EPTPracticing to Improve
People play games for many different reasons, and Starcraft is no exception. However, Starcraft has a following that is much more hardcore, a large percentage of which are focused on obtaining skill in the game. I am deeply entrenched in this subset of people. It isn't that we don't think the game is fun in-and-of-itself; what we are even more interested in is something about ourselves: the cultivation of our abilities, and ways we can overcome our opponent (read: winning).
That said, a surprising number of players who fall into this category don't really think about their learning process, and trudge mindlessly forward bashing their face into thousands of games until they aquire a modicum of skill. The raw amount of time one has put into the game will certainly affect how strong of a player they are, but there is so much more that goes into it. There are others who pay more attention to the game and their execution of things in it, looking back at replays and trying to figure out what went wrong and what adjustments could be made. These players might do such things as practice specific builds, or drill one match-up on one map at great length.
Then, there are the players who practice.
When I speak of practice, I am talking about something more specific than playing with the intent to get better. In practice, I am talking about doing something very specific to improve an aspect of one's performance as much as possible given the amount of time one has available. Being a musician, I can't help but draw the analogy of learning an instrument. Compare someone who spends most their time practicing a musical instrument versus someone who spends most of their time playing it. People who have played an instrument for a long time can become quite competent, but will not improve nearly as quickly as one who understands how to practice and spends most of their time practicing instead of playing. Our society doesn't really value video-games as a pursuit worthy of great effort, so it is rare that you will find someone who practices a video game as they might practice a musical instrument, a sport, or some other craft that our society has labeled as "valuable." Yet the process for obtaining skill in a video-game should be similar to everything else. This is a practice method that I think is extremely useful for learning any skill, though in this article, it is solely addressing SC and SC 2.
Of course the first step is deciding what you want to improve/practice. Second, analyze what your weakness is, and why it's there. This is probably the most difficult part, but will come naturally over time. Third, you must use the resources at hand to come up with a specific practice method to address this problem. Finally, you must be disciplined and put in the time practicing in the manner you have decided. Here are a few specific examples.
Let's say you want to work on a mechanical aspect of macroing. First, you must analyze what you want to improve, what your weakness is, and why it's there. I am not talking about "my macro is bad at such and such a time", or "I have too much money now", or "I need to add on two barracks at this point," etc. Though that type of analysis is very important, that is what you would want to do BEFORE this step. That type of analysis has been addressed in numerous different threads or day[9]'s podcasts and daily videos; thus, it will not be addressed here. I am talking about "I am having trouble spending my money at this point BECAUSE I am not able to cue units in my barracks quite fast enough, which upon further analysis I think is because my mouse movement isn't as precise or as fast as I'd like it to be" or "I don't have any marines cued in my barracks BECAUSE I'm spending all of my attention microing my army or BECAUSE I am not jumping from building cluster to building cluster as smoothly as I'd like." This analysis stage is difficult because you must be able to remove yourself from the events in the game and observe things about yourself.
Now, think of some ways to address this problem you have found. There are probably many different ways you can do this depending on the specific problem, though having a practice partner that is actually willing to work to get better--instead of just playing--is very useful. You might be able to use a UMS or a saved game file (there is less flexibility right now in SC 2 I know) that sets up a scenario for you to drill or practice how you would like. Just make sure you isolate what you are specifically working on as much as possible. Then, drill it until you (or your partner) lose focus, or until you feel the problem has been somewhat addressed. Some things are harder than others: so though you should see improvement very quickly, you probably won't get to the level you want with just one practice session (or even ten or a hundred sessions depending).
I'll give two personal examples: a good one from SC, and a not-quite-as-good one from SC 2 (I'm just starting to grasp things like which units do what, so I haven't done much practicing of it yet).
