EPTI'm bored, so I feel like educating you guys. It's going to get a little pretentious, and if you don't like it you can shove it (fuck yeah, expectation into satisfaction 2hit combo). I'm going to tell you why the way you think about strategy is all wrong, and then I'm going to tell you the right way. Hopefully, some images will make this post bearable to read all the way through. This was inspired by the sea of uninformed casters (which includes all except nony/tyler), especially those who do any discussion beyond narration. Though i mention Day9 and disagree with his analyzing style, I have nothing but respect for him. This is not about discrediting anyone; I simply want to share my thoughts.
![[image loading]](http://imgur.com/i5IAo.jpg)
1.1: The face of your doubt in my words
+ Show Spoiler [ tl;dr: nothing happens in this part] +
Firstly, I will tell you my story, and why Broodwar was just so hard for me. Then, I will tell you about heuristics (which is essentially day9's philosophy), and then I will discuss strategic decision making, and how all these topics relate. I will conclude with future topics that I wish to discuss. Please leave some feedback, and give it a read through, I really put some effort into this one.
I started playing BW in September of 2009 (My girlfriend bought me a copy (BRAG BRAG BRAG)). I had watched Cholera and Diggity and many procasts, but I had very little understanding of how the game worked. I switched between the races, but spent the most time as toss and hit D+ before I switched to the SC2 Beta. My only wins were based on specific timing attacks, such as Day9's 1base goon reaver, +1 speedlots, etc. I struggled very hard with the game, because I had almost no understanding of what caused what. If I won, I won, and if I lost, i lost (this is called the reflexive property of winning). I read all the literature, watched all the FPVODS, but my improvement was slow and frustrating. My mechanics were quite good; it was usually my decision making that resulted in my losses.
In sc2, I played Terran, but i never really honed it. I just found many abusive openings, and tested the limitations, and tried to enter mid-game with a solid lead, which I tried to extend into a win.
![[image loading]](http://images.cheezburger.com/completestore/2010/8/5/66b6fcd8-2d18-4fea-87fd-a33d72608095.jpg)
1.2: Idra disagrees with my life choices
School got busy and I stopped playing SC2 as frequently. With my free time, I played Settlers of Catan and started looking into chess. It was through these games that I finally understood what was wrong with my Starcraft mindset. Because I entered BW with others' advice instead of an understanding of RTS, I had no knowledge of the basic theory of the game. I knew the Bisu build up to 50 supply, but I didn't know why I was doing it. I blame this on my late entry into the game and its relative maturity, and the punishing skill levels of iccup, but I was unable to truly understand the game. Further thinking brought me here, where I feel confidant enough to share my views with you all.
ABRUPT TOPIC SWITCH HEADER: Heuristics
Heuristic is Latin for "find" or "discover," and though it strictly means "experience based technique," it can be thought of as a rule-of-thumb (yay, wiki!). Day9's strategic thinking uses a compilation of these heuristics to create an overarching strategy. He always says things like "Fast units punish mistakes better than slow ones," or "Expand while attacking" or other things to that effect. Other common heuristics include looking at supply counts, units lost, army value, or past events to measure who is winning.
However, these rules are very limiting. To use the "fast unit punish mistake" example, sometimes slow units are the best at punishing mistakes. For example, in brood war PvT, if you keep your army inside your natural or too close to your natural, the slow lumbering Terran Mech will punish you by setting up a contain; which will almost surely win the game. This is not to say that Day9 is wrong, it is only to point out a weakness in heuristics that you need a separate rule for every situation. This is why many players have trouble with situations they haven't played before.
![[image loading]](http://www.gosugamers.net/starcraft2/images/news/day9.png)
2.1: "Someone talking about me?"
It is my belief that while these are useful, if you do not understand the underlying and raw theory underneath, you're severely limiting yourself. By understanding the barebones, you do not need to rely on a list of things that may or may not apply to the game you are playing; this is something you we will talk about later.
ABRUPT TOPIC CHANGE TWO: Decisions and Strategy
A game is about making decisions. Each decision will affect your standing in the game, no matter what. Even if a decision keeps you even with your opponent, it is because he made an equally good or bad decision. (Good decisions I will call lead-maximising, and bad ones, deficit maximizing). In the first moment of starcraft, you build a worker and send your workers to mine. Though this is "Standard" play, it is only such because it is 99% of the time the lead-maximizing decision. These decisions are easier to imagine in turn-based games, because there is a set number of decisions, as opposed to infinite choices at infinite times in the game. (This is why people measure games in deviation from standard play; it limits the number of decisions to be analyzed.)
![[image loading]](http://blogs.sfweekly.com/foodie/youdroppedfood3.jpg)
3.1: Decision trees are illuminating
In every game, there is a point at which the game ends, and the result (or winner) is measured. In chess, its checkmate; in Starcraft, all buildings razed. These games are similar in that one player has won long before that point, and that is when the loser resigns. Aside from stalemates and base races, this "end" is rarely achieved. The point at which the player resigns is when he realizes that no matter what decisions he makes, it is one that lengthens his disadvantage.
