Choosing Prodoss

So, picking up from my last post, I stopped four-gating a while ago. I wanted more variety, so I went back to playing Random. I stopped relying on what had been easy win builds and just started doing whatever. And losing. A lot. I couldn't spend all my money, often because of a shortage of either minerals or gas. I realized I'm just not experienced enough to fly by the seat of my pants, because I haven't developed a feel for how to sustain whatever crazy build pops into my head. And I really wasn't learning anything or getting better. I guess I'm kind of stubborn about wanting to be a jack-of-all-trades. It's the same reason I never look up build orders online. Instead of following the instructions of those who have tested these things extensively, I wanted to get to the point where I would just naturally do efficient builds, based on an understanding of the game. But that's just too extreme. I should treat build orders like I treated joseki in Go (I'll eventually create or link to glossaries for my two games some time, but a joseki is sort of like a Chess opening). I used them as a study tool, trying to understand why each move was made rather than simply memorizing them. Similarly, I can study particular build orders (especially via replays of top players) and copy them to the extent that I understand them. After all, build orders only ever cover the first few minutes of the game, so it would be impossible to rely on them completely.

Speaking of BOs, this is an interesting thread on TL about someone who claims to have built a program that can tell you the fastest way to get to any set of Zerg units. There are the obligatory know-it-alls saying that it's crap because it doesn't take everything into account, but I think it could be a valuable tool to point toward better strategies.

So anyway, I've decided to stop playing Random so that I can focus on refining a small number of builds with only one race. In choosing a race, I ruled out Terran because I didn't want to put up with people saying that my race is OP and that's the only reason I ever win, blah blah blah. But then I was torn between Zerg and Protoss. On the one hand, I love how Zerg is fundamentally different from the other two races and how it gives you more flexibility in terms of expanding and tech switching. But I decided to go with Protoss just because I love using forcefields, blink Stalkers, Observers (who needs hax) and Void Rays. I've been opening almost every time with 3 Gateways and 1 Robotics Facility, but I haven't bothered to refine it to the same pattern each time, especially in deciding when to get the Robo. So establishing that exactly will be my next step. Another exciting thing I've been doing in quickmatch is working on late game macro. Earlier I made the decision to always stay on only two bases because I knew I couldn't handle the multitasking involved with keeping any more than that. But I can handle it now, and it's really fun. I can already see that if I get fully comfortable with that, it'll easily add at least 200 points to my rating.

Speaking of points, I haven't improved since after my 4-gate winning streak from last time. You get 1 bonus point added to your pool every two hours, and if you take that into account, my rating has actually gone down slightly since my last post. That's a little discouraging, but I want to be dedicated to this pursuit. I want to see just how good I can get at this game, no matter how long it takes me.

On another note, WTF?! They shut down Nexus Word Wars because the word "dyke" was in it?! That's a perfectly benign word (it means "dam", people) unless you specifically use it in an offensive way. I'm pretty sure I played that game about 50 times. There's something about this sort of typing game that I just love, probably the frantic pace and the thrill of having your fingers flying every second as you try to make intelligent decisions (probably why I like sc2 itself).

--

And finally, I made this post to reddit.com/r/starcraft, and I have to say I was surprised by the negative reaction (right now, 21 up votes, 24 down votes). I've been frustrated for a while about casters mispronouncing words, because (1) I'm always frustrated by people misusing English, and (2) I don't want to believe that video games, even at the highest level and most cerebral, are exclusively the realm of twenty-something basement dwellers. I've seen many posts on Reddit about individual words, including zealot (many people seem to finally grasp the first syllable but not the second), centrifugal, and barracks. Day[9] also recently jokingly bragged (I think it was during his sc2 release party) about knowing how to pronounce all the difficult words in the game, including chitinous, which is one I didn't know. And after hearing one or two people pronounce the word Templar (I thought it sounded like pulsar) as "tem-pler", I looked it up and found to my delight that I was wrong and those one or two people were right (although in both games, it is pronounced the "wrong" way). So I got curious as to whether there existed a "Starcraft glossary" to help with pronunciation and in some cases definition. I couldn't find any, so I had my idea for my first Reddit post. I've always been fascinated by languages and understanding the correct way of speaking and writing them (fun fact: I speak fluent French). And I love to find out about misconceptions, particularly when it's my own. Turns out not everyone likes to be corrected. Perhaps pronunciation is tricky because it's the least concrete of language skills since there are so many minor variations from person to person. There were some positive comments, but the general message was, "Don't tell me I'm wrong! This is how I'm used to it! Language can't be pinned down to what you find in a dictionary! Your pronunciations sound dumb, anyway." I wasn't sure how to feel in response. There was a part of me that just wanted to be angry and hoity-toity about people getting upset about me writing down what was in the dictionary. But in general, I don't like to be elitist. If so many people thought I should shut up, there had to be some sort of validity in that. I'm still struggling with coming to a conclusion, but here are my main ideas:

