|
WARNING SPOILERS
This season of code s has seen a lot of mech play, about 1000% more mech play than usual (MVP excluded over course). Terran don't seem to be winning with mech though. I looked over the round of 32 and round of 16 groups A and B and compiled the result, thus far code s terrans have tried mech 16 times, 17 times if you count game 1 of life vs 8thTeamBaby. Of those fifteen or sixteen games terrans have won with mech exactly three times, giving the strategy about a 25% winrate. Ouch.
Guimho beat hyun with mech in game 1 of their series
Hack beat roro with mech twice in games 1 and 3 in their series.
MKP beat Hyun with a mech timing in game 3 of their series.
Of those victories, hyun committed a lot to roach ling aggression against guimho. Guimho weathered the storm, maxed out, and then crushed hyun because he never reached hive tech.
Hack was able to beat roro twice with mech hitting the ~16 minute mech pushing timing. Roro's tech tree wasn't quite complete, and he pushed for victory.
Everyone else who tried mech was not so successful. No one so far has played mech perfectly, but a few people have come really close. They all seem to be going for a 16 minute timing push, and time and time again zerg is always able to reach hive tech and get broodlords, and then mech doesn't have enough vikings, and mech dies. Ryung played some great games against curious where, although he didn't go mech, he transitioned late game into something very close. On Entombed valley he crushed curious's broodlord army, only to sacrifice all his viking rather needlessly and die to the ultra switch.
There seems to be a theme in terran vs zerg emerging, and its hardly a new one. If zerg gets broodlord infestor and a healthy economy zerg has won literally every time. The one exception this season is Ryung played a game against curious on whilwind where he was able to cripple curious early. Curious managed to get a very nice Broodlord infestor army out anyway, and it went to a very scrappy base trade that Ryung was able to pull out because of his vastly superior economy.
Now before the zerg advocates call me out for making a balance complaints, that is not my aim here. I don't have an opinion as to whether or not TvZ is balanced, although I do think broodlord infestor, coupled with the ultra switch, is one hell of a one two punch. Terrans in code S this season or in any other season have not performed inspiring at all against the late game composition. I haven't seen a terran commit to winning the air battle yet and pull it off, the ryung game being the most recent exception to that, and I can't recall a game where terran was able to win the air battle and then follow up correctly to win the game. Its hard to say if it is really possible because you can't say that terran by and large have been playing that well against the composition.
The reason the recent mech trend is so interesting to me, is mech seems like the most viable option for terran if they want to beat zerg late game. Ryung has played really well with marine tank, even making wonderful transitions late game, but it seems to me that mech is still as superior strategy for making it to late game vs zerg and building up a proper transition to conquer the BL infestor + ultra tech switch. The problem seems to be right now that everyone is playing mech exactly like MVP played mech vs life. The 16 minute mech push didn't work even one time for MVP in the finals and with rare exception it hasn't worked in code s for anyone else either. Personally, I think it's just a shity strategy that hopefully will die very soon, or be delegated to a timing alternate to keep zergs honest in an evolving metagame. What interests me is the possibility of terrans turtling longer, and trying to win a heavy viking/raven mech push that hits several minutes later, and then transition back into tanks or possibly BCs for the super late game. No one is doing that right now, maybe its a bad strategy hell I am far from an expert, but clearly the way mech is being played at this time in korea is very much not working.
|
MKP beat HyuN in game 3 in Ro16 B with a viking anchored aggressively defensive mech timing designed to hit before BL
|
Ah, haven't seen those games yet. I will add it to the OP. Brings Mech victory percentage to a staggering 25% then, assuming there were no other mech games in the group.
|
The MKP game is interesting because he was expanding very aggressively, leaving HyuN with choice of delaying tech for aggression, or let MKP gets free 4 base econ. He chose former but didn't do enough damage to transit into late game, but even if he chose latter MKP would be able to max out and push before hive with insane econ
This strat obviously is very map dependent since it requires 4 bases that's relatively safe to defend with an immobile army.
|
Ganzi was able to win with mech against a zerg in an IPTL match going for air, unfortunately I do not remember who he was playing. He had the tech ready so when the ultra switch came he could produce banshees immediately. He ended up winning that game when he looked dead and gone.
|
can't be IPTL since he didn't play any TvZ...
