|
On April 09 2011 08:23 effdee wrote: Wouldn't factoring in queens in the ramp scenario change things a bit? Sure, but how much does a queen cost? Are you going to make that queen every game just in case I want to run up the ramp? Running up the ramp is just a single of the many options one can do.
|
On April 09 2011 06:09 SC2-Dethklok wrote: Ive seen him doing 13 hatch 15 pool in his stream lately. As far as I know 13 hatch 15 pool is supposed to be the most economical opening zerg can do. Also the hatch usually goes down before and scouts get to your base so the chances for a block to happen are minimal.
It's not. 14h15p and 15h14p are the most economic openings. A *lot* of testing has gone into demonstrating that.
regulator_mk wrote: 16 hatch 15 pool gets the pool down sooner than 14 hatch 15 pool since you have the drone that makes the hatch mining longer before he makes the hatch.
The faster pool doesn't matter if the hatch is slower, and that's why 16h15p falls behind.
He already answered this in the OP. In the second and third paragraphs he says he now does 13H/15P because it's more economic, and that he no longer does 16H/15P.
Sorry, I missed that somehow. But as a major contributor to that thread, I can say with some amount of authority that that the results changed significantly once a small bit of reality was injected into the builds (gas, spinecrawlers, etc), and that wasn't ever reflected in the OP, since he was only doing drone races. Which 13h15p was the best at, but which isn't particularly relevant, unless you're going to build up to 2 queens and 48 drones without doing anything else at all.
Here's a link to the post I made with conclusions about hatch-first builds as well as graphs allowing you to draw your own conclusions:
http://www.teamliquid.net/forum/viewmessage.php?topic_id=174374¤tpage=19#374
|
On April 09 2011 08:33 Skrag wrote:
It's not. 14h15p and 15h14p are the most economic openings. A *lot* of testing has gone into demonstrating that.
Do you have any links to a thread with that info in it? Last I checked this thread http://www.teamliquid.net/forum/viewmessage.php?topic_id=174374 13 hatch 15 pool is better. I know the thread is older so if you have a link to a newer thread that proves otherwise a link would be greatly appreciated. :D
Edit: Never Mind just saw the link at the end of your post
|
On April 09 2011 08:33 CecilSunkure wrote:Show nested quote +On April 09 2011 08:23 effdee wrote: Wouldn't factoring in queens in the ramp scenario change things a bit? Sure, but how much does a queen cost? Are you going to make that queen every game just in case I want to run up the ramp? Running up the ramp is just a single of the many options one can do.
This build requires 4 queens as part of its core, with more queens depending on the amount of aggression exhibited by the opponent, it is shown in the game vs Cruncher that Spanishiwa did not think twice about making extra queens once he confirmed it was a 4gate
My bigger concern is that a Queen takes more damage from stalkers than slow lings. Wouldn't it be better to have 8 lings sitting ontop of the ramp than a queen?
|
On April 09 2011 08:48 VictorX wrote:Show nested quote +On April 09 2011 08:33 CecilSunkure wrote:On April 09 2011 08:23 effdee wrote: Wouldn't factoring in queens in the ramp scenario change things a bit? Sure, but how much does a queen cost? Are you going to make that queen every game just in case I want to run up the ramp? Running up the ramp is just a single of the many options one can do. This build requires 4 queens as part of its core, with more queens depending on the amount of aggression exhibited by the opponent, it is shown in the game vs Cruncher that Spanishiwa did not think twice about making extra queens once he confirmed it was a 4gate My bigger concern is that a Queen takes more damage from stalkers than slow lings. Wouldn't it be better to have 8 lings sitting ontop of the ramp than a queen?
It doesn't. Queens aren't armored or light.
|
On April 09 2011 08:46 SC2-Dethklok wrote: Edit: Never Mind just saw the link at the end of your post
Same thread, it was just a really large one.
The graphs in the post I linked are from my AI testing (which was far more consistent and reproducible), using a build that tried to mimic timings from an Idra hatch-first replay as closely as possibly, building gas, a spine crawler, and a creep tumor.
