:On April 20 2011 04:09 ShinyGerbil wrote:
the idea of 2 roach warrens to speed up the upgrade just screams all in. so does the idea of 2 evos scream all in too? The second evo is only useless once you max out upgrades. The second warren is too, but obviously you max out roach upgrades much faster. The whole "all-in/not all-in" dichotomy is a false one in general -- many builds sacrifice future strength for present strength, some more extreme than others, with the idea that the present strength can be used to force an equal or greater longterm disadvantage on the opponent. The builds that people call "all in" are just on one end of that continuum. Building an extra roach warren is unequivocally this. It's sacrificing 150 minerals and a drone to get roach speed faster. This is sacrificing future strength for short-term power, in the hopes that by having roach speed for your first push you'll get more than 200 minerals of utility out of it. What it *does* scream is "I'm going to hit you in a very particular timing window with speed tunneling roaches, and I am going to hit you with a lot of them, because I'm expecting to get my investment back." It's a scouting telegraph, certainly. So, don't build your two warrens next to each other, unless you're using them to simcity and don't mind being obvious. Well, regarding this... 200 minerals is not a big deal. At all. Maybe it would be a good idea to try and hide it(heck, you have 4 queens, spread the creep everywhere), but it isn't the end of the world if its scouted. It is still very strong. But it still would be a good idea to build the warrens far apart. And have you thought about skipping ling speed (and giving up 2 roaches) to get +1 carapace?
Someone suggested that already, but I don't think skipping the ling speed would be useful. First because you never know what you may need in the game, and nothing says "FML" more than getting into the 25min mark and needing lings to harass somewhere just to see that you never got ling speed. Also, +1 armor won't make that much difference except to sentries' DPS, especially if the protoss is getting immortals.
Basically, its a personal choice, I just don't think the usefulness of ling speed can be skipped for a +1 armor here, even though I basically don't get lings. It's a good idea and probably would make the push even stronger however.
Also, I don't remember exactly, but I *think* that would make the timing off by about 5 seconds, no big deal.
|
The problem with this is that any decent Protoss gets Hallucination as he drops his Nexus and continually scouts. When he sees 2 Roach Warrens and a bunch of Roaches being made, he'll know that some attack is coming. It's standard to get at least 1-2 Cannons at your natural when doing the 3 Gate Sentry expand, and the safest course of action is to drop a Robo after you expand. This means that the Protoss will have plenty of detection and units to fend this attack off.
This will pretty much destroy a standard 6 Gate push, but it won't really be worth it vs. safer play or even something like a 5-6 Gate Blink timing. You're putting enough into these Roaches that your tech is far enough behind the Protoss that it doesn't really matter that much if you delay his third for a while.
|
I've started using a similar build (but probably more standard) in ZvP when it appears that they are 3 gating. Trouble is, it's extremely hard to tell, and the build completely relies on no warp gate rushes.
My version of the build rushes roach speed or burrow movement though, not both. It's also faster I think (maybe around 8.5 to10 minute mark) because I get gas quite fast for the lair, get less roaches, and don't get ranged attack upgrade.
Once I accidently built two warrens, and with it I did a double upgrade and won, I tried doing it again, but it seems to be spending too much money and comming too late for it to be as useful.
On April 19 2011 11:29 Zephirdd wrote:
This is how it works: At about 10:25, All three upgrades will be done and you will have 21 roaches completed, and 9 in production - perhaps more, because I just can't do this build correctly, since I keep missing OL timings -.- I think you mean 4 upgrades, because of ranged attack, right?
10:25 sounds a bit fast, but I can't say it's impossible. I open with 11 OP (for around 18 hatch) which is a bit worse economy, but then again I think it's really useful to completely counter any sort of cheese or hatch blocking. How are you going to hatch first when they probe/pylon block the natural?
|
On April 20 2011 08:49 gdalam wrote:
The problem with this is that any decent Protoss gets Hallucination as he drops his Nexus and continually scouts. When he sees 2 Roach Warrens and a bunch of Roaches being made, he'll know that some attack is coming. It's standard to get at least 1-2 Cannons at your natural when doing the 3 Gate Sentry expand, and the safest course of action is to drop a Robo after you expand. This means that the Protoss will have plenty of detection and units to fend this attack off.
This will pretty much destroy a standard 6 Gate push, but it won't really be worth it vs. safer play or even something like a 5-6 Gate Blink timing. You're putting enough into these Roaches that your tech is far enough behind the Protoss that it doesn't really matter that much if you delay his third for a while.
Excellent point, 5-6gate blink pushes would be really effective against this kind of timing, although these are speed burrow +1 roaches, so it'd come down to a very micro intensive battle.
The detection you mention isn't a problem however. The highest point of this build isn't to be "sneaky". It's to avoid forcefields and put lots of pressure on your opponent. As much as it crushes 6gate pushes, it should also deal a really good amount of damage - at least enough damage to pay for the second roach warren, which isn't really all that hard to do considering the cost-effectiveness of roaches compared to protoss gateway units.