Only TvZ I can find of Ganzi was back in Summer Arena, that can't be right
|
On November 15 2012 12:54 Ricoic wrote:
Ganzi was able to win with mech against a zerg in an IPTL match going for air, unfortunately I do not remember who he was playing. He had the tech ready so when the ultra switch came he could produce banshees immediately. He ended up winning that game when he looked dead and gone.
"a zerg"
a korean zerg
an eu zerg
or an na zerg?
|
United States2186 Posts
I don't want to be mean but this is why they are in code s and you are theorycrafting. They are doing it for a reason, so maybe try to figure out why instead of proposing fairy-tale ideas like purposely trying to fight broodlord/infestor. 9 times out of 10 you can't, which is why people don't try it on stage.
They are playing a strategy which relies on them performing close to perfectly to win, though it works pretty well when they do. In practice, when they can play perfectly, the build works out well. In the nerve-wracking gom booth, when both guys aren't playing perfectly, it does a lot worse because Terran mistakes count for far far more than Zerg ones. But that is something that is extremely hard to understand as a player, so it's natural they keep doing it. And frankly, even with that, it's still one of their best hopes of winning. Maybe they'll be able to overcome nerves and play flawlessly that day. Players have to think that way: it's almost absurd to plan ahead for getting nervous, because that vastly increases the likelihood of it occurring.
For example, when Flash does a similar strategy he is able to truly hit before 15 minutes, not 16 (assuming no Zerg aggression) as he doesn't make those errors. That's how he beat Soo on Metropolis in MvP: he attacked while Soo was hive rushing (off of a muta opening) because Soo didn't think it could that fast, so he was stuck with no lair and no hive army. Earliest broods are going to come at 1530-16 mins, probably 1630-17 if they make enough roaches to appear threatening. In the finals Mvp kept getting delayed due to bad decisions/life playing well. In Ryung vs Curious on Entombed Ryung loses 10 hellions and a banshee to random fungals when marauding around. That in itself almost loses him the game, even though it may not be obvious, because now his push comes later, and Curious makes broodlords in time. Mech has zero margin for error.
They all seem to be going for a 16 minute timing push, and time and time again zerg is always able to reach hive tech and get broodlords, and then mech doesn't have enough vikings, and mech dies.
You seem to be under some kind of illusion that making vikings actually lets you beat broodlords: it doesn't, beyond an extremely narrow timing window against a hive rusher who will have no corruptors for a minute (which, amusingly enough given your post, is what Mvp kept trying to hit in the finals and almost won on entombed had he not clumped his vikings and auto lost to 1 fungal; he also did win on Abyssal with it). Mvp's strat is really quite clever: you ram his support casters away with your excess of tanks; he can't broodling wall because he hasn't reached a critical mass of broods yet, and he has no/few corruptors to morph into broods because he needs broods to kill your ground army, so he loses them all to vikings and you win the game. It's extremely tough to hit but it works.
Outside of that timing, if you're making vikings you've basically already resigned to lose the game unless the Zerg has a seizure and clumps his brood/corruptor against seeker missile (like Soo vs Major on Daybreak from MvP); something which you cannot engineer. And even then, it's still dubious. The Ryung/Curious game where Curious made a horrific error and lost his entire broodlord army and still won 2 minutes later despite Ryung playing incredibly well is the kind of game that happens all the time in practice (except the Zergs are less likely to blunder in that way during practice and thus win even easier).
Calling Mvp's strategy a "shitty strategy" is beyond comprehension. Mvp is the most innovative player by far in the history of the game. Any build he does has very good reasons behind it. Instead of insulting him, learn from him.
|
Hey I don't claim to be an expert. If flash can hit the same timing one minute ealier then maybe there is hope for the strategy. I am a pretty intelligent guy who has watched these mech strategies a lot, and it just doesn't look like its working. As a matter of fact, it isn't working. I don't think it unreasonable to call the strategy bad if nobody is winning with it, and its premise that it is built upon is fundamentally flawed; a timing push meant to hit before broodlords that consistently doesn't hit before broodlords has some problems with it. If you are joe shomo on the latter or MVP in the finals it doesn't matter really, your strategy is bunk, or in the very least your execution isn't working.