The results in the original post are based purely on a drone race to 48 drones. If I recall correctly, 13h15p ended up falling behind economically in my testing because it was short resources to take full advantage of its larva advantage once you tried to do anything other than build drones, but that was a really really long time ago, so I might be misremembering.
|
I think this build isn't too good in more ZvZ scenarios than spanishiwa mentions. Concerning the replays, FXOtgun killed your hatch and you outplayed him in the mid game. So it's true you can barely survive >10pool-speed, but this build isn't advantageous. Against coLFireZerg you lucked out that he decided to kill your spine crawler and didn't walk up to your main to wreak havoc.
I think this is a very interesting build and it's awesome ZvP/ZvT, but heavy sling early (gas taken anywhere between 10-14 supply) aggression to expand can gain map control and out expand you, maybe even break your door.
I'd really appreciate it if you posted a few ZvZ losses you had when using this build.
--yuv
|
On April 09 2011 08:33 Skrag wrote:[...]and that wasn't ever reflected in the OP, since he was only doing drone races. Which 13h15p was the best at, but which isn't particularly relevant, unless you're going to build up to 2 queens and 48 drones without doing anything else at all.[...] Since I followed both of the previous Zerg Economy threads closely, one of the things that most interested me about this build was that it is about as close to a pure 'drone race' as you'll see in a 'real' strategy. The extra Queens (#3 and #4) aren't using larvae, the spine crawlers are few, and the gas is taken very late. It is almost purely a drone race to 36 drones before the gas is taken.
I know it isn't a perfect drone race, and so the results will vary from all of the modelling and simulation in those previous threads, but this struck me as smooth evolution from the idea of 'what is the fastest way to get to full mineral saturation on 2 bases while staying alive'. The 'while staying alive' part was modeled (IIRC) in the previous work by taking drones as specific timings for gas / spines based on a build used commonly by Idra at the time. Here it is 'Queens and spine crawlers' rather than gas, ling speed and spines.
In any event - with how close this build is to a drone race, I wonder if 13h15p is going to be the best outcome, or if the slight variations (extra spines/queens) will make another variation better. I suspect that it won't be a great difference, and Spanishiwa noted that the 13h is harder to block, a very significant advantage considering he does not fallback to a speedling expand (as many players will do against a blocked hatch).
|
On April 09 2011 07:28 mizak wrote:Show nested quote +On April 09 2011 07:04 Kogut wrote:I'd tune into a stream where the high-end naysayers of this build faced off against Spanish, even if they knew it was coming. I was hanging out on the stream last night seeing the ladder randoms try and fend it off, but I'd love to see the pros show us just how terrible this build really is. + Show Spoiler +Sure, you didn't say it's terrible, but you may as well have. They're just expressing their opinions, which have proven to be reliable hence the blue highlighting. I'm sure spanishwa is fine with discussion from high level players. I like their input, it gives more information on the build.
+1 for not actually reading any of my other posts in this thread praising DF and others for their insight and willingness to have a discussion about a 'non-standard' build.
|
On April 09 2011 07:51 CecilSunkure wrote:Show nested quote +On April 09 2011 07:42 imareaver3 wrote:On April 09 2011 06:50 CecilSunkure wrote: I read through this topic and decided that this build isn't going to work against a decent Protoss.
Whenever I scout hatch first I immediately go over and look for their vespene timing. If I see they are delaying vespene for too long (speedlings will finish at 6 minutes or later) I will probably opt for a double stalker opening. The thing about the double stalker opening is I can follow my initial two stalkers up with two more, and then expand, or expand behind the first two. If I see such a late gas as Spanishawa's I'm going to have four very fast stalkers at the zerg's natural.
A zerg player in this situation would need at least 2 spine crawlers to defend, and even then I can run up the ramp with my stalkers. You're going to have to make a ton of lings which will never be able to kill a stalker, and I get up a pretty fast expansion. All those lings, spines, and queens you made is going to result in even worker counts AT LEAST. Not to mention you'll have late tech and no map control.