On April 20 2011 09:35 Xapti wrote:I've started using a similar build (but probably more standard) in ZvP when it appears that they are 3 gating. Trouble is, it's extremely hard to tell, and the build completely relies on no warp gate rushes. My version of the build rushes roach speed or burrow movement though, not both. It's also faster I think (maybe around 8.5 to10 minute mark) because I get gas quite fast for the lair, get less roaches, and don't get ranged attack upgrade. Once I accidently built two warrens, and with it I did a double upgrade and won, I tried doing it again, but it seems to be spending too much money and comming too late for it to be as useful. Show nested quote +On April 19 2011 11:29 Zephirdd wrote:
This is how it works: At about 10:25, All three upgrades will be done and you will have 21 roaches completed, and 9 in production - perhaps more, because I just can't do this build correctly, since I keep missing OL timings -.- I think you mean 4 upgrades, because of ranged attack, right? 10:25 sounds a bit fast, but I can't say it's impossible. I open with 11 OP (for around 18 hatch) which is a bit worse economy, but then again I think it's really useful to completely counter any sort of cheese or hatch blocking. How are you going to hatch first when they probe/pylon block the natural?
Yeah, it's 4 upgrades, thx for pointing that out 
The timing is at 10:25 considering there were no interruptions - no need to waste larva on units, for example - as I've tested on YABOT.
Your version sounds like a less economic version of it, and with less units as well as you mentioned, but hits fairly earlier. Mind posting some replays for comparision?
|
i like it, i think the 4 queens are a great anti air and the Protos probably wont attack into that with void rays, and if need be you can always throw down i hydra den which takes almost no time at all to make.
i also think that this build can be used in mroe matchups but im nto sure yet
|
On April 20 2011 10:08 NDsOdapOp wrote:
i like it, i think the 4 queens are a great anti air and the Protos probably wont attack into that with void rays, and if need be you can always throw down i hydra den which takes almost no time at all to make.
i also think that this build can be used in mroe matchups but im nto sure yet
I don't think so, Roaches hardly have any time to shine on ZvT and it is very hard to "waste" so much gas on ZvZ instead of building more roaches and take the upgrades over time. Also, there isn't much use for Tunneling Claws on these matchups since there isn't a Force Field-like ability to justify getting them so quickly.
|
I generally opt for the spanishiwa build in zvp as well.
I actually stumbled upon this once, by accidentally making 2 roach warrens, as I had one in production, didn't realize it, and then made another, within like 5 game seconds of the other. They then both spawn right as I'm getting air harrassed, and the toss is like lol 2 RWs... you lost, and then my queens hold off the harrass, all the while, I was thinking, he's right, shit...
But so, I'm like, fuck it, let's do this, and tech both of them, all the while 1 armor 1 range was already going (i tweak it slightly such that my evos are ready to upgrade before spanishiwa reccommends them). I then begin to mass roaches meanwhile, and busted in his door right as he was starting to expand, and then flooded in the lings for the gg. I was thinking afterwards that I just was lucky, but Idk, this thread has me thinking:
RWs take forever to build, and they're key to early game zvp, in most cases. However, if you have two, say one near your main and the other at your nat, why not get double upgrades, and know that if one base falls, you can still make roaches, and not have to wait like, another 150 seconds for another RW to build. IDK, may be a slight waste of minerals, but meh, if you use it to wall off your base, and you have an excess of spanishiwa-build minerals (how the hell does anyone keep there minerals low with this? Not even Spanishiwa himself can lol) anyways, why not...
[edit add-in]: And as for the what to do with the excessive gas, do something techy that makes sense, given the circumstances. Infestors are a great way to spend gas, and they complement burrow + tunneling claws roaches quite well. Try working in a nydus. Go spire. Pepper in some hydras. The true fear of the Ice Fisher build is in the early game. If you can survive it, and take your third, you'll be maxed in no time, with possibly even a lead on upgrades and tech. If my games last longer than about 14 minutes, I generally win, unless they're macroing hard too, and in those games its really close, but that's only like 5% of the games beyond the 14 minute point it seems...
|
good slice of cheese, forget the contain and expand its a good all in for a best of series
im sure you can make it work into the normal spanishiwa macro game if you really want tho
|
On April 20 2011 13:46 darkscream wrote:
good slice of cheese, forget the contain and expand its a good all in for a best of series
im sure you can make it work into the normal spanishiwa macro game if you really want tho
You could make it an all-in, but heck, isn't that true with every kind of timing attack?
But cheese? Nah, it isn't. Cheese would be something that sounds retarded like BL rush or proxy hatch, but this isn't that retarded.
In fact, it is nothing but a very solid timing push.
|
No that is absolutely not the definition of cheese. But you are right. It's more of a timing push.
|
This + tweaked gas timings (like put up 3 gas then pull drones off 1 after launching the upgrades for constant roach production off 2 base) and actually staying on 2 bases might be a powerfull all-in against a forge FE from protoss for example, on smaller maps or positions.
But I don't see it working if you wish to transition from it.
Edit : You might be able to squeeze in a +1 range upgrade since it is all-in anyways, but +1/+1 like some people said seems a bit overblown.
Edit 2 : After some more thinking, if you use it as an allin this would work better imo with a 14g 14p or 14p 16h, you could cut the third queen used for creep spread instead to pay for the second roach warren. I'll try to find timings I like^^
|
|
|
|
|
|
EPT
