Hi, fellow players. You may remember my last attempt at a blog (link). It didn't really end well. Long story short, about the same time last year I wanted to go from gold league to platinum league by the end of the year. I probably would have achieved it, if it hadn't been for the release of Shogun 2 to distract me, as well as some other things. I basically let go of it. The attempt ended in the second week of January with a bit of QQ (link). Truth be told, I actually stopped playing a while later. My last replay is dated 29 January. I simply had to get things moving in real life (finishing Ph.D. in criminal trial law). Also, Mass Effect 3 came out, I decided to replay Mass Effect 2 in order to play the DLCs I had missed due to language version/regionalisation issues (thankfully, right now there was a way to swap languages while digitalising your DVD box CD-key, confirmed legal by EA's customer support), and then Crusader Kings was to release on 13 February and Fall of the Samurai (a Shogun 2 DLC) had been slated for the same time, to follow its elder brother Rise of the Samurai that had got released in the meanwhile. Plenty, isn't it? I actually haven't finished Fallout 3 since 2008 but let's not get there.
Anyway, while I didn't play, and still haven't, I haven't stopped watching Day9 dailies or those lovely, so superbly positive casts from Husky. I could actually play, I just didn't feel like playing. Couldn't get myself to. There was always something and if not anything concrete, then at least an inexplicable but very real psychological block. Not unimportantly, I was simply tired. And didn't want the losing drama for a number of reasons. Didn't want to pursue leagues. And... ...and, in reality, couldn't really explain what was going on anyway. But I kept watching dailies and some games. I'd occasionally drop by and check out TL too.
Of late, in direct connection with aversion to honest work (jusk kidding) and the general procrastination that sets in when I need to wind up that doctoral dissertation, I've started following IEM, Intel Extreme Masters. That was not actually the first tournament a part of which I've followed lately, but the first one where I've followed basically every single game involving my race, Terran. And some that don't (cheering for Nerchio and happy to see the compatriot Polish Zerg deliver so much pwnage ...even if he's a Zerg ...it was funny not to want the Terran to win when he played against DeMuslim; while at it: GL HF to Tarson the Hungover Terran, Mana the Serious-Faced Toss, DieStar the Other Terran and others... by the way, there was once a girl who 2:0-ed DieStar on one occasion, she was a pink Terran with a lovely a-move into siege tank lines (and successful), too bad she has retired).
And I also read Boxer's autobiography while on the bus to and from the law school library. Needless to say, it was far more interesting to read than whatever the Supreme Court handed down. Also, becoming a professional Starcraft 2 player felt infinitely more appealing than becoming a bonjwa lawyer at that point (note: I already have the basic law degree, this is a four-year Ph.D. on top of it, well, it has taken five years already but let's not survey this angle, shall we). Heck, screw that. I could be the perennial amateur, the gentleman gamer. I have a different job, one that I own, where I can work a third of the time I did at the last law firm I had worked for, and earn the same money. I could sponsor myself and tell companies whose policies I didn't agree with to go away. I'm still younger than Boxer, you know.
For the record, I could probably try casting. Or casting interpretation. I'm actually a professional translator. My other job. But one where I have more experience by now (three times the number of years) as in my default, chosen profession: lawyer. I've actually squeezed post-Master's translation school in. But don't worry, it didn't interfere with Shogun 2. Translation and interpreting to be precise but I'm not consecutive interpreter material because I forget the first half of a sentence before the speaker finishes the second one and I drop an f-bomb when that happens when I'm not particularly focused. Besides, most of the stuff those guys talk about (economy etc.) is so horribly boring, I can barely stay awake. This is not a problem with written translation because I need to be successful (and timely), not focused. And nobody looks at my face. Anyway, the other post-Master's I've just signed up for (guess it's addictive) is about perfecting the mouthpiece aspect of the legal profession, which involves diplomatic protocol, preparing killer presentations, perfecting your speeches, captivating your audience and minding the psychological aspects of it all while so doing. Perfect for me. Probably not bad for a caster. Besides, I have a cool accent. And could only benefit from the publicity. Also, it could propagate gaming. If a client doesn't think it's serious for a qualified legal translator (and lawyer, after all) to play games, then all the better. I'll have more time to play games. And good luck finding another good one in the current market! I don't care. Or rather I care to put gaming on every single CV that goes anywhere. I do that. I have a portrait shot with a sword hilt, I use it too. But I digress. And oh darn, this is becoming so much about me while I didn't actually want to say all that and, in the first place, I simply wanted to jot down some notes from a game I was watching.
So here's the deal: I'd like to commit my notes to some form of memory. I remember things better when I write them down. It doesn't matter if I read it later. Half of the time I can't even read my notes and most of the time I can't find them to begin with. It would be the same with a file on my computer. On the other hand, a blog is a different matter and what's a better place than TL? Beats having a shiny prefabricated online diary laden with graphics straight from the Blizzard SC2 Fansite Kit (not to detract anything from the latter), hands down. TL basically is Starcraft 2. Well, not really. But largely. So no, not fishing for attention. I only hope you like the writing. I do my best to make it enjoyable for you too and make your time worth it, perhaps even enjoyable, as I would always like to do with anything I write, especially correspondence. And this is a bit like a letter. Also, I hope that the notes will be useful to somebody else than me, too. Another reason for posting this, actually.
So, without further ado (there has already been more than enough), the game I last watched was MVP vs NesTea on Cloud Kingdom (this almost feels like blazoning), G1 at IEM. I would link to the Day9 write-up (he casted the game with Kaelaris) but he's running some overhaul of his site right now and I can't access the archives (he promises new features). It's a good thing I had loaded the page with the videos before this happened but it makes me worry what if I need to refresh. Well, let's hope I won't. Wait. Have I only just said, 'without further ado?' I did it again. Pardon me. Here's what I've noted of the game:
MVP (T) vs NesTea (Z) on Cloud Kingdom, G1 at IEM
You can build a Command Centre on a ramp. I obviously knew this but knowing is one thing, remembering is another. I've known this for long; I've never used it. Waitasec, I have. Once. I was in silver league, the map was I don't even remember... oh right, Jungle Basin, and it was a TvP. I was in the south position, no, wait, I started in the north position but we base-traded. Started from a failed cheesy rush on his part, I replied in kind and bunker rushed him as he was destroying my base. I took his former nat too and eventually he destroyed the rocks and I had the presence of mind to fly the CC atop the rap and start upgrading to PF. If I had already had an orbital there, I would have lost. Something like a hundred zealots died at that PF and the guy quit. So, well, I didn't actually build the CC at the ramp but I did use it for blocking. Only time I can recall. Should do it more often. This is vastly superior to relying solely or mostly on supply depots (which are easy to bring down for baneling busts and basically any Protoss pushes, let alone an early tank push by a fellow Terran).
Cheese can be an all-in win (or an all-in loss) ...or just an opening!The key is making it the latter. I need to 'transition' from having my twoport and other shenanigans as an entire game style and plan, and start thinking about them as just openings. I still need to do some things after them and it can't be just about anything. I need to have a plan, an idea, can't really leave it all up to improvisation. Again, like the previous one, this is nothing new but useful to bring to mind again.
Day9: 'Banshees shut down slow banelings, roaches.' Not really verbatim but something to this effect. And he's right. Banshees don't have the stopping power on the ground, as in roaches can't bump into them and stop moving; they don't have this blocking potential. But they can kill roaches better than most other things can. Roaches or slow banelings. Hellions can handle most of the rest, it's all light really from Zerg at this stage (early to mid-game). MVP did sometimes have to give ground or outright run with his hellions, this is normal and okay as long as those roaches are actually killed.
Many roaches flee from just 2 thors, 2 tanks (unsieged) and a bunch of hellions. I would actually have expected the Zerg to win any potential engagement but who am I to argue against NesTea? They probably have a reason. And even if not, they still run. Useful to know. (Point being a Zerg doesn't feel so confident even with a large pack of roaches against a modest mech Terran army, perhaps roaches aren't so scary.)
You can hit destructible rocks to make invisible units die from splash damage from the hellions. Nothing major but anyway. (It's always good to kill a tumor.)
Six refineries (6 geysers) = 18 SCVs = 18 supply. This because one of the casters (it was Day9) compared worker counts and remarked that MVP didn't really need so many, only really for the gas, with four orbitals (it was three bases, though). Eighteen sounds like quite a significant portion, actually. (It was, MVP had 46 total at that point, leaving 28 for minerals, which is a number you could feed off one base in terms of minerals.)