But I didn't write the thread to shit on MVP or any of the other best terran in the world! If you can provide some replays of the strategy being perfectly executed I would love to see what it looks like. I am just an esports fan, and I have no problem admitting I don't know very much compared to a progamer.
|
I def agree with Ver, especially on the point of vikings. When you make lot's of vikings, you are attempting to play zerg at their own game, and you won't win.
When vikings are made they are designed to counter broods, they do nothing else for you, and there is no way for you to transform them into something useful when they are countered by a swell of corruptors that can be made at will. However when zerg makes corruptors and doesn't need them anymore, he can turn them into broods, to serve an entirely different purpose.
|
I don't see a way to beat zerg with mech without vikings. Unless every one who has played mech in the last code s have all been executing the build wrong, the 16 minute mech timing doesn't work. If terran can't play late game at all with mech vs zerg that it is tempting to say mech doesn't work, or mech as it is being played doesn't work.
|
On November 15 2012 14:02 mothergoose729 wrote:
I don't see a way to beat zerg with mech without vikings. Unless every one who has played mech in the last code s have all been executing the build wrong, the 16 minute mech timing doesn't work. If terran can't play late game at all with mech vs zerg that it is tempting to say mech doesn't work, or mech as it is being played doesn't work.
You're wrong, and you are misunderstanding the point of the mech builds. Re-read Ver's post until you understand.
|
I feel like a common theme (or, misconception) when observing TvZ games is to look at individual blunders, and then use those blunders to explain away Terran losses. A deeper analysis would show that these mistakes are not always unavoidable pitfalls, but instead, they're artifacts of builds and mindsets that are meant to overcome disadvantages they are already aware of.
To give you an example, we saw Bomber lose to ____ recently (forgot who it was, someone help me - it ended up being 2-3 loss for Bomber, Day9 was commentating) on Antiga shipyard because he repeatedly pushed forward too far past the center watchtower while unsieged, and ended up in unfavorable engagements. The commentators explained it away as the reason for the loss of the game. Was this just Bomber not paying attention? Is Bomber unaware that moving past the middle of the map unsieged is risky? No.
We see players hitting mech timings prior to broodlord without vikings, and oftentimes losing because their timings got delayed and they run into broodlords. Are these players completely unaware of the possibility of losing to broodlords? No.
Terran mistakes in TvZ typically result in outright losses, whereas the margin for mistakes as the Zerg player is higher. My opinion is that the builds and tactics that seem boneheaded to you (i.e. not building vikings, moving past the middle of the map unsieged, etc.) are calculated decisions. The chance of gaining an advantage by your opponent failing to punish your mistake has a higher reward:cost ratio than trying to prolong the game into an unfavorable situation where the odds of you being caught unsieged, getting caught by good fungals, getting caught too high on vikings, getting caught too low on vikings, result in an insta-loss.
Even at pro levels, I'd take the chance that my opponent doesn't have broodlords in time or lets me move up too far before engaging over the chance of avoiding every single insta-loss opportunity that results from extending the game another 15 minutes.
Looking at these individual situations without more context is missing the forest for the trees.
|
On November 15 2012 14:05 Fission wrote:Show nested quote +On November 15 2012 14:02 mothergoose729 wrote:
I don't see a way to beat zerg with mech without vikings. Unless every one who has played mech in the last code s have all been executing the build wrong, the 16 minute mech timing doesn't work. If terran can't play late game at all with mech vs zerg that it is tempting to say mech doesn't work, or mech as it is being played doesn't work. You're wrong, and you are misunderstanding the point of the mech builds. Re-read Ver's post until you understand.
Lets aviod attacking the person here (namely me :p). Belittling me is fun but its not very productive.
