As I had mentioned previously, I will do some guides for beginners. Scouting seems like a good topic as it has often been misunderstood. This post covers the basics, and does not go in details for match up specific scouting.
Unlike chess, Starcraft is a game of imperfect information. Gathering information of your opponent and hiding your own information from him or her play a critical part to the game. I won’t talk about how to trick your opponent by giving misleading information, but rather focus on the basics of scouting your opponent.
Quite often I see “help-me” threads on forums regarding how to defend against certain timing or build etc., and the usual answer is along the line of “you need to scout it”. Indeed, knowing is half the battle. However, many have misunderstood the logic for scouting. It is not about seeing what opponent is doing, but rather seeing what opponent cannot do. In another words, it is an art of deduction.
“Eliminate all other factors, and the one which remains must be the truth.” – Sherlock Holmes (The Sign of Four)
For example, in TvZ, you did not see a Hatchery at the natural when you first scout. You can deduce that Zerg is unlikely to have gone for a Hatchery first build (assuming no Hatchery at your base or natural). Next, think about what Zerg can do without expanding at the natural early on. Let say, you scouted the main and saw the Spawning Pool and gas taken. Subsequently, you can deduce that Zerg did not hide a Hatchery because there would not have been enough mineral for the Spawning Pool and the Extractor. There can be three main types of one base play: early Roach, one-base Baneling burst and Speedling into expand. Try to deduce further, by looking at the Spawning Pool, is Zergling speed upgrading? This should tell you what opponent cannot do. The one build that is left is the answer.
Another example, you often see Terran pro-players position six Marines at the mineral line against Protoss even when they do not see a Stargate. This is because the information gathered (e.g., two Assiminator) does not allow you to cross out the possibility of a Stargate opening. If you only prepare for an Oracle after you scouted the Stargate, you are in trouble. In line with this thought, Terran pro-players will move the Marines away from the mineral line after a certain period of time. This is again seems like an appropriate reaction based on deduction. If Protoss made an Oracle, it would have arrived in my base now. It did not, so I deduce that opponent did not open up with Stargate into Oracle. With that being said, one’s overconfidence in deduction can be a weakness. For example, MC likes to make an Oracle, and does not send it to the Terran base. He only sends it in after the usual Oracle time is over. In fact, you can argue that the Terran did not exactly use deduction, but presumed that an Oracle will be sent over immediately after a standard Stargate opening. Deduction is an art with depth.
Have a purpose
You need to have a purpose when you scout. This is extremely important.
I have played against many relatively newbie friends who always make this mistake. They will scan my base randomly, and I would ask why they did that. Their answer was simply “I want to see what you are doing.” I then asked, “so what information did you gather?” The answer is nothing. You need to be looking for something when you scout.
Plan your scouting by having an aim with each scout. For example, at 7:00 in TvP, you scout the natural of the Protoss. The aim is to check the Assiminator at the natural. If they are not taken, then you can deduce that Protoss is not tech-ing or upgrading. So expect an attacking coming soon, while checking for a third base. Another example, in TvT, you can choose to do a scan with your second 50 energy because that is the time the buildings would have been down for any build. The aim would have been to check if there are two Refineries taken, and whether there is a Starport. The add-on should allow you to deduce away some builds.
All game long
You do not only scout at the opening phase. Scouting is an all game long thing. Although the opening phase scout gives you more information than any other phase of the game, you should keep scouting whenever you have gaps in your information. Scouting is not limited to deducing opponent’s opening and build order. It includes army positioning and intention. For instance, you move a Scv to Protoss’ natural at 7:00 and you see no units. Then Protoss may probably have the units at your side of the map and is warping in offensively.
To make things easier, it is good to have a pre-planned scouting timing for each match up. For example, in TvP, I have three main scout timing. The early Scv scout, 5:00 and 7:00. I will do a post regarding TvP scouting in the near future.
First worker scout
I have been asked when you should send your first Scv scout. It depends on the match up, map and build.
If the map is a two player map, you should scout at 14 or 15 supply against Protoss and Terran. Scv scouting is optional against Zerg on a two player map, if you do, 15 or 16 is fine. Against Protoss, a 14 Scv scout gives you the opportunity to put down the Engineering Bay at the natural to block the Nexus. Against Terran, 14 or 15 makes a big difference on a two player map. Opponent’s first Marine (if he did a 12 Barracks) can deny your 15 scout Scv from getting into the base. Scouting the gas timing is extremely important in TvT, as it allows you to deduce the opening.
On a multiple player map, I usually do a 15 for all match up just for spawning position. Personal preference.
I’m referring to the early game cheese proxy, like two Barracks or two Gateways. Your first Scv scout can look behind the natural and then go straight to opponent’s base. It is not effective to cover every plausible proxy location near your base. It takes much time and luck to find the proxy. By the deduction logic, you can never deduce that the opponent did not do a proxy because you did not scout it. At best you can deduce that the opponent did not put down the proxy at the scouted locations.
As you send the Scv to opponent’s base at a normal time, you should look for the standard structures. Let say, there is no Gateway, then you can deduce that there is likely to be a proxy even though you have not scouted the proxy itself. At that point, the location of the proxy is not important, because you react similarly. Therefore, it is better to scout opponent’s base to “look for proxy” than literally sending a worker around to look for it. The reason to check the natural is that it takes little time away from the main scouting task to spot a plausible location. Further, if that is TvT, you can try to kill the building Scv.