In the replay I saw of kcdc he was playing very risky with his forward aggression so early, so I think he lost because of bad play. If you watch the probe to drone comparisons he forced you to only have a 3-5 drone lead throughout the early to mid game, and you got ahead due to decisions that aren't caused by your build or style of play.
Imagine if instead of pressuring you with a zealot/stalker there were two very fast stalkers followed by two more. And once seeing you sink so many minerals into spines/lines/queens I know that you can't be offensive with anything but a lot of ling/banelings meaning I'm free to do nearly anything I want.
Looks to me like this opening against Protoss will only work if the Protoss doesn't know how to react. You can't run up the ramp because of a drone block--the Z can see you come within crawler range and just hold ~6 drones on the ramp. Two crawlers will kill a stalker before they can run up to the ramp, see the block, and run back, so you can't exploit the loss of mining time. At least, that's what I've seen Spanishiwa do against early ramp aggression, and I've found that it works for me. After all, if the stalkers can't run up the ramp, they can't do much at all, and chronoing out 4 fast stalkers does hurt the P. Drones blocking the ramp is risky and dumb. What if I don't decide to move up the ramp? What if I just peak up the ramp and see your drones there? What if I got a zealot across the map to tank the spine and just snipe it off with the Stalkers? Yes, a single spine crawler. You saying there will be two crawlers just shows you don't know what you're talking about; you need two crawlers to cover your natural against stalkers so you don't get anything picked off. You need two to cover the area, they won't both be in range of the ramp (on most maps). Plus if you had two crawlers ready to go by the time I have two stalkers at the natural you're behind already due to a sunk 300 minerals; I might as well make another nexus somewhere. The point is, is that in a realistic game you can't know what I'm going to do and prepare drones on a ramp to perfectly counter a single thing out of the many options a protoss player has with early stalkers. The thing is without speedlings you have to make both spines and lings that will put you at at least even worker counts, and you have no ability to be aggressive and your tech is delayed. The point of 2 or 4 Stalkers isn't to kill you, it's to force you to not make drones.
Zerg can easily adjust to you doing early pressure. If their drone scouts that you make your second gateway on 21/22, which is necessary to do a fast 2 stalker rush, they can just get their gas earlier, or they can make 2 spine crawlers instead of one. Especially on maps with smaller chokes, the 2-4 stalker pressure is really bad (e.g., it's horrible on shakuras).
If you don't throw down your second gate at an unusually early timing, you won't really do any damage, and once his speed finishes and he has the 12-16 lings you forced, he can do 2 or 3 straight rounds of drones, getting him at least back to even.
I think that the only way you can do the early stalker pressure against 15 hatch successfully is to do the pressure, fall back to expand, then cancel the nexus and 4 gate them. Unless you're playing a super aggressive zerg they're almost guaranteed to be droning really hard upon seeing your nexus starting after you forced them to make so many slow lings.
|
I just got bc rushed while doing this build. ;D Still held it off just because I was able to put down a 4th and 5th
|
On April 09 2011 13:51 PJA wrote:
Zerg can easily adjust to you doing early pressure. If their drone scouts that you make your second gateway on 21/22, which is necessary to do a fast 2 stalker rush, they can just get their gas earlier, or they can make 2 spine crawlers instead of one. Especially on maps with smaller chokes, the 2-4 stalker pressure is really bad (e.g., it's horrible on shakuras).
If you don't throw down your second gate at an unusually early timing, you won't really do any damage, and once his speed finishes and he has the 12-16 lings you forced, he can do 2 or 3 straight rounds of drones, getting him at least back to even.
I think that the only way you can do the early stalker pressure against 15 hatch successfully is to do the pressure, fall back to expand, then cancel the nexus and 4 gate them. Unless you're playing a super aggressive zerg they're almost guaranteed to be droning really hard upon seeing your nexus starting after you forced them to make so many slow lings.
An actual 2 stalker rush needs a 2nd gate before core/2nd pylon so its even more easily scouted than you think. Not to mention 2 stalker builds aren't exactly economical with probe cuts, no sentires, and chronos spent of stalkers.
