[H] 3 roach+speedling all-in PvZ - Page 27

On February 03 2011 06:39 methoD. wrote:
hum, i faced this allin today in ladder. Dont know, if the zerg do it very well. But, i scout the roach warren, and thats the crucial thing, after seeing the roach warren, you should instantly cronoboost warpgate research and warp some stalkers in. Its low diamond (2.6k) but eventually it helps

http://www.mediafire.com/?9joma9bk9mgky9w

On February 03 2011 06:29 Jayrod wrote:


This isnt necessarily directed at darkforce, but i wanted to use your post as a jumping off point

While I appreciate and understand your use of the larva use order as a means to justify a setback to the zerg. I think you may be overlooking a less than mathematical aspect of the game from that point forward. Sinking 600 minerals to defend an attack that might not happen delays any possibility of aggression from the protoss for an extremely long time. Since the threat of aggression is so far off suddenly due specifically to your choice to open this way... your decision of whether or not this larva will become a drone or a combat unit becomes substantially easier. More likely than not, it will be a drone knowing you are completely safe.

This delay in the ability to move out of your base as the protoss segues into my next point. There are obviously different phases in the game for map control and both players are constantly vying for map control and to regain map control, etc. Zerg get free map control early on with a couple overlords and some really quick and inexpensive combat units. To combat this the protoss will either make a sentry based army to ensure a safe retreat if you run into problems OR phoenix/VR (each operate differently in their role to obtain midgame map control). Now phoenix can give you SOME control and mobility, but basically what a phoenix based map control attempt does for you is show you a bunch of shit all over the map (expansions/creep spread/what have you) that you cant do anything about. Its tantamount to getting your observer into their base to see that a cloaked banshee just arrived at yours... All your doing is getting a glimpse and what is going to kill you sooner or later. 600 minerals in a forge and 3 cannons along with the delay in tech and expansion allows the zerg (or any race really that scouted you have such a defensive posture) to cut into this period of time where protoss can gain map control. Essentially you are skipping the early-midgame phase where protoss has a chance to seize map control back before the possible mutas hit or the creep has spread like crazy.

Having said all that, I am happy that this build has surfaced... I've actually seen this style quite a few times but none so precise as the one in this replay. I think we protoss got too used to zergs being predictable.... and they really are. Builds like this help the game more than they hurt the game and make it more dynamic. While I think roaches are a little too efficient for their cost (ive always felt 100/25 would be appropriate for their relative power and high hp) i do think there are some do's and don'ts in dealing with zerg in these first 5 minutes..

1) Do keep your probe in their base until a queen pops or lings pop. Your probe might live... might die... regardless of either... the information its giving you is worth the 50 minerals and production time, etc. Even if you are just seeing what kills it. I'm surprised you arent doing this anyways... I make sure 100% of the time that I see either lings or a queen before i try to escape because no matter what build hes doing how can you be sure those arent just more drones? The simple fact that he built lings covers some of the cost of losing that scout. If he sees you leave before he makes lings he can just make a drone with that larva instead. You are actually worse off in that situation than if you just let his lings kill you.

2) Do BUILD STALKERS if you see a roach warren. Regardless of whether its 7RR or this roach/ling all in... Protoss got so used to this 3 gate sentry expand. Its a great build... I love it... i used it in 90% of my PvZ games... but just be flexible... if they GAS steal you... another reason to build more stalkers and less sentries. If they gas steal you I would expect roaches to be honest. Start making logical conclusions if you cant see exactly what theyre doing. You might be wrong sometimes and it might put you behind and you might lose due to that... but you aren't learning ANYTHING by not making those theoretical leaps. There is alot of discovery to be done in those theoretical leaps whether you are wrong or not.

3) Don't rely on sentries to defend a 1 base all in from zerg. With OL sight range, sentries plain suck at defending against this type of thing. They can be good, but dont rush to get a ton, they dont make you invincible.

4) Don't be afraid to lose probes for information... To do so is assuming that you and your opponent are so evenly matched that even 1 probe can change the outcome of the game. Sometimes that is the case, but in EVERY game you should always assume you are a stronger player and that you will respond to what you see in a way that makes sense even if it doesnt work out.

all in all i think it was unreasonable on your part build a second sentry and that was mainly due to poor scouting. If you knew he would do it a third time... all the less reason to build the second sentry. Having said that, you tried it, it didnt work... and it wont work. Its not the correct response. I agree cannon is not the correct response either. Getting a fast +1 can help mitigate the setback you imposed on yourself... but overall you will be behind the zerg that sees the forge with his OL and decides to just not all-in. Its always better to defend with units when possible because units move (duh).

Earlier 2nd gate, more stalkers, delay third gate to ensure constantly produce units and chrono gateways instead of warpgates. Additionally, scout better. If you scout an indicator that hes not expanding quick, (saving larva for instance), get fewer sentries and more stalkers. The stalkers are crucial but its by no means a slam dunk. Basically hes cutting workers at 14, youve spent another 500 minerals on workers that he put towards units. His army should be about 500 minerals stronger than yours as such... why should your army win? scout. extrapolate. learn from the poorer extrapolations you make. improve.

On February 03 2011 07:00 JoeCrow wrote:
How does this build fare against a terran?

On February 03 2011 07:11 FrostedMiniWeet wrote:


If you actually think roaches should be 100/25 you're insane. The roach is only strong in the early game and eary mid game before protoss gets massive ammounts of sentries, or collosus, or void rays. After that roaches just get decimated by proper use of FF and Collosus, and making them more expensive would only exacerbate the problem.

And it should also be noted that while roaches can be strong early game, if you get them early game, your economy is completely screwed and you pretty much auto-lose if you don't do serious damage with them, which is why this kind of all-in strategy isn't seen that frequently by zerg players.

I'd much rather have the power of the 4 gate over this build.

On February 03 2011 07:22 CatZ.root wrote:


no its not you silly boy, I didnt want to play, so I was gonna go for a proxy hatch in plain sight, as you can see my minerals pile up to 300 early on, and my drone can't hatch so i just go back to my base and try to all in asap.

I was playing RANDOM, so the last race I wanted to get was ZERG. im glad you're an acomplished progamer, but you did not beat the strategy that's talked about in this thread, sorry.

On February 03 2011 07:31 Kefka.dancingmad wrote:
Hey just faced this on the ladder,

[url blocked]

THATS HOW YOU DO ITTTTTTT :D

On February 03 2011 07:12 FrostedMiniWeet wrote:


It would die to 1 bunker, 1 marine, and scv repair..

On February 03 2011 07:59 Jaeger wrote:
Just a few notes if you're plan is 3gate sentry expand,
Your core is ~5 seconds late.
You usually build a 2nd gas before zealot with this build.
Your sentry is ~13 seconds late because you don't have enough gas to start it on time.


You have several bumps in probe production because you don't have the money to constantly produce them.

1 second after gateway
3 seconds after gas
3 seconds after second pylon
1 second after core
1 second after second gas
11 seconds after warpgate tech
5 seconds after 2nd/3rd gateway

This might seem super nitpicky but as each chronoboost gives you an extra 10 seconds of production time thats over 2 chronoboosts of wasted production time.

So basically if you:
- place your core on time
- only chrono boost your nexus once
- constantly build probes from your nexus (this part requires pretty good worker micro actually)
- build a 2nd gas before zealot
- chrono boost out the zealot
- chrono boost out the first sentry

You end up with the exact same everything except you have an extra sentry and a second forcefield on your first sentry.