(1) I simply cannot allow myself to hold to a belief that others should use correct language if it is only for the sake of correctness. I get annoyed when people complain about there being too much foul language on TV, and my justification is that there is no inherent evil in it other than that which you create by the arbitrary rule that certain words shouldn't be used. It would be hypocritical of me to treat grammar/spelling/pronunciation/usage mistakes any other way. Just because I like to be right doesn't mean you have to.

(2) There are many good reasons, however, to be correct. The most important is that it facilitates communication. A great practical example of this comes up when you try to communicate with a non-native speaker of English. A mistake here or there might not hinder understanding since the context makes up for it. But if you combine too many mistakes into one sentence, you may start to have some real trouble. But that points to perhaps the core issue: it's not about speaking correctly according to the dictionary, it's about speaking according to what the listener is used to. In other words, we should make an effort to conform to the standard expectations of our community.

(3) What are the advantages and disadvantages of sub-communities developing their own variation from the wider standard? Perhaps it may instill regional pride, but it may also contribute to stereotypes from without. Consider, for example, the southern U.S. accent and all its effects.

(4) To what extent does social context play a role? Is there an obligation to have the ability to speak correctly even if we don't exercise it in informal contexts? Is pronunciation fundamentally different from the other language skills in this regard, since people tend to be comfortable with only one pronunciation per word?

So I'm going to try to relax about pronunciation, but I don't know if I'll ever be able to stop shuddering every time Artosis sits at a microphone in front of the biggest global Starcraft tournament of all time and says "zell-awt".

Games played: 514
League/Points: Diamond/1176

4 Warp Gate is kinda good

Yup, it's true what they say. I just got done with 18 games in three hours, winning 16 of them. I used nothing but 4 Warp Gate. One loss was to 8pool on Delta Quadrant, and the other was to a fast-expanding Terran on Lost Temple who managed to defend his expo with SCVs repairing bunkers. Kudos to that Terran player. But I'm not sure what I could have done against the Zerg. I happened to scout their location last, so by the time I knew what was going on it was probably too late. I guess I'll have to start sending a second scout against Zerg.

Anyway, 4 Warp Gate is stupidly amazing. (Here's an example replay.) I think I'll try adapting it to include taking my natural expansion. Sometimes it can be difficult to break my opponent's defenses, but then I can usually contain them in their main base. And once my second base is up and running while they're still on one base, it's only a matter of time before I can overrun them with pure numbers.

Games played: 395
League/Points: Diamond/1039

Greetings, Prophet. The great work begins! The Messenger has arrived.

This is where the real work begins and I start deciding how dedicated I truly want to be. After almost 400 games, I've finally stopped coasting. I don't just automatically win most of my games anymore. I can't just expand when I feel like it. Harassment isn't guaranteed to be profitable. Sloppy wall-ins get broken down, and sloppy macro gets outdone. And most importantly, just because I have a big army doesn't mean it's going to be a good idea to attack move into the opponent's main. Here's a good example of me losing to brazenness (ZvZ on Delta Quadrant).