I read ver posts. I respect his opinion. Maybe mech perfectly played looks very different. If so, for my own sake, I would very much like to see an example of mech perfectly played. Mech as it is currently being played in code s isn't working. Maybe the discussion can shift towards what these players could have done better in their games to make their mech play work?
|
On November 15 2012 14:14 iEchoic wrote:
I feel like a common theme (or, misconception) when observing TvZ games is to look at individual blunders, and then use those blunders to explain away Terran losses. A deeper analysis would show that these mistakes are not always unavoidable pitfalls, but instead, they're artifacts of builds and mindsets that are meant to overcome disadvantages they are already aware of.
To give you an example, we saw Bomber lose to ____ recently (forgot who it was, someone help me - it ended up being 2-3 loss for Bomber, Day9 was commentating) on Antiga shipyard because he repeatedly pushed forward too far past the center watchtower while unsieged, and ended up in unfavorable engagements. The commentators explained it away as the reason for the loss of the game. Was this just Bomber not paying attention? Is Bomber unaware that moving past the middle of the map unsieged is risky? No.
We see players hitting mech timings prior to broodlord without vikings, and oftentimes losing because their timings got delayed and they run into broodlords. Are these players completely unaware of the possibility of losing to broodlords? No.
Terran mistakes in TvZ typically result in outright losses, whereas the margin for mistakes as the Zerg player is higher. My opinion is that the builds and tactics that seem boneheaded to you (i.e. not building vikings, moving past the middle of the map unsieged, etc.) are calculated decisions. The chance of gaining an advantage by your opponent failing to punish your mistake has a higher reward:cost ratio than trying to prolong the game into an unfavorable situation where the odds of you being caught unsieged, getting caught by good fungals, getting caught too high on vikings, getting caught too low on vikings, result in an insta-loss.
Even at pro levels, I'd take the chance that my opponent doesn't have broodlords in time or lets me move up too far before engaging over the chance of avoiding every single insta-loss opportunity that results from extending the game another 15 minutes.
Looking at these individual situations without more context is missing the forest for the trees.
Hey I agree 100% with you. That is exactly what I think is happening with mech right now. The question is, is their strategy off or their execution off? There have been a lot of builds that have gone the way of the dodo, but before they did players lost using them many, many times. Is this mech timing one of those builds?
|
On November 15 2012 14:15 mothergoose729 wrote:Show nested quote +On November 15 2012 14:05 Fission wrote:On November 15 2012 14:02 mothergoose729 wrote:
I don't see a way to beat zerg with mech without vikings. Unless every one who has played mech in the last code s have all been executing the build wrong, the 16 minute mech timing doesn't work. If terran can't play late game at all with mech vs zerg that it is tempting to say mech doesn't work, or mech as it is being played doesn't work. You're wrong, and you are misunderstanding the point of the mech builds. Re-read Ver's post until you understand. Lets aviod attacking the person here (namely me :p). Belittling me is fun but its not very productive. I read ver posts. I respect his opinion. Maybe mech perfectly played looks very different. If so, for my own sake, I would very much like to see an example of mech perfectly played. Mech as it is currently being played in code s isn't working. Maybe the discussion can shift towards what these players could have done better in their games to make their mech play work?
This mindset is what makes this topic hard to analyze. Certainly, if "perfect play" was happening on both sides, Terran players wouldn't be caught by broodlords and lose. That's because if "perfect play" was possible, we'd be playing a completely different videogame, one where the most optimal composition in every single matchup is marine/medivac with perfectly-split marines fighting blink stalkers that blink away the second their shields deplete, leaving them invincible.
That's a hyperbolic example, but it illustrates the point that players have millions of inefficiencies per game, ranging from insignificant to game-losing. When you attempt to blame losses on micro mistakes (getting caught by fungal, for example), you end up missing the real lesson.