On April 09 2011 08:33 CecilSunkure wrote:Show nested quote +On April 09 2011 08:23 effdee wrote: Wouldn't factoring in queens in the ramp scenario change things a bit? Sure, but how much does a queen cost? Are you going to make that queen every game just in case I want to run up the ramp? Running up the ramp is just a single of the many options one can do.
I don't think running up the ramp would work if zerg micro's their lings too--the lings do catch up to stalkers when stalkers pause to shoot, meaning that the stalkers can be cornered with relative ease (the hatch creep + first tumor connecting main and nat cut down on your kiting area a lot). Not to mention you'll have 4 queens who, although really slow, will cut down on the area of non-creeped space you have to kite lings on. I tested it only microing the stalkers against 4 queens (coming from 1 side) and 16 lings on a move in a 1 screen space to mimic how much non creep space you have (like 1.5-2x what xnc provides to give the stalkers an advantage), and my stalkers all died.
1 round of lings (say 16-20) to kill off 4 stalkers doesn't really put zerg behind because they can make 2 rounds of drones right after they kill the stalkers, as toss won't have the FF numbers that they really need to push for a while.
As for harass options, if zerg saves 2 transfuses (ex hatch first, 2x tumor, 3rd and 4th queen right away, 2 injects and save the original 2's energies), I don't think your 4 stalkers are going to be taking down a spine crawler before another 2 transfuses are ready if zerg chases stalkers away once in a while with his queens (really hurt one can be swapped with main hatch one to preserve a transfuse).
That being said, I personally think no gas till 40 drones is a bit extreme and is abusable by a 3 gate 1 gas expand into pressure (prevent creep spread and force lings all the same as a zealot heavy army will destroy a spine even with multiple transfuses/ff can isolate a queen to beat on). Its what I did when I beat spanishwa, but that may have been while he was refining the build (this was like the day he was going on mr. bitter's stream). The pressure let me stay even with him until like 45 workers iirc (which is actually usually the norm for my pvz as I play very aggressive).
|
On April 07 2011 19:25 Fuzy wrote: Since alot of ppl are advising newer players to focus on just 1 build, and since there is a lot of discussion going on about this build, do you people think that it is an advisable build for someone trying to lern zerg macro, and getting out of bronze?
I kinda like the idea of using 1 build for all 3 matchups so i get better at doing that one. And in 99% of my games ppl put up early pressure so i rly started liking this particular BO. No, this build requires very good mechanics and good micro in order to hold off cheese, also, it requires a good amount of game sense and judgement calls later in the games.
|
On April 09 2011 11:45 Malthius wrote: I know it isn't a perfect drone race, and so the results will vary from all of the modelling and simulation in those previous threads, but this struck me as smooth evolution from the idea of 'what is the fastest way to get to full mineral saturation on 2 bases while staying alive'. The 'while staying alive' part was modeled (IIRC) in the previous work by taking drones as specific timings for gas / spines based on a build used commonly by Idra at the time. Here it is 'Queens and spine crawlers' rather than gas, ling speed and spines.
In any event - with how close this build is to a drone race, I wonder if 13h15p is going to be the best outcome, or if the slight variations (extra spines/queens) will make another variation better. I suspect that it won't be a great difference, and Spanishiwa noted that the 13h is harder to block, a very significant advantage considering he does not fallback to a speedling expand (as many players will do against a blocked hatch).
The ability to block is definitely something worth considering. It's just necessary to recognize that you're making some sacrifice for it. And the claim should be "using one of the best economic builds for this reason" rather than "using *the best* economic build". (in this case the sacrifice is pretty small, and I would say very likely worthwhile, but it's definitely there)
And the reason builds tended to fall behind when doing more than a pure drone race is because there would be points where they had to skip or delay a drone, either because there weren't enough resources to use the larva as soon as it appeared, or because the larva had to be used for something else. And keep in mind that even in the drone race showing 13h15p ahead, it was *barely* ahead. 20 minerals at 6 minutes. I strongly suspect that 2 extra queens will be more than enough to put those same kinks in.
|
Drones blocking the ramp is risky and dumb. What if I don't decide to move up the ramp? What if I just peak up the ramp and see your drones there? What if I got a zealot across the map to tank the spine and just snipe it off with the Stalkers
Sure, but how much does a queen cost? Are you going to make that queen every game just in case I want to run up the ramp? Running up the ramp is just a single of the many options one can do.