A nice nest of turrets MVP built on the edge... just the route NesTea later tried to use with his hopeful roach drop. Why noting? Because I'd probably have built it late or forgot and never got the opportunity to fix the oversight. Those turrets need to be there early enough, not just be there at all.
On the other hand, according to Day9, maxing is what must be done at all but not necessarily early. His exact words were, 'It's not about getting maxed fast, it's about getting maxed at all.' I hope he wasn't talking about the Zerg. Just kidding. The Zerg got maxed around 14:00 or almost maxed. On roaches. He dropped them off ovvies, except... (At any rate, just get maxed, until then you simply need to survive. But I love to be maxed around 16:00 with mech, if possible. If it gets delayed until 20:00, I'm uneasy.)
MVP blocked the landing zone for overlord drops with SCVs. Those were also auto-repair SCVs. It was a beautiful hold where MVP was at 150 supply himself. Lost 38 workers or so. This was the moment a calm Day9 reassured a concerned Kaelaris that MVP was doing fine with only so few SCVs (46 remaining).
Mech is much more expensive than a Zerg composition (not a superlate one), even though both are maxed. It is also cost-efficient. It makes sense that if you want an advantage and you can only have this much supply (and exactly as much as your opponent), you can dump more resources in the army and expect better results than someone with a cheaper army. This is by no means any brilliant discovery. It's obvious and borders on dumb. But it's useful to remind oneself of such a thing.
Back to roaches, unsieged tanks still kill roaches (they have better DPS than sieged ones, just no splash and smaller range, while no minimum range) ...at 1/2 of a thor's supply count. This matters because outnumbered thors do so badly against roaches, especially with a good concave, and it's hard not to get a good concave when you outnumber the opponent in terms of number of units.
MVP doesn't actually hurry to siege up. Most Terrans would, MVP would not. I think I wouldn't either, in theory, can't say how I'd behave in an actual game. For the record, I think sieged tanks could have lost or suffered greater casualties in those situations. Also, Terrans, even some high level Terrans seem to be afraid of unsieging, even in situations when they could a-move into a sieged line (or whatever else it is) and win (instead they take time to get sieged, and get pounded upon during that time). Another problem is sieging up when you move into the range of enemy tanks that are sieged. Seems like an unconditional reflex, except it takes time and it's no good vs tanks that are not clustered together but spaced apart. For some reason, I pay attention to this stuff. Even though I do what I can to avoid tanks in my games, as a rule. Reminds me of Effka, the lady player I mentioned before. She was no MVP but she surely knew how to bust a tank line. One of her descriptions mentioned, 'coming at you in more ways than you can handle.' I don't even remember where I found her replays but they were a great sight.
Well, the above was supposed to be a concise list of points but I guess my proverbial pen was itching too much. I always get a surge of energy around 3 AM. I would bold it but then that would be a huge block of bolded text, so I will not. This would have worked with one-two sentence long items. But I still hope the points I wanted to made are visible clearly enough. I am moving on to game two.
(And I'll probably have something to say about it. I certainly already know what I will digress about.)
(On second thought, I'll go up with the bold font, after all.)
Edit: While I give links directly to YouTube, I actually access the casts for myself via Day9's Archive, which is a vault of treasures. Day9 isn't the only caster, there are basically two duos: Day9 and Kaelaris, and Tasteless and Artosis. It's very easy to hit the Next button and see the next game without having to do anything else. Impossible when watching directly at YouTube.
As I let you on before, there was some chance I would cover game two before calling it a night. Indeed, here it is. But before we get started with it, I would like to tell you why I brought up the autobiography of Boxer. Boxer started as a Protoss and zealot rushed every game. When he smartened up, he advanced to reaper dropping. Does this sound like a gifted child? And he was not a child at all. He was in mid-late teens. Well, he had played arcade games before. He became a Starcraft 1 bonjwa through practice. He put in the time and energy. Most of all, hard work.
On a different note, I am telling you now because I forgot in the previous post. I brought up the subject of Boxer's autobiography specifically to tell you this but I forgot. I ended up basically giving you the information that I read it and the deeper sense was lost. This will, in a way, play into what we're going to discuss in a while.
I give you game two. Or rather, simply, here's game two. I'm not going to analyse it or review it or anything like that, remember? Simple notes. Here:
MVP (T) vs NesTea (Z) on Cloud Kingdom, G2 at IEM
This game is very special. Most of us know why, most probably know better than I do. But for those of us who don't: Nestea is perhaps world's best Zerg (and, incidentally, one of the oldest progamers at 29, of the biggest names only WhiteRa and Boxer himself are older, he is four months older than I am) and MVP is generally seen as the world's best Terran and best player overall. This is by no means undisputed but if you were to pick a candidate, you'd probably pick him. From yesterday (I don't remember when I actually last shut down my computer fully, it has all been sleeps the last days), I still have open Ver's guide about How to improve, based on Brood War (Starcraft 1). But the methodology is similar. He stresses the importance of good sources, learning from the best you can. That would be MVP. Another guide I've read lately is from Br3ezy, A Focused Approach to Perfecting Mechanics v1.1 who mentions MVP as the example when he talks about watching the best players play to improve. Not watching replays but rather watching VODs, to see the mouse cursor at least, to see how that player does things. In this video, we can mostly only see the casters' mouse movements--already with remarkable accuracy, Day9 was actually a very high ranked player in Brood War at A+ on Iccup, which I imagine few foreigners, i.e. non-Koreans, ever achieved; with his mouse, he moves like he owns the place. But we don't see MVP. But we can still try to get into his head a bit. This is what I'm going to be trying to do with a couple of games by him. My to do list of replays currently covers Tarson, Last Shadow and MVP but I still hear those guys saying that you need to learn from the best. That'll be MVP to you. But on to the notes.
Methodical scouting from NesTea – he spots the hidden barracks in a corner of MVP's base. This tip comes from our Zerg player but it's extremely important to a Terran. Terran needs this scouting solidity both in his own base and the opponent base, especially against Toss, with Korean fourgate (proxy pylon and then gateways somewhere in a Terran's base if he stops following the scouting probe) and all, though I once had a Zerg guy throw down a hatch in my base and, unfortunately, he won that game (after all, it was a functional meta hatch).
As Day9 notes, while it's the most common to hide something spooky, it's a good trick to hide something normal. In this case, it being rax, part of the most normal build ever. But if the opponent doesn't see the rax, he could panic, start to look for proxy (lose mining time, overinvest in defences), suspect cheese, or, which is in a way the opposite, feel overconfident, thinking you go CC first (i.e. begin to expand already before building your first barracks). Especially a Zerg always faces the tough choice between getting more drones to harvest his resources, or more combat units. You normally want to force him to get zerglings to avoid his economy growing exponentially. Except that when you actually attack him, you prefer to fight drones than zerglings, naturally. But you normally don't want him to drone. Particularly for a lower league Zerg it may be difficult to get it right here.
There's that moment around 7:30 (replay time, not video time) when MVP checks NesTea's third and, when he finds a hatch there, decides he's completely safe to build himself two E-bays. (I guess a three-basing Zerg won't have the active power to punish you for that.)
While the Zerg has supply lead early, around 10 minutes MVP catches up on supply despite NesTea's relative freedom and peace, no significant engagements etc. Later on, MVP will actually outplay NesTea economically.
Tiredness: NesTea the Zerg, not MVP. It was tiredness and also jetlag. While NesTea is a good multitasker according to the caster's, apparently his perception or orientation was hit and miss to the point he barely even qualified in the earliest rounds, being the best Zerg of the world! This is the state I often play in, actually. More often than not, perhaps almost always. As a concrete example, NesTea missed drops. Things like that cost you a game. It's important not to insist on playing when the conditions are so subpar, if possible. Not really possible for me (and one of the reasons my play is so erratic) but if it's possible for you and you don't mind me saying, don't pull all nighters, don't play that one last ladder before going to bed. Get up in the morning instead. I played Warcraft 3 around 8-9 a.m. before leaving for university during my student years. Incidentally, this is the same reason I forgot to tell you about Boxer's beginnings. If I were playing a game, I'd probably miss SCVs, be late on timings, react too late to attacks.