SC Example+ Show Spoiler +
I noticed a while ago that in some high-adrenaline games I would often have bad clicking precision and miss sniping medics or single marines with my muta stack, resulting in running them in to do nothing save getting damaged by stimmed marines. I also noticed I had a lot of tension in my arms, which I hypothesized was a large contributor to the bad mouse precision. I played a handful of 2 hatch muta ZvT games while observing my physical state and realized that right after my mutas came out I became excited and I would start to get tense in my abdomen and shoulders. When actually microing them against the opposing army or base the tension became far worse, resulting in (I guessed) more jerky motions and less mouse pointer precision. After becoming aware of this, I had something to work with: this tension was making my muta control worse, and I needed to come up with a way to iron it out. First, I used a Python Muta Micro UMS (created by Grobyc), which was the lowest stress environment. When starting to work on muta micro, I would usually play this for 5-15 minutes to make sure I was relaxed and the most fundamental control was clean. Then, I grabbed a willing clan-mate to play various different Terran styles against me, making saved game files around 30 seconds before the mutalisks came out for each scenario. Then we drilled and drilled, resetting back to the same save point every time the muta harass ended (whether it ended well for me or not). Since I was specifically addressing the state of tension in my arms, I would spend a lot of attention on my physical self trying to stay relatively relaxed. This of course resulted in the horrible loss of entire muta stacks while I was getting used to it; however, after a relatively short amount of time, I was able to stay much more relaxed when microing mutalisks, which, as I had hoped, greatly improved my clicking precision. Problem analyzed, and solved. Suck it, medics!
SC 2 Example+ Show Spoiler +
In watching a handful of my replays, I was shocked at how horrible my larva inject timings were. My queens sometimes had 50+ energy at only nine minutes into the game, and trust me, it was not for a strategic reason. Thanks to the ridiculous amounts of information one can get off of the replays, it was very easy to see that I was focusing nearly all of my attention on army placement and less important things like telling my overlords to scatter and generate creep on my side of the map. It seemed to me that the issue was that I simply didn't have the larva inject at the front of my mind, as opposed to not having the attention or speed to jump back to my bases to larva inject (I mean, who am I kidding? It's not hard at all to do this, I am just a noob). So, my strategy was to try to ingrain constant larva cycling (selecting hatch 1 and its larva, then selecting hatch 2 and it's larva, etc., but in a very fast way so it doesn't disrupt what ever else you are doing) into my muscle memory while doing basic army manipulation, so I would have the larva inject timings more in the front of my mind.
[Small tangent here-I think larva cycling is both easier and more important in SC 2 than in SC. You can have all your hatcheries hotkeyed with Jaedong-style easy-to-reach hotkeys (doesn't use keys 8-0) since you practically never need to add on hatches for extra larva, AND it really helps keep tabs on larva inject timings. Just something to think about...but now I digress]
I didn't really have a practice partner, so I first hopped in a game with a computer, macro'd an army while putting up three bases, and ran my unit groups through an imaginary obstacle course while larva cycling, double-tapping back to a hatchery each time it was ready for an inject. After I lost my focus doing this, I tried playing games against humans: not with the intent of winning, but with the intent of the continued address of this problem. After a few games of this I got a lot better and stopped practicing; but as I sit now, my larva injects are still at an unacceptably low level, and need more practice.
[Small tangent here-I think larva cycling is both easier and more important in SC 2 than in SC. You can have all your hatcheries hotkeyed with Jaedong-style easy-to-reach hotkeys (doesn't use keys 8-0) since you practically never need to add on hatches for extra larva, AND it really helps keep tabs on larva inject timings. Just something to think about...but now I digress]
I didn't really have a practice partner, so I first hopped in a game with a computer, macro'd an army while putting up three bases, and ran my unit groups through an imaginary obstacle course while larva cycling, double-tapping back to a hatchery each time it was ready for an inject. After I lost my focus doing this, I tried playing games against humans: not with the intent of winning, but with the intent of the continued address of this problem. After a few games of this I got a lot better and stopped practicing; but as I sit now, my larva injects are still at an unacceptably low level, and need more practice.
Of course I would not espouse this method of practice to the point of excluding others, as it is very important to play games and see the "big picture" of how your development is coming along. However, I do think it would be a very good use of most players' time to augment their playing with this method. It's certainly not for everyone, but for people whose main goal is improvement in a game (i.e., a lot of the enjoyment they get out of it is the process and result of getting better) or for people who want to try to compete but don't quite have the time to mass game, I strongly recommend you try out this disciplined style of practice. I am totally convinced it is more efficient than just playing thousands of games, even if you are playing them with the intent to improve.
Because after all, Starcraft is serious business.
~Hungarian
In sc1 i couldnt connect to battle.net at home so i couldnt really play games vs others. So what i did was going through a similar process like you describe vs. computers and getting all mechanics, builds and so on in place. Also analyzing your own play, and watching your replays to find what goes wrong and why helps alot with the input of watching good players and seeing how it is supposed to be done 
