Beginner's Guide to Starcraft
Thanks to Shockwave[xpow] from the B.net forums.
Here's a compilation of tips for beginners who are interested in learning basic principles behind Starcraft. The post is long but you don't need to read every part to get started. Just look at whichever sections are relevant to what you want to learn; in particular, the latter half of the post is mostly just explanation of theory that you may or may not be interested in. As a general caveat, there are exceptions to every rule, so I'm not claiming these are inviolable and will always be beneficial. They are just good principles to follow absent any compelling reason not to; as you get better, you'll start recognizing when to deviate from certain rules.
THE "BIG FIVE" BASIC PRINCIPLES
Here are the five basic skills that I think every new player should start out trying to get decently good at.
1) Always be building peons (SCV/drone/probe) and put them to use on minerals/gas. Don't ever stop. Your CC/Nexus should always be flashing, your main hatchery should always have at least one drone building from larva. If you need to queue up peons, do it, but try to avoid having to do that since it ties up resources. When you get gas, put maximum 4 peons on it if it's close to your town hall (CC/hatchery/Nexus).
2) Never be held back by your supply limit. Always build supply early enough in advance so that you're not waiting for your next depot/overlord/pylon. E.G. Generally you should start building your first supply at 8/10 (or 8/9 if Zerg). In later game, build it much earlier if you're pumping units faster.
3) Try to keep your resources as close to zero as possible by spending whatever income you're getting. If you have a barraks, make sure it's making marines. If it's making marines and you have leftover money in the bank, build another barraks or tech, expand, etc. Just don't let your money sit in the bank.
4) Dont make too much static defense (cannons/bunkers/turrets/sunkens/etc.). Build those minimally. E.G. 2-3 cannons is about enough for each base you have, use mobiles for the rest of your defense.
5) Scout early. You'll want to use either a peon or an overlord if you're Zerg. When your supply reads 10, take a peon off your resources and send him around the map. Just hold down shift and click on various spots on the minimap.
LEARNING THE GAME
I would recommend the following steps to actually getting better at the game, working your way up from first computer opponents to real ones.
1) Go through the single player campaigns, at least the Terran one, they're fun and they teach you various basics. Make sure you log onto b.net at least once so you can get patched to the latest version. Don't worry too much about strategy when going through this; e.g. you may not be able to practice the Big Five in some missions if resources are limited or you're forced to build a lot of static defense to survive.
2) Next, try playing some 1on1's against the computer. I would pick simple maps like ladder/LostTemple or Challenger. Play until you can beat the computer consistently. Stick with one race, I recommend Protoss because they have the simplest build/economy structure but any race is fine.
3) Then try playing 1on1 against the computer without building any static defense until you can consistently beat it. Try to beat the computer as fast as possible, play aggressively and send your first units to attack when you get them, don't let any units sit in your base doing nothing but instead be constantly on the offensive.
4) Get online and play against real opponents. Start by creating "1on1 newbies only" games and kick out anyone who has an appreciable record, or not if you don't mind getting slaughtered. If you lose, save the replay, watch it, and figure out what your opponent did that beat you. Then try doing what he did against your next opponent. This way you're always copying people who are better than you. If you need practice, you can always play against the comp and get good at executing your learned strategy until you're comfortable with it.
5) If you can't get online and are creaming the computer regularly, try adding more computers. 1on2 should be manageable without having to build any static defense. If you can't do 1on2, do 2on3 where you ally yourself with one computer.
6) If you have specific questions about strats or things you've seen, post them here. Do not post very generic questions like "How do I get better at Terran?", but rather stuff like, "My opponent made a bunch of wraiths in my last TvT game, I have no idea why but he did beat me, can you explain that strat?"
THEORY OF ECONOMICS
Let me explain some of the rationale behind the Big Five now. For basics, my philosophy is that economic management is the key to any RTS and thus should be the first thing you learn how to get good at. If you can get money in as fast as possible, and spend it as fast as possible, then you'll do pretty well, and that's what the Big Five focuses on. If on the other hand, your opponent attacks you and you die with 1000 minerals in the bank, that's wasted money that could have e.g. been another 10 zealots or what not and you may have survived the attack. Making peons constantly means that you are getting resources as fast as possible. Peons make up for their cost very quickly. There is a limit to the number of useful peons you can have especially on smaller maps since only one peon can mine minerals at once, but I wouldn't worry about that for now. Most newer players forget to make peons and thus never come anywhere close to reaching that point. If you are oversaturated, it's no big deal at the recreational level.
At recreational level, I also believe it is more important to make sure you're spending your money on SOMETHING, versus fretting about the best unit or tactic to use. We can talk about specific tactics for race matchups, but that's more advanced and is likely to cause you to focus more on advanced stuff versus making sure you have the basics. I could e.g. teach you one tactic that can win you most of your newbie games with little effort, but that doesn't buy you much in the long run. Instead, for now, just make sure you build anything with your money so that it doesn't sit around in your bank account. Good things to spend your money on include: 1) more units 2) more unit-producing buildings like rax/gateway/hatcheries 3) upgrades 4) expansions (building a CC/Nexus/hatchery near another group of minerals).
For Protoss, for example, you'll start off making probes and probably build your first gateway. When that gateway is done, start making zealots out of it while you're still making probes and pylons. At some point, you'll have enough money to build another gateway, an assimilator for gas, or a cybernetics core for tech. Doesn't matter what you make, just choose one of those and build it. In later game, the principle is usually:
1) Look at your production buildings, make sure they're all making something (zealot, goon, carrier, whatever).
2) If they are all making units and you still have enough money for another production building, start on one.