Even MVP makes mistakes/can't manage everything perfectly at the same time, as evidenced by lings literally walking up and killing three hellions near his base, as well as a couple of different occasions later. Although at this level you could actually suspect that a player might actually want to sacrifice something on purpose in order to give his opponent a sense of false security and induce him to let his guard down (which I did myself in Warcraft 3, perhaps already in Warcraft 2), MVP does follow up with more apparent mistakes in this game, notably around 15:00 when the banelings manage to connect and he doesn't lift up the marines. Despite being described in these words (by Liquipedia): 'Mvp is regarded as an extremely solid all-around player, who relies on excellent macro management combined with impeccable timing and superb unit control,' he is entirely capable of not controlling something, and not capable of controlling literally everything. He's not a machine, no progamer is a machine. Neither is a ladder opponent. If even MVP makes mistakes, everybody should be able to get away with some of them on ladder. For the record, one of those mistakes MVP immediately corrects: a while later he does lift up the next pack of rines to be put in danger when banes roll in.
Quick decision-making: MVP doesn't get stuck when he needs to make the decision to leave some marines behind due to not having enough medivacs to pick all of them up. He fluidly escapes with two full medivacs while leaving seven marines to ling/bling before infestors catch up (I couldn't really sense a moment of hesitation). There's no point thinking about losses you can't prevent, MVP doesn't do it. He especially does not lose two full medivacs of marines (16 marines plus two medivacs) due to getting stuck feeling bad about losing seven marines. (From what I've read, it is important for pros not to get hung up on things like that, they need to move on with a clear head.)
The thought that makes the (effective) APM: Around 14 minutes, NVP unloads a large group of marines. Before the last of them hit the ground, the first are already stimmed. This is happening with the D key trick (Medivac moving while unloading units, instead of standing in place, this is achieved by pressing D while clicking the move order). This entirely within a newbie player's mechanical ability (mouse skills). It doesn't even take exceptional speed. All you need to do is think about things to do. Come up with things to do. Do something while waiting for a different thing to finish, not after. Switching between tasks instead of queueing. Can you imagine me waiting until medivacs reach their destination, then unload over a dozen marines in one place from totally stopped medivacs, then selecting them and stimming them all only when all unloaded (after perhaps making sure I had indeed unloaded all)? I certainly can. (Which needs to change.)
HSM in the final moments of the game. No comment. Seriously, though, if a Zerg of the caliber of NesTea can't predict it accurately (16 ravens got made without his knowledge, even despite MVP's SCV and possibly marine sacrifice to clear supply), a ladder Zerg shouldn't really either. It seems to be only two ravens running up, getting successfully shot down by corruptors (rather many of them)/infestors before firing the heat-seeking missiles, at which point NesTea obviously knows what's going on, except those other ravens from the side come as a surprise (about 16 total ravens for MVP). NesTea doesn't even manage to split up instantly or focusfire those two ravens down (I suspect a former WC3 player might have but I'm not 100% sure) but he really can't do anything about like fourteen of them flying in from the side, many of them probably having enough energy for an HSM. (It's just a bit Zergs to tend to know not to gang up too much, some actively split their fliers.) For the record, it's vaguely possible to go all raven against Zerg (higher league than mine but not really GM).
Day9 dissected MVP as going for precisely two-prong dropd with marines and medivacs, where tanks allowed him to stay on the defensive, while buying himself time to get more ravens (in this light, it makes sense that at least one of the attacks didn't even appear to be particularly reasonable; it makes even more sense if you consider than he would eventually need to free up supply to make ravens anyway). Obviously, buying time is good with ravens, they need 125 energy for the missile, which isn't fast to farm. I once had a TvP build that made a huge point of that. It was Olorin's one base marine/banshee/thor/raven push vs Toss. Almost the entire timing of it concentrated on hitting 100 energy on the raven for the PDD at the most important point (that being the most significant confrontantion with stalkers). Even my own TvZ includes some time-buying devices. I'll tell you about my TvZ tomorrow perhaps.
(Again, this is not an analysis of the game: it's my notes to help my learning. Things that capture my attention. I am hoping for them to improve my game-sense and to give me an idea of what to do, which closely relates to mechanics. My average APM is about 1/10th of what I can deliver when I simply need to be issuing commands. If I start coming up with better things than the above, it will mean I have improved. For now, at my skill level, I am coming up with these. Also, the more effort I put into my notes, the more I will gain from them. Naturally, I'd be delighted if someone else could gain from them too. And it's not about these precisely things that I mention: they are ultimately only examples, it's more about the ability to find them, or rather generally things like them.)
(On a different note: EMP + HSM vs Toss? Impractical but sounds attractive, hehe. But could be used in those catch-all compositions where you need both ghosts and ravens. You know, my experience with anything close to a successful late-game TvP is mostly as a spectator. )
There is one final thing I'd like to turn your attention to, which is not actually connected to gameplay. Notice how, near the end of the video, the host would like to ask MVP some questions, MVP would presumably like to answer them or just simply say something, but there is no 'translator' (a person who 'translates' speech is an interpreter in proper English, by custom this is not 'oral translation')? Learn languages when you still can (and you always can, except not always with such ease and such promise of success as at an early age), use every opportunity. In my case, it got me a living when there was a massively bad market situation in the field I actually got my degree in. Translators and interpreters are made by three things: fluency (and easy and confidence communicating in a foreign language), correctness (this one requires some study but not necessarily an MA in translation or linguistics) and the knowledge of a field or more fields. I waited for somebody from the audience to stand up and fill in but nobody did, I guess they didn't have a confident enough speaker of both Korean and English there. But the same shouldn't happen, for example, in any part of Europe with English, German, French, Italian and so on. I believe the gaming scene gives a lot of opportunities to practice things that can be useful in the future to a translator or interpreter. Myself, I owe a lot to gaming in this regard. You wouldn't believe how all the vocabulary from games is proving useful at every step, offering a convenient retroactive justification of all my years of gaming while at it . (Not to mention live language, something people with relevant MA degrees are often lacking in.) Also, if you find yourself facing an opportunity like that, take it. It doesn't matter that your grammar isn't perfect, that you may stutter, that you won't catch all the speaker says (at 15 or 16, I acted as an interpreter for about two hundred people without preparation or warning, and I was far from proficient at that point--and that was my most successful attempt of all). Grasp it, and absorb the experience. You never know when it could be helpful to you later in life.
Okay, really last thing, I promise: I'm sorry if I sound as though I actually were good at the game or showing off. I'm not. I just have this manner of communicating. And when I explain the obvious (e.g. spell out what it means to go 'CC first'--for the same reason I put links on names and documents), I only do so because some of us might be new to the scene and I wouldn't like anyone to feel lost, not understand what we're talking about and so on; I would be extremely delighted if any of this actually proved helpful to anybody in his Starcraft 2 journey or RTS journey in general. Everybody else (most of us) will please excuse me.
Years ago in the European scene, especially back when I played Warcraft II Battle.net Edition, a 'pro' was not understood the same way as today. While it had implications of tournament play and it was understood to be a bit of an exaggeration, it referred to what we would call a recognised gosu player. In fact, the term was frequently and blatantly misused, being a mixture of wishful thinking and admiration, depending from which angle you looked. A ladder hero would certainly be referred to as such, while somebody who didn't have gosu skills couldn't hope for it no matter how much he got paid for playing. This old understanding has probably been deprecated. I say 'probably' because I have not followed the scene since 2008 and even then I was mostly sketchy, with only first-hand experience. Kas (EmpireKas) would be a 'pro' in both understandings. His ladder account on the EU server shows 13468 games as of this writing (Saturday, Aug 25 12:53am GMT (GMT+00:00)), and he has been in grandmasters ever since the league has been introduced (his only finish in master league is Season 1 because GM started in Season 2). In his matchup history, the oldest game you can see before you need to click, 'Show more entries,' is from yesterday. Apart from being a ladderer, he is also of special interest to me as a former Wacraft 3 human player. And as a European perhaps. In IEM a week ago, he played this game:
NesTea (Z) vs Kas (T) on Metropolis, G1 at IEM
And here are the notes:
Kas goes macro, he won't punish you, he will have later timings, perfect mechanics and non-stop reinforcements when he gets there, according to the casters. But even he can go mass banshees these days. Also according to the casters. This game he picks the former option. This is going to be a macro game. Well, you already knew when you saw the length of the video anyway (around 40 minutes).
