Tip of the day: Scouting

In Starcraft, you continually make decisions. You can, at any point, decide to stop mining with all of your probes, or attack with all of your units, or suddenly switch tech. This is quite a contrast to a turn based game, where players have limited windows in which they can make decisions. Starcraft, however, is real time, and you have nearly infinite opportunities to change the way your game is going.

Using a build order or just having a general plan you're working towards can reduce these decisions to a manageable amount. You know that you build a cyber core right after your gateway finishes and that you start warp gate research right after that. Some decisions, though, like when to get observers, are a bit fuzzier. You get some advantage by going warp prism first, but if your opponent has cloaking units and you do that, you can lose. How do you solve this dilemma? Scout!

You can send your initial scout in one of three timings:
a) well before your opponent's tech diverges
b) a bit before your opponent's tech diverges
c) not at all

Timing a has the advantage of spotting your opponent's gas timing. This will often be a strong signal about what he wants to do with his tech, so this can easily pay for the slight economic disadvantage this choice puts you at. Timing b is a bit more standard (except when you expect your opponent to wall off). You get a little bit more mining time in exchange for having your scout arrive later. If your opponent is truly doing something off-the-wall, you'll be a bit less prepared for it, but you should arrive in his base just when his cyber core or his barracks finishes. Scouting Z at this timing can be a bit trickier because of zerglings. You're going to lose your probe in a few seconds to marines or stalkers, but you get to either spot or delay any additional tech that your opponent is working on. This can be a bit less reliable than spotting his gas timings, since he can cancel after he kills the worker, but it's usually not too huge of an issue since he's down mins / gas from cancelling. Finally, timing c is the most powerful economically, since you don't waste workers or units, but you don't get to see what's going on in your opponent's base. This can be okay, though: if you're confident that you can deal with any tech your opponent might be getting and that you'll have an attack such that if he expands earlier than you he'll pay for it, it's fine to choose option c.

This brings us to an interesting point: it's never a bad thing to have a perfect image of what your opponent is doing, but unless you're waiting for him to tech / expand / attack so that you can do something in response, there's no need for constant scouting. In fact, there's no need to scout at ALL unless the information gained by scouting will inform a decision you're about to make. If no matter what your scout reveals you continue on with the same game-plan, then you need to re-consider why you're scouting at that point in the game and think about either changing your scout timing or changing your build. On the other hand, if you keep making a certain decision blind, with no knowledge of what your opponent is doing, and that knowledge could help you make the right decision, it's worth figuring out a timing where you can either poke in with some units or send in a scout to help you make the right choice.