Post your ideas: Strategy Talk

Question by machinehead:
I saw some of the games you streamed last night and it was fantastic stuff. I liked your logic behind not making marines on big maps, but on fighting spirit -- after acknowledging how strong 12 nexus can be -- it would seem that maybe making rines to be able to punish/beat 12 nexus would be a good move. I say that on the grounds of watching flash play, where it seems he can auto kill 12 nexus at any position but cross -- granted he has marines to send. Basically, I'm curious to know how you respond to 12 nexus if you don't have the ability to hurt them early.

Reply:
the probability that they are not cross position is 2/3. the probability that you scout first to react quick enough(assuming normal scouting) given they are next to you is 1/2. So the probability that he is not cross positioned and I scout him first is 2/3*1/2=1/3. We already know it doesn't even matter if they are cross positions which even if i scout that doesn't allow me to do anything. Therefore, I'm not going to have my game dictated on getting marines. If the player even does 12 nexus, I have a 1/3 chance of even reacting with marines which is not the majority of the times. Hence in the long run, making marines is a poor decision. That being said, i generally need to get back into the game some other way. I generally do this by optomizing my build. Knowing that the toss has to go observers, I shoudln't ever be getting an engineering bay at all. Instead, getting a starport and threatening a drop should both keep him in his base long enough for me to take position for another expansion, and let me play optimally to know when i should get any detection if at all. THe toss on the other hand has a late scout so the player will never know how to react otpimally and you can get even from optomizing your build and denying protoss a deliberate path.

What do you think of infernals PvT build where he builds his second nexus before his second pylon? The strength of the build is to hopefully fake the T out by a lack of a second pylon.

Well the strength of this build is actually to take the expansion and have an early macro economic lead at the cost of a couple of probes temporarily, and a timing window for the terran to be able to damage the protoss. Time will compensate for the lack of probes since the 2nd nexus will be up before normal timing.

Would you recommended this build? Why or Why not?

Though I don’t have much experience with this build, I would not necessarily recommend it as a Protoss. Because you get the 400 minerals before you get the second pylon, you will be capped on 17 for a while, giving you a maximum of 2 dragoons(assuming you cut on 13 probes). If the protoss goes very macro aggressive, they will only have 1 dragoon(15 probes), and probably set down 2 gateways after the 1st to compensate for the lack of dragoon count. Obviously the timing to attack is when the protoss has that 1-2 dragoons and is waiting for the pylon to be finishing. As long as pressure is done then, the protoss should lose the nexus or at least waste tons of mining time. We have to remember the overall goal is to hinder the protoss from getting a macro lead. If we just pressure, it should force probes off minerals. With the already low probe count, the toss should be dying for minerals, while the SCV count is still plentiful. This is when you can either apply more pressure or expand and take your economy lead as terran.


Do you have a similar build that you think is better? Even if it's just putting that second pylon down before the second nexus at your potential 3rd base?


I believe this build is geared to certain players styles that tend to be much more passive in beginnings. This build is probably used more as a tactical play rather than sound opening that’s consistently used. It relies on the terran assuming normal play and can be very gimmicky to a player that isn’t able to react on the spot for the first time seeing it.

There was a discussion on gas steal (pvt) lately, and I noticed quite a few people agreed that gas steal is virtually never advantageous for the protoss, unless the trick is part of a build especially designed to benefit from it.


I play both PvT and TvP, and so I have 2 questions.


1. What would generally be the ideal response to a gas steal (as T.)
Let’s first go over what the gas steal does and how it affects the players. Then we’ll look at the dynamics of how things change the end results.

For those of you that don’t know, a gas steal is when a porotoss will send out an early scout(usually the pylon probe) and make an assimilator on the terran’s refinery. This delays the timing of the gas, making a later timing for tanks at the cost of 100 minerals at beginning game.

Some of the subtleties:
This will also delay tech for protoss since the 100 minerals is being allocated to an offensive gas steal rather than a tech investment.
The terran will also have much more minerals since not only are there no scvs allocated to gas, but the 100 minerals that protoss must spend puts terran ahead 100 minerals in relativity.

That being said, terran has two different responses that I believe are the only ways that can be played. I first want to refute trying to kill the assimilator before the probe is dead because the probe can just gas steal again. It is nonsensical to use 6 scvs to kill an assimilator that costs 100 minerals only to be made again. By you using 6 scvs, you remove 300 minerals from productivity temporarily and mining time, which puts the protoss ahead. Therefore, any response must be done after the assimilator is made and the probe is dead.

The two responses are to either all in with 2 barracks, possibly proxying the 3, or to fast expand. With the 2 barracks, the thought is that the 100 minerals were spent when the cybernetics core should have been made. If the core is delayed 100 minerals, the dragoon will come out 100 minerals later, hence there exists a small timing window to do damage(but must all in with SCVs).

Now using the same logic that the cyber core>>dragoon>>range will be 100 minerals later, that means any offensive capabilities are limited temporarily giving you timing to expand. You can take that timing window to expand knowing that protoss has to catch up economically.

Once you expand, the protoss cannot just simply expand as they have already committed 100 minerals to the assimilator and the fast expansion by terran has been up way before protoss. Therefore, protoss must deal damage and get back in the game or else terran will get ahead incredibly.