Equal supply happens around 13:00. (Wonder if I shouldn't do some come comparative research here. I'd expect the equal supply moment to be a good moment for a battle, especially since I'd expect Zerg to have more drones and I'd expect a Zerg army to a bit less potent than an equal supply Terran army, barring the application of infestor abilities.)
NesTea is forcing Kas to attack him. NesTea is a reactive player, Kas is a safe player (had to develop a safe build in all that laddering where he didn't know his opponents). This is a safe vs safe battle, sort of.
Kas has nice wall-offs with depots in front of bunkers (not difficult to do but not everybody cares), there is no place for a surprise attack to happen anywhere (which is important).
In spite of heavy losses in the mid game, Kas defends off his sheer macro until about 30:00, when supply is equal again. This is what being able to produce tons of units does for you.
At 32:26, there is a nice HSM killing all the banelings that are rolling in. So much damage averted from the Terran army! And this is apparently doable.
NesTea is no pushover himself when it comes to macro, and has many bases. Both of them are broke before or around 40:00.
Just like MVP, Kas goess mass raven in the end. He had been massing vikings before. Unlike MVP (who didn't get a chance), Kas throws down plenty of auto-turrets, which takes a large part in actually bringing down bases (destroying buildings).
NesTea (Z) vs Kas (T) on Atlantis Spaceship, G2 at IEM
He goes CC first. Well, this isn't ladder, after all. (I wonder how he'd survive a sixpool.)
While he later gets praise for his great macro, he doesn't seem to know right away that three marines won't score any kills against 2 queens, he realises in the middle of combat, after losing 1 marine.
Kas doesn't get medivacs, he spends more on units doing damage. This is actually a bit strange for a WC3 human. As a WC3 human, you sometimes literally had more priests than footmen. (He makes up by stimming wisely.)
He relies on stimshield rines and just a couple of hellions for survival. The hellions being so few, they give an impression of being meant for kiting, not for actual damage dealing.
11 minutes and they are on equal supply. Looks like fights around this stage of the game have a tendency to speed this up, which means they must be bad for the Zerg.
Around 15 minutes, Kas loses an entire attacking army. This is probably the typical Terran death moment, with people not macroing well behind the push. Half a minute after the battle, he is actually back up to equal supply with the Zerg. I really ought to take a closer look at his replays and all the macroing that goes on there.
Thor drop in one battle with comfortably placed tanks behind (plus marine buffer). Another battle, a siege tank dropped behind a gap. Making a mental note. Need more imagination. Imagination is a huge limiting factor.
Around 25 minutes, it shows that he was basically playing hall drop while herding thors in the background. This does look like WC3. I think I spent a year busting halls. And it's good to see pros herd thors in the background too, I almost feel home. Except I'd never really dare move a mass of thors out without at least some hellions for support. I used to but it didn't end well.
In fact, around 26 minutes he actually moves out with thor/marauder/medivac (a lot for Kas)/viking. Strange composition. I can't see in the cast whatever NesTea is sporting right now. Well, in the end it seems he wouldn't have needed hellions. Well, there's a reason he plays those tournaments and I don't.
Now around 27:45, Tasteless (brother of Day9) says thors don't normally beat broodlords but apparently they do. Well, no wonder with like 12 thors vs 4 broodlors but I can tell on my own that thors beat broodlords in less favourable proportions. Especially when you have hellions hiding behind thors to deal with broodlings, without suffering damage. Especially when the thors have +3 armour and the hellions +3 attack plus BFH.
Bottom line: Kas outmacroed NesTea again. He does the same to Protosses but I just can't get his 'Very Easy TvP' right. Building a third at 4:20 against a Toss on ladder, yeah...
Hi again! As you see, the posts are beginning to resemble normal notes more and more, with less narrative and more bullet points. This is how I'd like to be. I'm still looking at IEM games and here are the new ones.
(Probably everybody knows but Bomber is a very strong Korean Terran. He's particularly good with macro-based builds. Slivko is a rising Russian Zerg.)
According to Tasteless, Zerg can't do much on Ohana in super-late game (due to there being only five bases for him to take).
Bomber has a token supply lead around 8:00 (but Bomber is this type of guy, like this TvZ build).
Combat shield finishes right on the verge of creep for the marines. Seriously, did Bomber actually know how long the marines would take to travel there and time it accordingly?
Around 9:00, when attacked from both sides, Bomber first fights against the spine crawler, then turns around to face the roaches after driving the crawler to half HP. There could be some improvement in just making the choice and sticking with it, except it's better to change your previous choice if it's a significantly worse one than to stick with it for the sake of consistency, but it would be great to be able to make the right choice right away. With approximately equal choice, it would be better to make even a coin flip choice and move on to execution (indecision costs you still more than a slightly inferior choice that you make in a split second and execute right away and stick with it)
By 10 minutes you already have two E-bays fast at work for Bomber plus stim about half-way done.
According to Artosis, Slivko was a bit confused because the expected hellion timings didn't materialise, so he didn't really know what he needed to counter whatever Bomber was planning. Personally, I've had non-standard hellion timings in TvZ too, or especially non-standard building placements (esp. in terms of addons; a reactor factory is more obvious than a naked factory, and a naked rax at a blocked ramp is ambiguous).
Around 15 minutes, it shows that despite a bad engagement, Bomber's in a better macro position and actually takes a large supply lead just after losing that engagement (so good is his macro). I see a pattern in this IEM's TvZ. Optimising your mining and production seems to be so popular and so powerful among the IEM's Terran participants in TvZ.
Bomber doesn't do any creep control (creep control is the basis of TaeJa's winning style), he scans to see what's ahead. Some players do it like this.
Around 17:46 Bomber loses an engagement despite having supply lead, because broodlords force him to unsiege, he generally doesn't have many tanks, and infestors and banelings can mop it up. (However, his macro will eventually pull him out of even this trouble.)
According to Artosis, the Zerg army is hard to make cost-efficient with broodlords and infestors. Which is useful to know.At any rate, one can't just think about his current army composition in the abstract. Have to feature reinforcements in!
Bomber (T) vs Slivko (Z) on Daybrak, IEM G2
Bomber sets up a bunker rush while sending hellions behind it (timing is around 7:00). Unfortunately, hellions don't arrive before the bunker rush is killed off. It would have been rather powerful otherwise. Personally, I've combined both in TvZ but sequentially. Not like aiming to have both the bunker rush and the hellion push going at the same time. Which would be very cool to have.
I didn't really like retreating the two hellions on a path leading through queens instead of pushing wherever the drones were hiding. This even though two other hellions were coming. It's hard to manage two very small groups of hellions but it's probably equally hard to manage drones etc. against them.
Bomber goes two banshee. That's also how I do it. But I wonder if it wouldn't have been better to split them up and harass two mineral lines. One-shooting vs forcing the Zerg to split attention to control damage is a hard call.
Interesting note by Tasteless: Bomber doesn't know how much of a supply lead he has (or any). So he doesn't know he's 100 supply ahead. More like 120, actually. He's cautious as the Zerg may have resupplied. People who are cautious when being in the lead look overcautious in replays or casts but people who assume a lead sometimes lose the game.
Still a close game. HSM and PDD mattered a lot.
Bomber (T) vs Minigun (P) on Cloud Kingdom, IEM G1
(Minigun is a young American Toss that plays LoL in his free time. Probably does good to his micro.)
Good decision to step back and kite a bit instead of continuing to fire at an almost killed Nexus.
Steady infantry production seemed to be enough to deal with a gateway army, i.e. stalker/sentry(FF) and some zealots.
(Yeah, this is all I could come up with.)
Bomber (T) vs Minigun (P) on Ohana, IEM G2
So little happens in those TvP games for Bomber compared to his TvZ early game.
According to Day9, Bomber's timings on stuff are fast. Might be a good idea to take a look at the replays to see what he builds when.
Bomber places spotter supply depots around the extremities of his base. This seems so easy. But one thing is easier: forgetting that you can do it!
Newsflash: Tosses use observers as spotters for drops.
As Day9 says, it may be a good idea to 'invite' a drop (then perhaps corner it, or attack while the unloading is in progress), instead of chasing it away.
Bomber had 2/1, lots of units, invested everything in the attack (means he all-inned, basically ) around 16 minutes and won. My personal struggle with TvP is that I can't win such attacks, not enough stuff or not good enough execution. Or wrong planning. Something just goes wrong for me. But maybe if I actually played TvP the standard way I'd catch up after some time and it'd be as comfortable for me against my matched opponents as it is for Bomber against his opponents.