Pure marine isn't viable because players make mistakes, and creating a composition where bunching up for one second results in a loss isn't sustainable over a 25 minute game, let alone a series of 25 minute games. Marine/tank is a more sustainable version, and mech is a furthermore more sustainable version. Take this logic to its conclusion and you can understand why we see people doing things like 16 minute mech pushes instead of slow-pushing on creep with maxed marine/tank.
|
I hear way to many commentators saying "oh he caught caught unseiged thats why he lost" but in reality a code s terran isnt stupid and he wasn't "caught". Have you every seen anyone actually leapfrog tanks before or have you done it yourself? It takes at least twice as long and with maps as big as they are these days it takes just too damn long.
|
On November 15 2012 14:17 mothergoose729 wrote:Show nested quote +On November 15 2012 14:14 iEchoic wrote:
I feel like a common theme (or, misconception) when observing TvZ games is to look at individual blunders, and then use those blunders to explain away Terran losses. A deeper analysis would show that these mistakes are not always unavoidable pitfalls, but instead, they're artifacts of builds and mindsets that are meant to overcome disadvantages they are already aware of.
To give you an example, we saw Bomber lose to ____ recently (forgot who it was, someone help me - it ended up being 2-3 loss for Bomber, Day9 was commentating) on Antiga shipyard because he repeatedly pushed forward too far past the center watchtower while unsieged, and ended up in unfavorable engagements. The commentators explained it away as the reason for the loss of the game. Was this just Bomber not paying attention? Is Bomber unaware that moving past the middle of the map unsieged is risky? No.
We see players hitting mech timings prior to broodlord without vikings, and oftentimes losing because their timings got delayed and they run into broodlords. Are these players completely unaware of the possibility of losing to broodlords? No.
Terran mistakes in TvZ typically result in outright losses, whereas the margin for mistakes as the Zerg player is higher. My opinion is that the builds and tactics that seem boneheaded to you (i.e. not building vikings, moving past the middle of the map unsieged, etc.) are calculated decisions. The chance of gaining an advantage by your opponent failing to punish your mistake has a higher reward:cost ratio than trying to prolong the game into an unfavorable situation where the odds of you being caught unsieged, getting caught by good fungals, getting caught too high on vikings, getting caught too low on vikings, result in an insta-loss.
Even at pro levels, I'd take the chance that my opponent doesn't have broodlords in time or lets me move up too far before engaging over the chance of avoiding every single insta-loss opportunity that results from extending the game another 15 minutes.
Looking at these individual situations without more context is missing the forest for the trees. Hey I agree 100% with you. That is exactly what I think is happening with mech right now. The question is, is their strategy off or their execution off? There have been a lot of builds that have gone the way of the dodo, but before they did players lost using them many, many times. Is this mech timing one of those builds?
I wish I had an answer, but I only have an opinion. I don't believe (and never have believed since the beginning of this game), that a Terran player can do the same build, same timings against an equally-skilled Zerg player every game. The matchup revolves around the ability of the Zerg player to drone/produce army according to what they expect their opponent to be doing. For this reason, I'm not a fan of trying to find "standard" TvZ builds. I actually think the mechanics of the game fundamentally don't support it.
The most successful TvZ strats are the ones that either:
1) Exploit the metagame.and are used in an unpredictable fashion (i.e. MKP's refinery-first hellion/medivac play recently)
2) Are unscoutable (i.e. MVP's GSL finals play where he used the same opening all 7 games, but compositions varied - although that needs more development still, as there are only a couple variations)
My prediction is that, with time, people will realize that "standard" TvZ builds that you can execute exactly the same many times in a series don't work.
I don't think mech will go away, but I think this specific timing will see less and less success as time goes on, for this reason.
|
You bring up a good point about perfect execution. Perhaps better questions to ask then would be, how come terrans are having such poor success with this build in this match up at this time? This is the discussion I originally wanted to have. Perhaps that point was side tracked or lost at some point by something else I said.
|
i was wondering:
if instead of 3 ocs, what if terran make the 3rd a pf instead.
Yes that is bad in the typical sense, but for mech where gas is much more of an issue then minerals when massing those tanks/thors, It would be then MUCH easier and faster to secure the third and gather the gas necessary for the mech timing.
and your mining would not be disrupted by some ling/roach pushes before you get a solid tank defense up that forces you to lift for quite a while
|
|
|
|
|
|
EPT