It's time to start thinking a little more and polishing my game up a little more. I think I'll focus on just one race for a while, probably Protoss, my weakest of the three. Perhaps I'll start with the typical 4 Warp Gate push and see where that takes me. I like the idea of starting by using only one strategy. Eventually, you perfect it, and then you can look at the games you're still losing to see where that strategy's weaknesses really are, at which point you modify it as little as possible to accommodate encountering a similar problem in future games. As you continue to modify it over time, you'll in theory eventually end up playing a fully versatile, many-faceted game that may or may not even incorporate the original strategy.

Games played: 372
League/Points: Diamond/806

O_O

I'm finally feeling better, so I've been playing a little more. First impressions of Diamond League: Oh my god, people actually know how to make units! Like, as many as I do! Am I going to have to employ... strategy?

Starcraft a-slow-slow

Haven't been playing much recently due to being seven kinds of sick. Just now I tried some games and ended up winning 3 and losing 6. I kept making ridiculous macro mistakes, immediately falling behind like I did in the initial weeks. I guess my mind isn't as sharp right now. I'll get some rest and come back when I feel better.

I made it!

Seriously, though, I'm terrible.

Games played: 337

League/Points: Diamond/838

Lessons in the Fundamentals of Starcraft

There is a famous book about Go called Lessons in the Fundamentals of Go, written by Toshiro Kageyama, a top professional player. Instead of using lots of variation diagrams, individual joseki analysis, etc., the book takes a small number of simple examples and uses them as a platform from which to speak about how you should think about the game. I remember a line in there somewhere about how the way to improve at go is to start by studying the fundamentals vigilantly, building a solid foundation for your game, and then, when you begin to progress to the higher stages, to study the fundamentals again. Kageyama seems obsessed by this idea in his writing, as though frustrated that more people don't understand and appreciate it.

So does Day[9]. Just look at some of his recent Dailies, especially #177 "Simple Winning Techniques", which I am now halfway through watching. His basic idea is that people become too concerned with intricately countering the opponent's strategy and/or unit composition and lose sight of the fact that you can just win by getting a huge army and rolling all over them. And it makes perfect sense. Sure, you can beat those Colossi by getting Vikings, but you can also beat them with more and more Marine-Marauder. You will lose more army value in the fight, but that's value that would have been spent on the tech anyway. I'm speaking in broad theoretical terms, of course. Continuing the Colossus/Viking example, it seems the main differences between the two approaches are... 1. You only have to pay for tech once, but you have to pay for army size repeatedly. If the game goes longer (that's an important "if"), you will still have access to Vikings as you build and rebuild your army. But every time you rebuild with huge numbers of MM, you have to pay again for that bulk. So teching to a counter could be seen as an investment just like expanding. 2. Specialized units are just that--specialized. After your Vikings take care of all the Colossi, what are they going to do? If you end up using them a lot in ground assault mode, you basically just have a bunch of slow, overpriced infantry that can't stim. Since you cannot exactly predict what units your opponent will get, this phenomenon is essentially guaranteed to occur every time; the only question is to what extent.

It's ironic that Day[9] has started harping on this subject recently, because it's exactly the stage where I find myself. I am still just hideously awful at micro and harassment--both using it and countering it. It's because I instinctively interpret macro (for better or worse) as the initial requirement that you must master to at least 90% or so before you can even begin to play the real game. If I let a Queen get to 50+ energy because I'm forgetting to inject, and then my opponent shows up at my base and kills my whole army, then as far as I'm concerned, that fight didn't really happen. It was totally invalid, imaginary. I've heard of some Go teachers who despise being asked questions about how move 50 could have been improved in a game where move 40 was some fundamental error. "It's a meaningless question anyway! That position is impossible." To these people, playing a game is like doing math. Who cares what the square root of x equals if x was obtained by dividing by zero? Logic says that if you assume a falsehood, you can prove anything. I will now prove that I am a god. Let us assume that pigs fly. Clearly, since pigs don't fly, it is true that if pigs fly, then I am a god. But I have assumed that they do. Therefore, I am a god. Now let us show that that wave of Stalkers my opponent is sending to my base is going to utterly annihilate me. First, let us assume the absurdity that I, for no particular reason other than forgetfulness, did not inject larvae. Clearly, if I do not inject larvae, then my economy and army will be inferior. Thus, as we now see from the mass carnage on screen, I lose the game. Well, so what? The situation was impossible anyway. Let us not, oh please for the love of Aiur, let us not begin postulating some new strategy to deal with a similar wave of Stalkers in a future game. The fact that I made a series of dumb macro mistakes and then got killed tells me hardly anything at all about whether my build/strategy was any good. Would you think that Hellion harass was bad because you got distracted and accidentally ran your Hellions straight into a ball of Roaches? That, to me, seems just as silly of a conclusion. But here's the difference between those two things: Hellion harass is something you try from time to time, and macro-management is something that you have to do every damn game. So how can I know if, uh, getting units and attacking with them is any good until my macro is reasonably close to ideal?