Never any direct damage before the one and only battle, strong macro up to that point.
(Don't tell me you don't know Grubby. Everybody knows Grubby. He was the closest Warcraft 3 had to a bonjwa.)
Grubby doesn't scout. According to the casters, Bomber always goes fast expand anyway. But here Bomber goes four rax instead of three.
Bomber seems to have won on decision-making and surprise. Wonder how much of a gamble it really was.
Bomber (T) vs Grubby (P) on Ohana, IEM G2
Bomber puts just 1 marine up overcliff to find out what's going on there, the rest remains downcliff.
Grubby war prisms (basically: drops) while Bomber is pushing.
Grubby defends thanks to that 9-gate style (SaSe) where you warp into your base when attacked. One needs to be prepared for a Toss doing that. Mass gates aren't so rare on the ladder.
Vikings as spotters and scouts.
Hey, ghosts are quite cool when there are no more templars (not only against HT)! They can just EMP the army units before an engagement... (As in, everybody knows this but still.)
Bomber got his good moment to EMP a good number of templars, as well as smaller-time moments to snipe some. As bad as storms are from a T perspective, I guess TvP could yet be an exciting matchup. As good as Grubby is about unit control (and a Warcraft 3 bonjwa probably has it well ingrained to watch out for spells), Bomber still gets a window or two. Perhaps PvT isn't as flawless as one would think.
Killer (Z) vs SuperNova (T) on Daybreak, G2 at IEM (was 1:0 for Supernova, G1 didn't get casted)
(Cast starts around 15:45 into the game. According to Kaelaris's introduction, basically greedy expands for both players, including a third at 6 minutes for Supernova. Constant aggression from Supernova in the form of hellions (BFH) with a large body of thors added in the mix, but you can also see some tanks and cloaked banshees at 16:18, which is a good moment to see the composition of Supernova's army. And a raven already. Roachfestor for Killer.)
Standard but thor-heavy mech for Supernova, with banshees. There seems to be a tendency among pros to reduce the number of tanks in favour of more thors. I actually have gone further for a long, long time. My own TvZ mech actually doesn't involve tanks. And I can live without them unless the opponent goes like max roach. While I don't have a nest to fall back on like Supernova has around 19:30, I have more thors, which usually works just as well (and is way more convenient in the average situation). Supernova has more toys but he doesn't have 3/3 upgrades on the thors when broodlords come, in the battle around 20:45. Hellion/thor with 3/3 upgrades can actually handle that type of broodlord push pretty well. (And all the better when the Zerg makes corruptors and you don't have any air.) The trick is that you leave broodlings to hellions and focus-fire down a broodlord after broodlord. Works like charm for me. Easier to beat than a 100% roach 'composition', actually.
Supernova doesn't finish off the almost dead spawning pool in the back (which would be an inflexible fixation on target), he takes advantage of the opportunity to pursue drones. Taking advantage of opportunities and reacting instantaneously to them is important. (It is also extremely subjectively important to me, and I emphasise it more than the making and execution of plans.)
Supernova scans ahead before committing his army.
Supernova rocks for having like 80 SCVs to 60 drones.
Supernova should probably have pulled some of those mules to repair the thors around 23:30. I'd do it right when I saw those roaches, personally. This hurts your econ but being dead hurts your econ more. Just pull those workers right away.
Banshee/viking/raven attack groups. Yup. Good old Polt. Polting a Zerg is so cool. In fact, there are more TvP strategies that work great against Zerg, e.g. Olorin's Thor/Banshee/Raven/marine push. I wonder how many classic TvP strategies could be transplanted into TvZ and used successfully.
Looks like an overall very successful Terran mech-air that uses the strengths of every unit.
Supernova (T) vs MC (P) on CloudKingdom, G1 at IEM
(MC is the highest earning SC2 player and a tournament bot if you didn't know. )
SuperNova is capable of beating MC, he already did to get there. But MC is a tournament winbot who doesn't get stage freight. He's probably more comfortable in a tournament than I am in a 1v1 vs Very Hard AI. That's actually a good living plan, come to think of it. You run in more tournaments, you get better at specifically playing tournaments and you win more money, and then you win even more because of how good you are at specifically playing tournaments. So much cooler than practicing law. Oh wait, that guy is in law school. Oh the lolz. Figures. (Just kidding, he actually bought a car for his mother with the money he earned. Respect.) MC, I hate to break it to you but playing SC2 is cooler than working for a law firm. Seriously, I know what I'm talking about. I've heard he wants to become a judge, well, I guess that's cooler than a law firm from some angles.
According to Tasteless, MC is good at, 'sort of imposing his will on the game.' Wonder what this means for SuperNova (reacting? getting slightly cheesy, as in relying on surprise? feeling like he has the system against himself?).
According to Artosis, SuperNova is one of those players who do mix some tanks into a TvP army, though not always. Apparently, I wasn't stupid for thinking about it, even despite the patches. Back in the time, Jinro delivered some Toss pwnage with MMTV. It worked for me too in silver league until something got patched or the metagame changed or something else happened. In this case, there are just some tanks in the early siege push. This is actually done on ladder, one of the popular pushes.
Double gas fast expand proxy shees against MC. Dude's a baller. (Bunker + tanks means he's also safe while doing that.)
According to Artosis, if you catch the Toss with two cloackshees after his first observer, you should get at least 10 probe kills before the second observer comes out. You don't know when the obs comes out but waiting for a second banshee increases the chance that obs will already have been made. Also, two shees one-shot a worker. I always use them in pairs myself.
SuperNova spotted a spotter obs (a ripple in the air) and scanned and killed it. Means he was looking at the army when it was moving (the ripple obviously won't show on the minimap), even though no battle was going on. I would probably have been looking at my base and missed the situation if I had been in his shoes. Ambushes on the marching route are probably a not so infrequent cause of death, I guess. Especially for armies with tanks.
This is extremely obvious and I think I remember it pretty well but: a shee must move. I'd have kept scoring sure kills on the probes at the first mineral line, actually, but apparently SuperNova moved it to make room for the second shee, to push two lines at the same time in effect. That second shee, despite having only 3 kills, succeeded at shutting down/preventing the mining in MC's nat. Also, no shees lost. And a 10 SCV advantage over probes.
Banshees harass the mineral line instead of plodding along with the siege push.
Around 14:20, once again SuperNova instantly stops attacking the target almost the moment he sees it's not the best for him, even though the target (a nexus) is firmly in the red, almost dying. He gets that nexus, not even a cancel, a moment later. Perhaps there would have been a last second cancel if he had kept firing at it before.
Storms do kill off SuperNova's army but his drops are successful. I'm not sure I'd have gone for assimilators in his place, I'd have targeted pylons instead, I think. A nasty Warcraft 3 habit. I even had that human vs orc rush where I'd kill the supply burrows and avoid killing workers all the while I built farms and more army for myself. I must implement this concept in SC2 some time ('Where are my pylons? And why is he not targeting my workers? Wait! Wait!'). Then again, if you have a worker advantage, it's probably reasonable to keep making it bigger.
SuperNova faces a problem with disparity of upgrades, while having supply advantage.
Not microing out of storms is not good, even if you feel you can take it.
According to Tasteless, you should basically always know there's an observer on your attack path and just scan and kill it. Sounds like it, especially after attacking down that path so many times.
SuperNova has a heckton of medivacs, which is different from the Kas style. I used to play TvP with having a lot medivacs in my army. I somehow preferred that style over more infantry and fewer medivacs but then I had periods when I preferred the opposite style. With the healing power SuperNova can hold a lot and isn't even microing... but with microing that would have been better, more units would have remained alive, victory would have been faster. I suspect if you force yourself to micro even the battles you're sure to win, you thereby improve your micro a lot, get it more under your skin (push it into the subconscious), train the habit of always being alert etc.
Back to those banshees, I can't believe he only got two. And they were so successful. I'm sure I'd have built more, especially if they'd been so successful. But then their successfulness might not have been that much bigger with four banshees or more, and banshees are a bit like hellions: you don't really need so many. At some point, it's losing its cost-efficiency. But cloack still feels expensive for just 2 banshees. For the record, those particular two banshees might as well have done without cloack, I think. I think they never really took any advantage of it, and cloak costs 200/200, which is expensive early on. I often skipped cloack when I went banshees (and it's so sweet if the opponent still invests in detection anyway; actually, some guys fake the research, as in they show a working tech lab to the scout).
