On June 02 2013 00:51 TheDwf wrote:Show nested quote +On June 01 2013 23:08 opisska wrote:
It's maybe too eearly, but I would really like to read some serious analysis of the games from someon, who actually understands what's going on.
Because at first sight, it's very difficult to comprehend, why Innovation doesn't play just a tad bit safer. He is supposedly amazing in long macro games - and having seen his brutal multitasking when dropping multiple places at once, I must agree that there is something to it. But he is always trying to get there with an advantage - be it with hellbat drops or very fast 3rd. The question is: is this advantage necessary? When he is so good in long games, couldn't he play them from slightly behind?
My feeling is that he actually needs this good start. He is at "his game" only when HE is the aggressor. When HE is the one constantly attacking at every front and that is very difficult to achieve when you don't have a comparatively stronger economy. But that would mean that his play is in princpile solvable and he could very quickly fall apart. Yet, only very few people can seem to find the solution - and only for Soulkey is really worked. Was it because he is a superior player to the other Zergs that failed, or he (or his teammates/coaches) were the first to really see through Innovation?
CC rax gas bunk fact reactor CC lab stim EB EB
is safe—but it's a razor-edge opening, you have little room for mistakes to defend things like Roach/Baneling busts. Terran don't play (yet) defensive Tanks as a standard because it slows their development. People forget about it because they're mesmerized by "Mines" or "Hellbat drops" but Terran has a lot of stuff to do by midgame—basically you want to cripple heavily Zerg before ultras, since for infrastructure reasons (Terran's production being much slower to Zerg's) you get bashed if Zerg gets his hive units (usually ultras, rarely broods) in a commanding position. Even if broods/infests is extinct, the time bomb is still there to some extent (in this case, Zerg threatening to kill you with a 5/3 ultra timing while you don't have enough Marauders to handle it, or an "oldschool" brood push as Life did several times) and you have to defuse it by pressuring Zerg heavily during midgame—hence the triple OC and double EB openings so you match Zerg's upgrades and economy for your 4M initiative during the Lair phase.
Basically the issue is that you want to optimally develop your infrastructure if your opponent plays a standard 3-bases lings/banes/mutas, while you have to concede defensive measures such as bunkers/Marauders/Mines or Tanks against the various forms of agression at Zergs' disposal. The second issue is that larva inject's outbursts of production make Zerg difficult to read. Even when you have Hellions in front of his bases and you see Roaches pushing them back you still have no way to know for sure what he's doing, and to what extent he will commit; for example the difference between defensive Roaches and an eco Roach/Baneling bust is simply what is in the eggs at the moment (drones or lings). So either Terran gets +3 bunks regardless, at the risk of overreacting against a Zerg simply moving with Roaches to scare him, or he calls "bluff" with an instant loss if he guessed wrong (e. g. Bogus vs Symbol, Daybreak).
Bogus lost the Red City game partly because he didn't build a round of Mines after his 6 Hellions; he directly lifted his fact to make another reactor, which isn't good against a Roach/Baneling bust. He could still have held with better control (in particular you have to evacuate the bunker in the wall shortly before it dies because half of your bio units are there), but it would have been difficult. The "2 Mines vs reactor" decision is directly related to the "optimal development vs standard | defensive tweaks vs agression" theme I talked about above.
As for the Atlas game, 3-bases Speedroaches/Speedbanes bust are another thing. It plays upon the same ambiguity, but at a later timing—i. e. your opponent gets a quick third, starts to saturate it but stops at some point (55-65 depending on how "all-in" he wants to be), turns 100% of his production to units and goes for the kill. This, again, is difficult to scout for Terran, because seeing dual evo + Roach Warren + Baneling nest doesn't tell you what Zerg is actually producing. Without Tanks you have really little hope to hold this kind of attack in an efficient way.
So, to sum it up:
1) When playing 4M, Terran needs to damage Zerg before hive.
2) This is best done using the optimal builds, i. e. triple OC dual EB, and relying on excellent control/scouting to hold agression.
3) Zergs can nonetheless try agression knowing a) Terran has troubles scouting it in advance; b) the margin of error for Terran is thin, which is advantageous when people feel pressure (= more prone to mistakes) like in tournaments; c) larva inject allows them to fall on their feet
as long as they don't fail too miserably [naturally, depending on how things went with the agression, Zerg can end up behind, not necessarily on economy but with lack of creep spread and delayed tech, in particular upgrades, but if they deal a certain amount of damage they can have a game].
4) This means Terrans have the choice between sticking to the optimal openings and aim at playing them perfectly, or they can try to adapt using less razor-edge openings such as Hellbats drops into bio, Hellions/Banshees (very rare nowadays) or 1-2 defensive Tanks (best on maps like Neo Planet S and Star Station in which your Tank defends the natural from the highground).
5) But those openings, being less ambitious, give you a weaker midgame than the optimal triple OC dual EB, while the 1) point above is still hanging over Terran's head. For example, if you open Hellbats drops into bio, your third and dual ups are delayed, so if your drops don't yield results, Zerg achieves a better position in the game (this is what happened in the last game). Same thing happens with Hellions/Banshees (which is similarly a way to pressure + scout while being safe against things like Roach/Baneling busts). The problem with defensive Tanks is that they are what they are—defensive; they don't threaten anything, don't disrupt Zerg in his droning, and are thus a waste of resources (even if you have no longer to search Siege Mode, they come with a price) should Zerg play standard. But the point is precisely that Zerg isn't telling you if he's playing standard or agressive.
Last point about the "but he has the best TvZ in the world so he should play safe" argument. Yes and no. At this level of play there is always the issue, fueled by what about I wrote above about Zerg's ambiguity, of safe being "overly safe". It's incredibly easy to get "punished" by caution: all it takes is one misread and you weakened your development for no objective reason. When playing someone like Soulkey, who is particularly obnoxious with lings/banes/mutas (I say this as a compliment; from a Terran's point of view, having to deal with constant lings/banes counters or mutas wreaking havoc in your base while Zerg defends with lings/banes at home is horrible, and this is truly the right way to play lings/banes/mutas; Life is another example of a Zerg with a similar great tactical sense), you definitely don't want to bestow him any advantage.
When Bogus plays Symbol, he can probably concede blind defensive measures each game without problems because he's one league above him anyway, so he might come back simply through superiority. But as evidenced by the last game on Daybreak or the Whirlwind game he lost on Proleague, things don't work this way against an opponent of similar skill; you don't want to leave too much room/initiative to Soulkey—else he smothers you to death with his guerilla style.
EPT