To answer your question, I don’t know what the ideal response to it is, as there are many interpretations of how to deal with this build. What I enjoy doing is going 2 barracks in my base to kill the probe and force protoss to make more units(rather than the dragoon range) because I threaten an all in. Then I just simply expand, and go for marine range before factory. My thought process is to defend against dragoon range, and that I have all this extra minerals but I won’t have gas for a while so going straight to factory would incur a surplus of minerals. I then mass marine medics on 2 barracks keeping all of my units very defensive knowing that some sort of pressure will be done to me. If no pressure is done, I know I have a large advantage since economically, I am ahead, and the M&M can still be used well for pushing. From there I play straight up.

2. What sort of PvT build would actually benefit from the usage of gas steal?

Again, this build relies on the tech timing advantage. The delay in tanks needs to be capitalized, therefore getting protoss tech as soon as possible will allow damage to be done. If you are able to get 3 dragoons to be dealing damage to a wall, and you kill 2 scvs, then your build has worked successfully. However, if terran is able to pressure a toss in the beginning, it decreases the amount of opportunity protoss has, and forces them their gas to be allocated to specific areas. This is why I normally do 2 barracks to leave an option for me to apply pressure if I feel the protoss has been too greedy with tech.

I wouldn’t say this build is necessarily refuted and protoss instantly has a disadvantage. For example, if the map lends itself to really nice reaver harassment, then gas stealing would be potentially advantageous such that the reaver can do more damage with a smaller tank concentration while defending. There are too many variables to refute it on every map.

what should i do if i chose the flash build on Desti/HBR... 33armory, 3rd after +1 attack, 4th after +2...etc camp style. What would i do if i scanned him going carriers?

Should i just ditch my camp plan and attack, or keep camping and just add goliath instead?

Your advantage is in your upgrades. No matter what build he goes you will have an upgrade advantage especially against carriers. Just add goliaths and push when you have enough, as his ground will be underdeveloped since his money and upgrades are probably in carriers, and you will have an upgrade advantage on his carriers that is, you have a superior army.

When I played standard TVZ, fast ultra builds always gave me a headache and lots of trouble dealing with. The problem usually was I wouldn't know they were going fast ultras until it was too late to really deviate to a strategy that didn't suck against it. "Fast" ultras could be teching straight to them, or a muta, ling open with a transition to fast ultras. So the questions would be, is there a tell-tale sign when players are going for a fast ultra build? And, what is your strategy once you realize you are against fast ultras? Perhaps it's an upgrading problem on my end that makes it so challenging, I dunno.

The tell tale sign will generally be the 2 evolution chamber upgrades going on much earlier than normal. For ultra-ling to work, there must be an upgrade advantage. The dynamics of the ultra-ling unit composition is that the ultras tank the damage and the zerglings deal the damage. If there exists no melee on the zerglings going up against a 1 or 2 armored marine, that’s a 20%-40% decrease in attacking capabilities, meaning the marine medics live 20%-40% longer(not actual but estimates since there exists more variables of lasting longer than just true damage). You will see 2 evolution chambers always and a fast hive. As soon as you see the 2 evolution chambers, you should realize what you need to prioritize. The three main things I always think to do are to make sure all my upgrades are working, my science vessels are producing, and my barracks count is way up. Notice I didn’t say tanks or vultures! This is just my style

Upgrades have already been explained.

Melee units rely on flanks and arcs around a group of units. How I think of it is that I increase the actual size (volume) of my army with marines and medics rather than getting vultures and tanks. Therefore I’ll generally stay on 2 base and go straight to 7 barracks to accommodate the ultraling army.

I will also get two starports and have constant harass with vessels AND dropships. The important part is to keep the vessels alive(OBVIOUSLY:-P) but realize what having these threats do. By having vessels irradiating and dropships roaming, you force scourage which detracts from the ultra count. This consequently reduces the amount of tanking ability an army has increasing the potential of your army for a battle win. Therefore, when you have vessels and dropships, make sure the zerg sees the count and the dropships. You don’t even need to drop, but only let the zerg see it, that way scourage are made. For every 6 scourage, you’ve effectively killed one ultralisk 

Lastly, I cannot stress enough how important a good unit composition is. Firebats are pretty essential against any zergling/ultra army. Remember that the zerglings are dealing the damage and the ultras are tanking. Ultras alone are terrible as they don’t have the capacity to kill huge groups of marines unless they are in huge numbers. Therefore, when you have group fights, make sure firebats are in your army to undermine the attacking capabilities of the zerg. This will kill the main threatening part of their army, allowing your marine count to stay high and kick the zerg butt 

After a two factory build, many times I find myself having wiped out the protoss's fast expansion and have driven him back up his ramp. At this point, what should I do to follow up? Usually, the protoss will clear out my containing forces with speedzealots and/or a shuttle,. and then proceed to attack my nat and/or take many bases. As a result, even though I was able to take out the nexus, I still end up losing the game.

Great job! You killed the expansion, killing many units and taking out his nexus. So now what? :O! Well here’s the part you need to realize the situation. You have just done a lot of pressure with units, and were able to take out all of his stuff. How is he going to react?! Is he going to macro normally? Of course not! He has to do damage back to you so that he equals out the fact you just killed his economy. Therefore, what do you do against a protoss that MUST be very aggressive as he has no choice?

Defend.

Take all your units at the contain and move them all back to your choke. This will give you extra time as travel time gives you more time to get more units. Also his reinforcements will come later as the travel time also effects them as well. Just expand and get all your units back to retain them. Defend up as he has to do something. If he does nothing, you’re already economically ahead. Let your expansion kick in and push when he gets a 3rd base, as he will have no way to deal with the amount of units you have since you’ve been on 2 base much longer than him!