According to Artosis, SuperNova stayed in MC's space most of the game and he generally plays like that, sometimes losing when he overcommits to aggression. I remember Koreans having a reputation for that kind of thing in late Season 1 when I bought the game.
Supernova (T) vs MC (P) on Daybreak, G2 at IEM
Quick CC into double rax double gas.
Scouting for pylons around 5 minutes.
Artosis says banshees keep the opponent in base as a minimum. That's what they're for. It helps to keep your opponent in base when you're maxing on battlecruisers, doesn't it? Just kidding. But I've actually really done that sort of thing. The sweetest part is when you have like 3 functioning bases, already 2/2 air and 5 working techports. Just saying. Worked in silver. Most of the time.
Oh the drop shenanigans. Around 13 minutes SN prevents charge from finishing by killing the archives, then nexus at main, then cybernetics core. He dies to storms but the damage is done. He also kills the nexus at MC's third. The armies used don't look impressive. It's the damage they're doing in all those places.
One drop could successfully land because the obs was busy watching the banshee in the corner of the map.
Apparently, SN can operate two cloakshees while building an army while calling down mules while expanding. Well, I can't. But I'd be cool to do at least the majority of those (production is probably the easiest to combine with harass control but a screen jump to control mules could take as little as one second, so anybody should be able to fit it in; if anything, it's the deliberation you can't afford, execution of any of those ideas you can).
On August 28 2012 04:23 _Search_ wrote: I'm master league in KR and NA and I find the Korean Terrans do a number of things that make them much scarier:
- They don't always fast expand directly at their natural. When I scout with my overlord and don't see an expansion down I start going into anxiety mode, dropping spines and delaying tech. Then the fully completed OC drops into the natural, losing only a few seconds compared to if the Terran had built it right at the natural. And the other half the time they're actually doing a risky one-base rush.
- They know what sucks about Zerg and abuse the hell out of it. Zerg can't shoot up, so TvZ early game is all about using medivacs and banshees to mess up the Zerg economy. Zerg needs to see to survive, so preventing creep spread and sniping overlords kills Zerg's chances of survival. Zerg has no ranged units, so Shakuras Plateau and Antiga Shipyard are basically auto-wins for any Terran willing to exploit the cliffs. Zerg falls hard to unexpected tech, so showing up at Zerg's natural with 16 hellions (after preventing any sort of Zerg scouting) is an instant win for a zerg with bad sim city and no roaches. Zerg can't kill units being healed by medivacs, and they definitely can't kill marines in mineral lines, so the game for a lot of Koreans is to just get to those sweet spots where marines are invulnerable.
- Korean Terrans play very risky with their medivac drops. They lose them constantly but keep doing them because they know it just takes one good drop to end the game. Once those marines are sandwiched in the minerals they can't be killed. They also do multi-pronged drops while clearing creep with their main army.
- They never. leave. games. Shoot me now.
- As one mega bio/tank push is being swarmed the second is on the way. Day9 said Terrans generally wait 2 minutes between marine tank pushes. Not Korean Terrans. I have lost games because I literally did not have 20 seconds to morph banelings during 3 solid minutes of a constant stream of units. Ahead, behind, Korean do NOT care. They will keep attacking until someone dies.
- Koreans tend to trade a bigger economy for constant aggression. They tend to take their third very late and their fourth even later, if ever.
All in all the best way to describe the difference between Koreans and foreigners is by describing each's idea of a perfect game. To a foreign Terran the perfect game is getting away with taking a ridiculously greedy third. To a Korean Terran it's winning in the first 6 minutes. Zergs aren't scared of greedy Terrans, but they're terrified of risky rushes.
(It was noted in the thread that while Koreans are better than everybody else, which we all know, Korean Terrans in particular seem to be like two leagues above foreign Terrans. Speculations arose as to why. Personally, I was inclined to think it was the mechanics but then apparently not only. And not really the fact they practice more, either.)
(Thanks to Search for kindly letting me paste his entire post in here.)
Edit: My thought about drops is that after a million fail drops those guys finally learn doing good drops. They just do it until they do it well, not start doing it when they are already good enough to do it well. I suppose if you do like one million gamble drops, you can finally at some point separate the good ones from the bad ones and know the timings and other windows.
SN goes CC first. What's with those Terrans going CC first in TvZ and surviving it? Seems it's difficult for a Zerg to punish without blind-countering it. Might be useful to now. But I'd expect blind counters to be frequent on the ladder, in the form of relatively early one-base pushes. Maybe I'm exaggerating.
Equal supply around 10:30, like in most games.
Turret placement tip from Kaelaris: no glaive bouncing damage from mutas if you place turrets as a chess knight would move (i.e. L-shape).
SN drop around 14 minutes. I don't understand why it wasn't a D-key move-drop. Perhaps hoping to be unseen and move everything in at once? But you can see it on the minimap anyway.
SuperNova (T) vs Vortix (Z) on Daybreak, G2 at IEM
Multi-cloakshee harass can be really efficient and allow you to mech up behind.
Both hellion and banshee harass can actually be kept up even if mutas are on the field. I only wish I knew how to do it and not die. I guess I'll need to start from doing it, while not dying will have to come later.
Sieging late is bad, just to reiterate. Just siege already if it seems the Zerg will break through to you. Insta-sieging is kinda crucial with tanks. Why I don't like tanks, 'cause I'm slow.
Six banshees are guaranteed damage that a ground-pushing Zerg can't really do much about, according to Kaelaris.
According to Day9, ZvT ultralisks against mech are only good if the Zergs does a lot of counter-attacking. I don't know what this means but I guess it's a wise thing to say.
Hellions wandering around the map do more damage than hellions sitting in base (except you don't want your mech to be caught hellion-less by ling, so this is kinda tough).
Around 20:30 a banshee keeps killing drones without moving an inch when mutas come. Perhaps SN just wasn't there or perhaps he made a decision like that. But many people would have executed a futile attempt at escape, failing to kill what drones they could still have killed if they'd kept doing it. This is something I've been paying attention to.
Day9 summary: Four siege tanks held back a roach counter-attack but twelve siege tanks had died because they were unsieged. Wicked proportions.
Supposedly 3/3 thors, and generally mech, are so great but those ultras actually still did a lot of damage around 28 minutes, almost killing the mech.
Vikings take long to land, just land them already if you're going to do it anyway.
In that engagement around 30 minutes, I'd prefer to get my back on the PF instead of putting up a long line of tanks. I personally hate long tank lines because you always know the first tanks will die, so it seems quite artificial for a war simulation. I prefered more concentrated defence. On the other hand, concentrated defence may've been just eaten by the Zerg in the situation, hard to say.
SuperNova (T) vs Vortix (Z) on Entombed Valley, G3 at IEM
Actually no CC first (1 rax FE).
Against roaches and speedlings, Day9 believes marine shield would be devastating. Maybe I should force myself and try bio or M(M)(M)TV vs Zerg once again.
No third for Zerg = bunker time. Hey, gotta memorise this one!
Around 10:00 is why I never go hellion/marauder, like not certain enough to work.
This is obvious but you can't really kill hellions with roaches. A nice person in the Strategy forum noted that both hellions and banshees were units that could easily retreat to cut losses. Yeah.
Semi-random thought: has anybody ever tried to load infantry into a medivac to chase wounded, escaping roaches (medivac speed is 2.25, roach speed is 2.5)? Perhaps I should try.
The 99% e-bay delays Toss nat at the price of about 30 minerals plus being unable to use the entire 125 for a time. I should probably try it, if only as a means to learn doing more things in a game and be more active.
MC fakes the probe ready to expand. Note to self: Don't jump on scouting info, it can be fake if it's too obvious. I've faked stuff like that too.
Another note to self: behind those pesky little stalkers the Toss may well be expanding like mad and not making units.
Yet another note to self: check third, you never know. Especially when there doesn't seem to be a big army.
Very quick max, gotta see rep, perhaps steal the build.
Attacking separately with reinforcements in a second bioball instead of balling together looks like a good idea.
Not overcommitting with a successful attack: Kas safely kills the fourth after killing the third, doesn't force his way through the nat, so he avoids throwing away the victory.