You write like you have totaly NO IDEA about this build. If ure gong with 1 stalker1zealot than Z can either focus fire a stalker or just take damaged drones back. U have to focus fire a drone cuz of AI, and you need to actualy shot them 4 times (or 5 due the hp regen) to get the kill, while spine does its dps. You would need damn great luck or play against no-reasonable player to get a drone kill like that.
And for god sake Queens are not only there just to block the ramp. Basicaly now I do 2queens per base in almost any game, just becouse of nice creap spread and nice insurance against early aerial attacks. And oh damn, transfuse. I think every1 of us saw how great lifesaver it can be.
And all this is not only speculation. I think spanish proved it in quite alot of games. Especialy by abusing drone AI.
|
On April 09 2011 03:04 JediGamer wrote:Show nested quote +On April 08 2011 23:53 tgun wrote: JediGamer: That would only really beat Zergs who aren't very familiar with the build, since they can drop 1-2 spines at their natural, be safe to hellion runby's and destroy your drop with 4 queens + slow lings.
Spanishiwa's build is amazingly strong, as most terran/toss haven't figured out a way to adapt to it. Zergs are ceeding a small amount of mapcontrol early on (fast slings) in exchange for a faster, stronger economy, allowing for a better outlook entering the midgame. Seeing as how the only Blue highlighted players here are saying it won't work vs consistent opponents. Anyhow I won't bother saying its not an epic build because there will be about 50 midmasters-bronze posts that will shout 'NO!' at me. How many people here actually looked at CrunChers macro during the game. His first warp in of units he had 450 gas. Warping in zealots. It was an experimental build that failed.
Yeah, you can pretty much disregard what I'm saying without actually providing a factual retort, but the fact stands that the way his build is constructed leads you into a much stronger midgame, ceeding early-game mapcontrol and having a very tedious opening. People like you are the reason that we move so slowly forward; you're too careful to try new things.
This has pretty much become my standard ZvT opener, as it counters most terran openers (2rax bunker pressure / 1rax reaper into hellions) and allows me to get my creepspread moving early, which is one of the most underrated things in the zerg aresnal right now.
However, I do agree that in ZvP, it isn't the greatest opener as it almost immediately loses to 2gate proxy and cannon rushes. I'll stick to 14gas 14pool expand until I can find a way to get an earlier hatchery down that doesn't get hurt by early pressure as badly as 14pool 16hatch does.
EDIT:: Only basing your opinions off 'blue posts' is horribly narrowminded.
|
On April 09 2011 18:35 tgun wrote:Show nested quote +On April 09 2011 03:04 JediGamer wrote:On April 08 2011 23:53 tgun wrote: JediGamer: That would only really beat Zergs who aren't very familiar with the build, since they can drop 1-2 spines at their natural, be safe to hellion runby's and destroy your drop with 4 queens + slow lings.
Spanishiwa's build is amazingly strong, as most terran/toss haven't figured out a way to adapt to it. Zergs are ceeding a small amount of mapcontrol early on (fast slings) in exchange for a faster, stronger economy, allowing for a better outlook entering the midgame. Seeing as how the only Blue highlighted players here are saying it won't work vs consistent opponents. Anyhow I won't bother saying its not an epic build because there will be about 50 midmasters-bronze posts that will shout 'NO!' at me. How many people here actually looked at CrunChers macro during the game. His first warp in of units he had 450 gas. Warping in zealots. It was an experimental build that failed. Yeah, you can pretty much disregard what I'm saying without actually providing a factual retort, but the fact stands that the way his build is constructed leads you into a much stronger midgame, ceeding early-game mapcontrol and having a very tedious opening. People like you are the reason that we move so slowly forward; you're too careful to try new things. [...]