It's not uncommon in later game for you to have 10 gateways by this principle, especially if you're playing on large maps with lots of resources.
As a general principle, try to avoid queueing for creating units as Terran/Toss, because this ties up resources. If you queue up 5 SCVs in your CC, that means there is 4 SCV's worth of minerals that's waiting to be used but is not actively being used, this is 200/0 worth of resources that could be going towards your next barraks or supply depot immediately. For a more extreme example, suppose you have one starport with 5 battlecruisers queued up. Four of those battlecruisers aren't actually being built so the queue is just tying up resources. You should instead have 3 starports building 1 battlecruiser each.
Queues are a convenience but have their drawbacks. In late game, you can't be producing at 100% efficiency such that you start a new unit everytime the previous is finished, and you'll probably have some surplus of resources so queueing may not tie up your resources significantly. But at least in early game, try to never have more than 2 units, even if that, in your production queue.
As you do get better and start figuring out what to build, that's when scouting becomes important. You send an early scout out for two reasons. One is to start getting to know the map, even from basics such as whether you're on a land or island. The second is to find your opponent's base and get some idea of what he's doing, such as what race he is. In advanced play, recon is very important, as most decisions are not made blindly. Heavy emphasis is placed on figuring out some way to take a peek at what your opponent is doing and then to adjust what you're doing based on that.
Static defense is all structures that have an attack. A common question is, "Why not build a lot of cannons and static defense? Cannons are great units -- they detect, fire at land and air, have awesome range, and straight up will kill units better than most other Toss units for the same cost." Cannons are deliberately made better than mobile units where the tradeoff is that they DON'T MOVE. You cannot beat your opponent by having a well defended base. A good opponent will see this and just leave you alone while taking expansion spots around the rest of the map. Remember that at your level of play, economics has a huge impact on winning or losing. If your opponent has 2 bases and you have 1, it doesn't matter how well defended your base is, you're going to lose in the long run.
Sadly, playing single player missions tends to encourage mass static defense. Building cannons works great against the computer because it'll just mindlessly throw forces at your cannons and get them slaughtered. Some single player missions are in fact very hard to beat without building static defense. This encourages very bad habits for multiplayer. The computer isn't smart enough to figure out that it should just ignore your cannons. You can build a bunch of tanks and bunkers and then go out to lunch, and when you come back the computer will have no units and no resources because it wasted them all against your defenses. Players will not do this, though.
Getting more concrete, suppose you start out building 3 cannons early. That's great defense and your opponent will see this when he scouts you. This is good news for him though because he knows you've been spending money on defense versus mobile units so he's safe from being attacked by you. 3 cannons is a bit more cost than a Nexus, so this means that if you both are getting resources at the same rate, he can for that same money plop down a Nexus at the nearest resource node for free. Now you're essentially going to be fighting one base against two, and you will lose unless out of some miracle.
I will now address the issue of map choice, as an addendum in response to comments below. My intention is not to make a strong statement that will cause controversy, but to share some knowledge that can help you make informed decisions.
There are generally two broad classifications of maps, so-called normal maps versus money maps. Money maps have overwhelming resources, enough that you will never have to worry about expanding or running dry. This includes zero clutter map, big game hunters, and fastest maps. Money maps simplify the game in many ways because you can focus just on one base strategies. And it makes some parts of basic skills more challenging because, for example, it's much hardeer to keep your resources near zero and it's not atypical to max out on 200 supply and still have near infinite amount of resources. On the other hand, money maps tend to reward strategies that are typically not feasible in other games, such as turtling (building a lot of static defense) and teching to capital units such as carriers, i.e. "turtle & tech".
If you folow the Big Five principle on those maps, you should still have an advantage over someone who does not because e.g. he builds many cannons. But, it is far easier at a beginner level to win by turtle & tech than to play conventionally, so money maps tend to favor these kinds of tactics at that level.
The type of map and experience you choose for yourself is up to you, and e.g. the vast majority of recreational Starcraft players prefer money maps because those simply the game to an extent that they don't have to learn some aspects of the Big Five. My only strong recommendation is that you start with land maps, since island maps add a whole new level of complexity and at beginner level usually involve memorizing some canned build order and attack regardless of the situation.
One of the frustrating parts about Starcraft is learning to defend against rushes. Even new players can learn some build order that will work against most players since some attacks are easier to execute than defend against. A common one is a zergling rush where you don't make any drones and just build an early spawning pool, this is a "do or die" attack but it often works against recreational players. If you follow the Big Five, you should be well equipped to deal with any early game attack. If you are attacked before you have many units, use your workers for defense. If you are attacked and have units but don't have enough, the solution is NEVER to build static defense. Watch the replay and see how he got so many units, it's likely that he either has better economy management, or more unit producing structures. For example, if you build a cybernetics core and gas for tech and he builds 2 gateways, then you will probably get run over by superior forces. Just modify your build order next time so that you don't tech as fast, but instead make early units. Just remember, there is NEVER a need a to build static defense to stop early rushes in recreational level. You should always be able to adjust your build or use workers.
If you want to learn more about advanced tactics, I would recommend watching replays on sites such as WGTour
( http://www.wgtour.com/rep.php?datab=broodwar ).
Just click on the replays and save them to your "maps" directory. Watch how the pros play, and ask questions here about things you don't understand and we'll be happy to explain.
Overall, pay attention to your resources and learn by watching your opponent's replays when you get beat. Ask questions here for anything you're confused about. Check out the Starcraft Compendium
( http://www.battle.net/scc/ )
since it overall has pretty good advice. Get good at the Big Five. Watch your replays afterward to make sure you are actually building peons nonstop and that you're not letting money stockpile. Good luck.