Love the EMP on colossi while vikings are sniping. For a moment I thought Kas was a-moving into a bunch of colossi and losing that, silly as this may make me look.
According to Day9, the Terran E-bay in the Toss's nat makes it obvious the barracks are delayed and so Toss can be sure there will be no push.
What good micro can do. Around 26 minutes. Stepping back a bit, microing out of storms, getting a good concave, you can actually get yourself out of a situation that looks really bad. Just don't lose your spirit like that.
Can't really do much without room to fall back on and kite. Can do a lot with room to fall back on and kite. Solution: remember to get room, practice kiting.
Don't overmake vikings like 12 v 2 colossi (Tasteless).
MC (P) vs Kas (T) on Atlantis Spaceship, G3 at IEM
CC first means you must go 2rax after it (Day9).
Map doesn't look 3CC friendly to me.
Late SCV pulls lose you bunkers. It pays to have a group hotkeyed.
It seems about the same time a drop killed robo and the army killed the CC in nat (started about 13 minutes) Kas's own third goes up behind it. Dropping while expanding while killing his expos is good.
Day9 says (+/-) drops are effective and all but not as much as more direct pushingin this type of situation.
Day9 says the no-medivac stim for Kas (around 17 minutes) in his third is a mistake. But Kas wins. Wonder who was right. Not sure Kas would've won without that stim.
Sometimes force fields actually protect you from Toss's own zealots. (Lol, yeah, noticed that before.)
It's curious how Kas first had vikings when he needed medivacs, then had only medivacs and like a couple of vikings when he needed a lot of vikings. Ignore prism, improve army, just go kill him or at least his expos instead?
MC (P) vs Kas (T) on Cloud Kingdom, G4 at IEM
Dunno, carrier imba? (That's what MC said, hey!)
As a side note, the interpreter is a great guy. I like how he's confident, nothing seems to set him back, he just says, 'I didn't get you,' and the host repeats. Those can be really stressful situations, you know.
VORTIX (Z) VS FORGG (T) G1-G5 at IEM (very tired when watching these games, perception bad)
A couple of hellions with just one banshee can do a lot.
Micro is extremely important to hellion/banshee harass. Never giving up, always trying to improve it, is important to that micro.
Banshees are good hatch killers. Gotta be careful and control approaches to your base/nat/third, zerglings popping out of nowhere and getting a surround when you're busting down destructible rocks is not good.
Violet (Z) vs Bomber (T) on Entombed Value, G1 at IEM
Lol at the scary fake 2 rax opening.
Check third. IMHO, whatever you suspect your opponent might be doing, just check it. If it's important enough, scan. Mules don't matter when you're dead.
8 CCs at 15 minutes...
Infested terrans landing just at your sieged tanks => unsiege tanks.
Don't overbuild vikings.
As long as you're maxed, you're in a pretty good shape (Day9).
Plenty of macro OCs = good. Enables the much needed production vs Zerg.
Slow mech army can't deal with Zerg attacking in various different places at once (Nydus and all).
Violet (Z) vs Bomber (T) on Daybreak, G2 at IEM (& G3)
Nice banshee flying around the outskirts of the Zerg base.
Nice combination of sieged + unsieged tank to avoid friendly splash.
Vortix (Z) vs SuperNova (T) on Entombed, G1 at IEM (new one, quarterfinals, they did play before in this tournament)
Continuing to damage the hatch for when hellions and banshees come.
Nice marine micro and hellion micro at the same time. (Likely kinda makes the difference between successful pressure and lack of time sometimes.)
Depot wall on the wide ramp (to the entire larger plateau where your bases sit), concentrated defence at the wide entry, no individual defences/best defence = offence kind of thing.
Getting away from the Zerg base just before speed when harrassing.
Cloakshee doesn't chase a runaway queen, gets back to drone hunting. I really love that type of decision-making, especially when it's correct.
30 supply advantage when those zerglings finally got those hellions in the minerals. (So where are the thors? )
I don't understand running away with an entire army from a couple of broodlords, where especially taking losses in the escape attempt is a given. I would have stayed and fought. SN had to fight with a very similar composition anyway, at his base ramp, where some of the thors couldn't even come into range. IMHO he should've just stayed, focus-fired the broodlords, fried the broodlings with BFH. PM me if you disagree (you can PM me to tell me you agree with me too if you want).
On September 07 2012 16:24 Ryhzuo wrote: Very very long read, took me bloody ages. 5 star for you.
Also hope you're still gunning for platinum. Maybe hope back on the ladder for another go when HotS comes out? I hear the warhound is pretty good.
Thanks, Ryhzuo. Both for the 5 stars and above all for reading through all this and the solidarity. I'm glad you enjoyed the read. I definitely am going to hop back in once I'm done with the Ph.D. but probably not before. I shouldn't even really be watching games right now, it's a compromise of sorts. This stuff is helping me get a better sense of the game, so that I can play with more confidence and make better decisions when I come back. And that'll be around the HotS release, yeah, or even later. I predict the air mine should give Terrans more ability to leave base vs Zerg or preempt drops in TvT, while the warhound should give some a-moveability, which should lead to some security, which should help Terrans stabilise overall. Besides, not knowing the current metagame and trendy builds, I might get some slack when everybody is in the same situation. And playing the campaign should perhaps give me some insight into Zerg that I normally wouldn't ever be induced to get (other than from the vs AI achievements). I do worry about the loss of thors from the game, especially in TvZ (in TvZ, it can be tricky without thors when the cost of warhounds includes the price of their special attack vs mech that will be unusable vs Zerg unless it's like vs armoured and not just vs mechanical).
Getting surrounded by lings with your hellions and dying sucks. Note: don't get caught behind a geyser. It's probably hard to mind hellions and banshees at the same time.
About 9:40, I guess it'd have been better to kept flying near the edge, harassing those mineral lines to death, instead of flying south. Just jump those mineral lines one after another.
Get drone kills, don't attack queens that have transfuses.
Going hellion-thor-banshee vs roaches and mutas, thors need to target mutas (obviously). Banshees can deal with roaches later.
(Skipping G3, don't have anything to note down from there.)
G4 of above:
Marine hellion marauder isn't bad, it seems (I have some kind of personal aversion to that composition, dunno why, maybe it seems unnatural to have your bio part beefier than your mech part?). Kaelaris says most Zergs die here. Day9 says Vortix is always unprepared and it happens all the time. Timing's around 10 minutes.
Slivko (Z) vs MVP (T) on Ohana, G2 at IEM (watched G1 too, no observations... or I was simply too sleepy)
For the record, I'd suggest not overdoing caffeine if you want to play. I have tons of experience (not so much results) and I can tell you the ease connected with the absence of absurd tension/overfocus/hasty pulse and whatever else is good and allows you to play in a normal way. I play better after sleeping pills than before, myself, being a caffeine junkie that works irregular hours and has had sleeping problems like forever. I recall playing Warcraft 3 on the ladder at a pretty high level at law school in the mornings before classes.
Banshee target prioritisation: drones. MVP doesn't even look at the hatch being in construction (queens can trasfuse buildings, by the way, while transfusing a drone would be impractical to the extreme). Then he attacks a single queen when it seems a sure deal. Then he kills some random zerglings on a better than nothing principle.
Harass optimisation: six hellions, two banshees (I think), one viking (for ovvies). This does mean getting close to minimal numbers of the respective units instead of milking the tech (like I'd be inclined to do).
I think it's important to have this next step after everything, and transition fluidly. This helps keep the opponent occupied and the build fluid.
According to Artosis, MVP can just waste hellions all over the map anyway at the stage he does (around 13:40). Kind of a map control thing, I guess.
MVP gets himself in the sweet spot Slivko can't do much more than roaches (which he can negate with banshees), so he can just expand at the rate of 2 CCs a time if he wants, without much trouble.
The number of banshees grows to four over time but no overkill (like 3 bases, 2 more setting up, just 4 banshees).
Again according to Artosis, MVP is maxing out, getting his deathball to roll over Slivko. While Slivko has some roaches, zerglings and infestors (with probably broodlords coming). I wouldn't mind being able to do it like that. Artosis and Tasteless refer to this as Slivko being in a fish tank, where he thinks it's an ocean but it's not. I think I'm sometimes in that fish tank in my games, while I want the opponent to be there. Terran can sort of get ball in TvZ.