Thus (in the broad sense of the word "Thus"), your goal as a strategist is to force your opponent into a place where all his decisions lengthen your advantage. This is the pinnacle of accurate play. In Starcraft, this is usually done through army/macroeconomic value, simply because of the nature of the game.
I will now attempt to explain Idras playstyle, for i find both Idra and Zerg most illuminating in this concept.
Case Study: Idra ZvT
Zerg has 3 ways to invest its resources: Tech, Macro/economy, and Army. It cannot do all 3, as Teching detracts from economy (lost drones and opportunity cost of gas) and Army size (less money), Economy detracts from Army size (less larvae), and Army size detracts from Economy (larvae constraints). Because of this triangle, Zerg players choose which to invest in at what time.
![[image loading]](http://imgur.com/HFu7Q.jpg)
91.~ fdsfgs
Idra's playstle is simply defined. "Invest everything into economy, so long as: at any point, my opponent's decision to attack is a mistake." In the early game, he does this by having just enough units and larvae so that any attack can be defended without loss of economy. In midgame, Mutalisk contains accomplish this. After this point, he hopes that enough of his investment was into macro that he can close out the game. This is why when Idra wins, it is always crushing and fearsome. In Idra's wins, there never seems (to the untrained eye) to be a weakness in his play. In his losses, you think, "Damn, if only he built 4 more roaches." It is the fine line that he walks, and the narrow escapes that give him huge advantages in mid-late game. I find that the only exception is when Idra is confronted with a build he's never encountered (i.e. idra vs nony showmatch with the 3gate blinkstalker timing).
Case Study: Choose your adventure/Selfquiz/Choices choices choices
You are playing cross position TvT on Shakuras, and you open blueflame hellion drop. you complete the drop, and you kill every single one of his scvs. he kills your medivac full of hellions while you're trying to escape. You are faced with 4 options at this point: Attack, Expand to nat, Double-expand, tech to banshee, or turtle with tanks and 2 barracks. Which ones are lead maximizing? Which ones are not? What situations make one stronger than the other?
Attacking is a mistake. His army is almost equal in size to yours; your advantage lies in the future. Why rush confrontation when time makes you stronger? the longer you have an economic advantage, the better your overall army advantage.
Expanding is a mistake. Many with the "expand your lead" hueristic would have chosen this, but the rule is inadequate for this situation. At this point, your lead has yet to materialize into your army. you've spent 500/100 on a drop that has yet to damage his army. Why would you sink 400+ more into an expansion that will make you more vulnerable to attack? Because hes has no producing power at the moment, his army will only become weaker relative to yours. If he is good, he will attack immediately, while this weakness is minimized. you strengthen his timing by weakening yourself at the moment for a future gain. To put it shortly, he has a mild army disadvantage, but you nullify it by investing in something besides your army. Until your CC has paid for itself, his army has a decent chance at annihilating yours, and winning the game. Instead of limiting his choices to deficit maximizing ones, you have giftwrapped a winning decision to him.
Believe it or not, Double-expanding is surprisingly valid, as long as the variables are correct. If he has an army with a large amount of tanks or thors, and you can expand to sufficiently out-of-the-way places, he has no way of taking out both bases. by the time he kills one and moves on to the second, (especially with you distracting his army, forcing sieges, poking at his base), your economic advantage will have paid off. The decisions he has are all wrong. If he stays in his base, you will crush him with 3base macro vs 13scv macro. if he attacks, he can only move slowly enough that one of the bases will have paid off already. attacking your main base is his best decision, but cross position on shakuras, we're pretending that you can delay it long enough for your investment to pay off.
Banshees are a mistake. Banshees will blow up his buildings while he counterattacks, but it will not stop him. In the meantime, he will crush your army that has invested in a banshee that is at his base, not defending. He already has no scvs, so he will only be losing supply depots, which he doesnt care about in this all or nothing situation.
Turtling is the standard play. You are operating at a huge economic advantage to him; you want to prolong this period of time as long as possible, so that it pays off for the maximum amount of time. Playing passively opens up the opportunity for him to expand, but with no scvs and no production, this will take a long time, and you have ample time to build an army and crush him.
Conclusion and Summary
Watch a few of your replays. At every decision point, you should ask yourself, "Why did I do what I did? How does this put me closer to a situation where all his choices further my lead?" If you cannot answer it, then it is probably a mistake. If you do this rigorously, you will understand the fluid and underlying theory behind all strategy games. You will be able to outsmart and outmaneuver your opponent so consistently that they will throw up their arms in anguish, "No matter what I do, I lose." Thank you all for reading.
Future subjects
Playing from behind: Why Idra resigns prematurely
Punishing mistakes: Why players 4gate when ahead in a series
Legitimacy of cheese, 1base play, and scv-all-ins
Game theory and strategic analysis of social interactions (Sociopathy 101)
Decision-based replay analysis (imagine day9daily, but much, much, slower)