Sure u get a decent midgame up with that econ ... but you need to GET to midgame which is considered one of the biggest problems of Z. We are not trying to slow down the process we even make it faster by pointing out obvious counters to this build. It is more that most of the people here posting have played the build only 30-50 times (yes only) and it is not that well known on ladder (for P/T at least, saw several Zergs doin it now). It is a good thing by spani but imo this build is just pushing over the top. Try to get a good timing for your gas instead of blindly playing this build because it works on ladder RIGHT NOW
edit: imo on my level of play(dia+) most T and P dont really get the whole thing about mapcontrol yet. but this will change drastically with such builds for zerg.
|
On April 09 2011 08:33 CecilSunkure wrote:Show nested quote +On April 09 2011 08:23 effdee wrote: Wouldn't factoring in queens in the ramp scenario change things a bit? Sure, but how much does a queen cost? Are you going to make that queen every game just in case I want to run up the ramp? Running up the ramp is just a single of the many options one can do.
With all the respect due to the blue highlighting, 4 queens for 2 bases is not particularly costly or uncommon even at a pro level. In fact they are amazing for air defense and creep spread as well as transfuse (which shuts down a lot of early pressure when used on queens or crawlers).
Note as well that what 2 (4?) queens and a couple crawlers can do to stop stalkers is very map dependent, on some maps the crawlers can cover 2 ramps and get in the way of anything trying to get by. Obviously on XNC you absolutely need the queens to block that ramp.
|
This build is stupid and dumb... on the surface
Im having so much fun with it right now. It make me think about all the wasted money and tech i get early on, when really i dont then.
Sometimes for me, my money can get a little on the high side so i get an extra macro hatch. Plus it helps with the queen production. I know the build has a goal of 4 queens but since they are soo good, i end up getting a few more.
Though for me at the moment i am struggling late game with toss, although getting 8-15 infestors has also made that fun now. I just need to keep on top of things.
Havent tried it in zvz yet.
note: i have been doing the 13h 15p version.
|
On April 09 2011 08:33 Skrag wrote:Show nested quote + He already answered this in the OP. In the second and third paragraphs he says he now does 13H/15P because it's more economic, and that he no longer does 16H/15P.
Sorry, I missed that somehow. But as a major contributor to that thread, I can say with some amount of authority that that the results changed significantly once a small bit of reality was injected into the builds (gas, spinecrawlers, etc), and that wasn't ever reflected in the OP, since he was only doing drone races. Which 13h15p was the best at, but which isn't particularly relevant, unless you're going to build up to 2 queens and 48 drones without doing anything else at all. Here's a link to the post I made with conclusions about hatch-first builds as well as graphs allowing you to draw your own conclusions: http://www.teamliquid.net/forum/viewmessage.php?topic_id=174374¤tpage=19#374
That post is really interesting. I didn't know that the hatch/pool timings have been analyzed that much, thanks!
I've been going 13H 15P simply because that's what the guide said I find that as I start to get around the ~30-40 food mark my income exceeds my ability to make larva. I've recently switched to grouping all my queens together and using the backspace trick to spawn larva, and I'm not great at it yet, so that might be one cause. I'm gonna try using 14H 15P and 15H 14P as well, see if I can strike a better balance between money and larva.
I've also noticed that it helps to have two control groups of queens. One group which has one queen for each hatch which you use for spawning larva, and another group of your creep tumoring/transfusing/defending queens. I found it hard to keep spawn larva up in the later stages of the game when I'd have all my queens on one hotkey, because I'd want to send my queens into battle to heal units and I'd forget that I need to leave some behind to spawn larva to rebuild after. When I went to respawn again my queens would have to run halfway across the map and I'd end up with a couple thousand resources in the bank because I would just completely fail on larva injecting. I think I get way more queens than the build asks for (I try to get like 6-10 out eventually) because I love transfuse micro... even though I suck at it.
|
|
|
|