Tasteless emphasises the need to control the Zerg's growth as Terran.
Slivko (Z) vs MVP (T) on Cloud Kingdom, G3 at IEM:
Proxy fail, Slivko spots it (it would have been hard not to spot it en route to base with drone scout) and denies any sort of bunker rush. Thing is: denying are at least nine drones. Only three SCVs are engaged in the rush.
I think the early drones and the kited and killed or otherwise at least not mining drones are a good gain for whatever MVP invested minus maybe the early pack of marines prior to 6:00 and the loss of the ability to come up with a scary push (marine packs not being spectacular).
Notwithstanding the unspectacular nature of an unassisted pack of marines, MVP can still force the cancel of Slivko's third while getting his own third in place. Which connects to the fact that being on the same number of bases as Zerg or especially greater is good.
MVP seems to be doing the cute little trick whereby he disengages the excess units (thors) that are not needed to kill the current target (an ovvie). If you manage to make such decisions correctly throughout the entire game the stacking effect is probably noticeable.
Slivko (Z) vs MVP (T) on Metropolis, G4 at IEM:
Not much to note down (despite a great game) other than methodical pushes being great, same for multidrops (these ones here are actually possible to follow, no insane stuff), finally Skyterran.
Getting closer to the finals. It's semifinals now. I had no idea how long it would take me to watch all those games and especially note down my observations. I probably wouldn't have started if I had known beforehand, to be honest, but right now I want to take it to a conclusion, get to the finals and then focus on work and work alone. So no watching and writing up the MLG, nope. Still potentially up to ten games ahead (only one Terran there in the semifinals), hopefully no more than five hours of real time. So here we go.
(This is gonna look interesting because Vortix has a weak early-mid game, while MVP is strong at that timing.)
Both Artosis and Tasteless attribute MVP's successfulness to how he chooses build orders appropriately.
Seeing MVP overlap the Zerg in supply before 9:00. May be due to some specific build order from Zerg but still remarkable. This leads to 56 vs 72 around 10:30, briefly before Zerg skyrockets with 9 mutas.
Yeah but ladder Zergs wall off. Tournament Zergs don't.
Response to mutas getting your SCVs: get his drones (with your hellions). (Incidentally, MVP recovers the supply lead at that point, around 13:30, leading in worker kills 38 to 22.)
The smaller the map, the better the mech. Thanks, Artosis. (Also, found out about his quip, 'when you are ahead, get more ahead,' which makes a lot of sense in SC2 from what I can tell. I'm more used to trying to cash in on any advantage immediately, myself.)
Hellions always gotta be round your thors (for the reason of zerglings and stuff). I've always been insisting on this in TvZ. Thanks, Tasteless, anyway.
MVP is kinda playing Zerg against the real Zerg. He's trying to starve the Zerg.
Those plenty of tanks that weren't useful against anything in particular made sure there was not enough anti-air (thors or vikings). This is why I prefer to skip tanks. I firmly believe in hellion thor being viable against broodlors, no matter what people say. Eventually (around 28:30), there are plenty of vikings killing the Zerg air but the key point IMHO is the absence of tanks taking up supply while not being useful. Even thors (focus-firing the broodlords) with a couple of hellions (dealing with the broodlings) should have won that engagement.
Off-topic but those guys have great headphones due the metal frame replacing a stiff plastic bar.
G2 of above (Antiga Shipyard)
Problem for everybody with his third on this map.
The longer the game the harder for the Zerg on Antiga, according to Tasteless.
Hiding rax to fake proxy (and make the Zerg overreact.
Too far away for early aggression with cross spawns (Tasteless). Gotta adapt those TvZ builds to reduce early aggression and expand more in such cases.
Third CC so near the edge of the base that the Zerg's overlord misses it by seconds and has no idea about the third!
Two evolution chambers and three queens wall (complete wall-off for Zerg). Finally some ladder reality.
Good drop: two medivacs full of marines. Losing 1 marine for an overlord and a queen. No point tempting it further. Just cash in on the advantage by simply withdrawing from that point.
Loading up marines on medivacs near mutas is a very bad idea, even if MVP does it. They're gonna die anyway. They might as well kill something in the meantime. SuperNova seemed to be good at this kind of decision-making.
What good macro does to you: All of MVP's push dies and they still end up equal on supply. And MVP leads by 20 soon thereafter (around 17:00). Counterpush by the Zerg and the supply lead gets to 50. Another counterpush on a maxed Terran by a non-maxed Zerg and it's GG. Conclusion: macro (means more than battles).
I suppose ending the game before broodlords is a success for the Terran. On the other hand, it may be that the problem securing the fourth is the reason Zerg just can't go big time broodlord.
G4 of above (Daybreak)
Supposedly every Terran opens hellion vs Zerg (Artosis). Wonder how prepared Zergs are to handle different openings. Or even hellions coming after something else, not before.
Poor micro hellion harass = bad. (Esp. with blue flame investment.)
Mech does die to roach muta...
G5 of above (Cloud Kingdom)
1 on 1 maps are best for cheese because you already know the location of your opponent (Tasteless). Which also makes cheese more likely to happen. This one is obviously ermm... obvious, but it's still good to note down (and remember in a more conscious way.
Not much to say except this is a case of MMT (marine-medivac-tank) vs everything. Just some late thors and vikings. Otherwise MMT and drops everywhere (not so many, actually).
The finals are MVP vs Nerchio. Nerchio has defeated Puma, Happy, Grubby, WhiteRa and more. The Polish Zerg is definitely gaining ground. He's probably the best Polish player right now. This Bo5 is the last games of this blog entry. Thanks for staying with me so long, hope my notes can be useful to you as well.
MPV (T) vs Nerchio (Z) on Ohana, G1 at IEM Finals
A small bunker rush can lead to an all drone pull.
After gaining the above advantage, MVP proxies a rax to keep up the pressure (seems to be 1-base all-inning vs the damaged 2-base Zerg).
When you're ahead, get more ahead, didn't we say a post ago? Don't gamble it away. The opponent has much less to lose than you. Later, the casters say the all-in could've worked with better micro.
G2 of above (Cloud Kingdom)
Good harass pays.
2CC + turret wall-off on wide ramp, seriously...
G3 of above (Atlantis Spaceship)
CC first, playing to the (large) map. Zerg can't do anything about it. Note to self: abuse this more. (While Zerg may even be going 3 hatch before pool (Kaelaris says NesTea did once vs MVP).)
Eight geysers optimal for Zerg (Day9).
MVP seems to win be taking down hatches while the big battle's going on near his base.
Double raven, HSM, corvid reactor...
HSM, well.
Marine marauder (emphasis on marauder) popping out of nowhere to shoot down the ultras gaining the upper hand in the battle. Good reaction speed/awareness.
Turns out huge maps and late game tip the scale in favour of MVP, not his non-Terran opponent.
G4 of above (Metropolis)
Ling infestor broodlord crushes most marine tank medivac play (Day9). Why I don't play marine tank medivac.
And thus so unspectacularly ends this blog entry (I think; if I look at MLG or some other major thing, it will be a new entry, if it's something smaller I might keep it here). It took so darn long. Go to Day9's Archive and see the award ceremony.
Bonus: Day9 also noticed that the last IEM was a great showcase of mech TvZ and he decided to make a daily about it, featuring some of the games I took notes from. It will be a good thing to see them from his perspective with his comments:
He picks SuperNova vs Vortix on DayBreak. The second game from this post.
Day9's essential idea is: expand early, hellion banshee, faster third. (This is kinda similar to how he summarised Bomber's TvZ as revolving around 3 command centres, 2 rax (reactor/tl), 1 e-bay and 1 tech fact.
Things I thought important to note down were that about 12 is the earliest lair on 2 bases, maybe 11, and that you should keep poking with hellions, not just hide them from mutas while farming thors, dump minerals in hellions and OCs, for each attack add about +1 minute to his Hive timing (17 base, 15 if left unopposed). And a strange notion: you want to have more than a half of the bases on the map, with expansion lead you can play passively.
EPT
)
) around 16 minutes and won. My personal struggle with TvP is that I can't win such attacks, not enough stuff or not good enough execution. Or wrong planning. Something just goes wrong for me. But maybe if I actually played TvP the standard way I'd catch up after some time and it'd be as comfortable for me against my matched opponents as it is for Bomber against his opponents.
He was the closest Warcraft 3 had to a bonjwa.)