I've now played over 300 games with this attitude, and I still see myself as a total macro noob. (Oh, but did I mention I'm back in Platinum, rank 2 in my division, my most played race is now listed as Zerg, and I usually play Random? Yes, things are a-changin'.) I recently played a game (here's the replay) as Protoss where I totally blew my own mind by managing to maintain four bases and then using my massive income to overrun my 3-base Zerg opponent. Being able to easily and continuously produce off of ten Warpgates, two Stargates and a Robotics Facility gave me a childlike glee like what I felt when I played my first 1v1 on 7/27 and wreaked havoc with a fleet of Battlecruisers. It made me think back to when just the idea of expanding to my natural was a lot to handle, and it reminded me of how much of a way I still have to go before I gain the skill to actually go around the map and do what I want to do. Strategy is all fine and good, but I'm still working on moving my pieces across the board.

Here are my main macro points to work on:
1. Understand the progression of your income, and anticipate the immediate future. Don't wait to get to 1000 minerals before you realize you need to build more production buildings. Understand that as you get more workers and expand, you will need to have things in place to be able to spend the added income. Always know approximately how many workers are at each base and how many are collecting gas.
2. Don't panic. The speed of the game can make it feel like you'll never have the time to build enough stuff to spend all your money, while making sure you've got three workers on each gas and minerals aren't over- or under-saturated, etc. But that's just not true. Really, how much does it take to start spending your money? Maybe building, say, four more production buildings and adding them to your hotkey groups? Is that really so complicated?
3. Similar to #2, don't get sidetracked by less important but more practiced activities. If you are gas blocked, stop figuring out how many Overlords to make and start making Extractors and putting three Drones on each. Especially in a panicked state, it is easy to fall back on the most ingrained habits. But if the habits are ingrained, that means you're already doing them well and don't need to devote as much attention to them! It's a vicious circle. By definition, you have to put more effort into things that are more difficult.
4. Don't stop producing workers! No really, don't stop! Ever! I mean it! Keep making them! Do you have a lot now? Oh, that's cool. Keep making them!

Some other points:
1. The enemy is not the units attacking you. The enemy is the guy clicking on those units and telling them to attack you. And that guy is just as limited as you are. Gaining an understanding of what that guy is doing and why is fundamental to understanding the progression of a game. As you play, imagine casting your own game as a spectator with full vision of the map. If you don't know enough to be able to do that, scout! (Or, as Day[9] suggests, just reason out his limitations.)
2. Chokepoints are death chambers. The DPS of a Roach that can't get close enough to the enemy to attack is 0. It is perfectly normal to have two comparable armies on either side of a choke in a standoff situation. That's why it's so difficult to attack the opponent's main.
3. Don't get chained to army hotkeys. Box off units and issue orders appropriate to that chunk. In particular, don't repeatedly press 3a-click (I use 3 for my main army) on the same spot. That's like trying to make Mario jump higher by squeezing the A button as hard as you can. A waste of energy.
4. Have fun! Remember how awesome this game is and what an opportunity it is to be a part of such a large, active community.

Games played: 317
League/Points: Platinum/887

p.s. Fun fact: Glancing at the analysis in Sc2gears (highly recommended app, by the way), I see that since September 1 when I started using the program, Random has landed me on Terran 16 times, and I won 15 of those games. XD