This post is a sequel to the previous post (scouting basics), whereby we look at how to hide information from the opponent. It is essentially about scouting, but from the receiving-end perspective.
Clearly, as far as possible, it is best to completely deny scout. But that is too idealistic. This post is about hiding information from the opponent by thinking from the opponent’s perspective. It is not about giving fake information, but simply enlarging the gaps in opponent’s deductive thinking.
First, we need to know what cannot be hidden from the opponent. It is not possible to go Command Centre first and expect not to be scouted. The very opener choice is almost impossible to hide if the opponent does a standard scout timing. Sometimes opponent may not scout the optimal direction in a four player map, but still it is likely that opponent knows. So if we decide to do a Command Centre first, then don’t stress about opponent knowing it. The same applies for choosing to go Refinery first or Barracks first. Simple. This post focuses on the next two points.
Second, we need to know what the opponent is looking for. It is about thinking from the opponent’s perspective. This depends on match up.
Third, how to hide it the information. It is not as simple as not letting the opponent sees the specific building. This is where minor details can make a big difference.
There are two main things to consider in terms of scouting against Terran. The first is the Refinery (timing and number). The second is the building and add-on (timing and positioning).
The timing of the first Refinery gives much information on what build Terran is doing, and it allows the opponent to deduce away many things. Whether Terran take the second Refinery at a specific time is another big telling sign. Importantly, you cannot hide the timing of the first Refinery.
Regardless of the opening, the key is to hide the second Refinery. Even if you do not take it, don’t the opponent knows it. How can opponent scout it? Obviously, if opponent uses a scan, there is nothing you can do about it. Reaper scout is what we should play around. I am not going to discuss about how to position your Marines to prevent the Reaper from scouting. The important thing is the choice of the first Refinery. Like I said, you almost cannot prevent opponent from scouting the first Refinery, so you should build the first Refinery on the geyser that is more vulnerable to Reaper scout (map dependent). This minimises the chance of the Reaper from checking the second Refinery.
The time frame when the Factory is building or just complete is the point whereby Terran will reveal the choice of build order path. It can be a third Command Centre, Banshee, Hellbat, Viking, early drop or even blue flame. The second Refinery allows Zerg to deduce whether possible Banshees or Hellbats are coming. Unless the Zerglings somehow manage to sneak in, Overlord is Zerg’s only way to get information. Using the same logic above, the first Refinery should be built on the geyser that is more vulnerable to Overlord scout.
An example on Overgrowth
The timing of the second Refinery does not really matter in this match up. Rather the key is whether you pull Scv off the Refinery after Reaper expand, as it gives away the follow up. If two or three Scvs are pulled off, it is unlikely to be a Factory follow up. The Probe can stay barely long enough to see the number of Scvs pulled off, while get out of the base before the Reaper can catch it. We can pull one Scv of the Refinery (two mining gas) to make it ambiguous for the Protoss. You can remove the rest of the Scvs or pull the third one back depending on your build choice, after the Probe left. The amount of gas mined by doing this may not be perfect for your build, but it creates gap in Protoss’ information gathered. If the Probe is sacrificed by staying in the base to count the Scvs until the Reaper appears, then that is the price on his or her side.
The basic idea is to hide the building from opponent’s vision. However, most key buildings that give away the build order are usually built near to an add-on, so it is hard to really hide the building completely. Stemming down from this thought on the position restriction, there are still details to consider. Look at the following two examples,
Both did a Reactor Hellion build against Protoss. Taeja’s Factory position is better than Bomber’s in terms of hiding information. Like I have mentioned earlier, the Factory placement is restricted to be placed near the Reactor. However, if the Factory is placed on the left of the Barracks, it signals to the opponent that the Factory will be swapped onto the Reactor. This is because the Factory cannot build an add-on, since the Barracks is on its right. How does that matter?
The two main variations in TvP after the opening is whether you choose to put down additional Barracks or tech up to 1-1-1. Thus, a Probe may try to sneak in at that 4:30 timing to check the build choice. In the above scenario, let say the Probe scouted the Factory. This information tells the Protoss that Terran is likely to either do a Reactor Hellion or a Widow Mine Marine Medivac attack. The difference between these two is whether the Factory is swapped onto the Reactor. For the Widow Mine build path, the Factory will later build a Tech Lab after two Widow Mines are produced, while a Barracks is built next to it. This is to shape back to the convergent point of three Barracks bio production. Therefore, if the Factory was initially built on the left of the Barracks, it is inconvenient to put down the Tech Lab while microing at the other end later. Then, logically, building the Factory on the left of the Barracks is suggesting to the opponent that we will be doing the Reactor Hellion, as the Factory is not in the optimal position for the Widow Mine build.
With the Factory being placed vertically next to the Barracks, it creates gaps in Protoss’ information because it can go either way. Normally, it is a good habit to build the buildings vertically next to each other.
In a standard TvZ opening, Reactor Hellion is a key early convergent point. Right after the Factory is swapped onto the Reactor, the Barracks needs to be landed for Marine production or Tech Lab. It is a good idea to place the Barracks slightly farther away from the ramp away from the Factory. Again, this enlarges the gap in opponent’s information. By placing the Barracks next to the Factory, it gives away the information that we are unlikely to get Banshee. This is because the Barracks is used to build a Tech Lab for the Starport. The Barracks is therefore often landed next to the “hidden” Starport, and not next to the Factory. Therefore, if the opponent sacrificed a few Zerglings to look at the main ramp and the Barracks is not seen, Zerg cannot be sure if Banshee is coming.
All these details I mentioned above do not change our build at all, but it creates gaps in opponent’s deductive thinking.
4 thoughts on “Hiding information”
As a clan we are using BTSync as a private cloud for our members/admins to share StarCraft Stuff (like programs, links, videos, music, images etc) so they can stay up-to-date with our big StarCraft “Member Package” (~700MB).
I am willing to share it with you. I would gladly send you an one time key to access it if you want!
Keep up your great work! As a Terran player I really appreciate your effort!
gl hf gg
Sure, thanks. I won’t mind getting more exposure to other materials.
I think your TvZ and TvP refinery analysis is good, but I disagree about TvT. If you opponent is willing to SCV scout, they can deduce refinery timing from the gas mined. Because of that, I think forcing the opponent to travel a bit further for a scout is more important than the small likelihood the second refinery is relevant and your opponent just so happens to manage to scout it (which shouldn’t happen, without a reaper sacrifice or a wonky reaper timing). For example:
gas first (9) – 2 fact hellion or 2 fact widow mine. harass with scv and check for gas pull to confirm, 2nd gas will not be taken
gas first (11) – you already know something is up, and its not banshee. the addon that the barracks creates should tell you most of what you need. tech lab = very fast tank push, reactor = hellbat all in/hellion elevator, none = widow mine drop/hellbat all in/hellion elevator. I can’t think of a single 11 gas build that opts for an early 2nd refinery
reaper (11) – what’s important is the presence or lack of second barracks
gas first (12) – all good 12 gas builds get a 2nd refinery around 35-40% factory completion. of course, you can scout gas first hellion elevator or gas first widow mine drop if there is no second refinery, but both are terrible builds that hit at easy to hold timings
reaper (12) – same as reaper (11)
gas (13) – only case 2nd gas location is useful for.
gas (14) – antiquated wol build. likely 2 rax pressure or safe 1 marine expand. will move out before 2nd gas is made. 2nd gas might be useful to identify if build is all in or not.
gas (15) – 15 gas expand, you will never be able to scout 2nd refinery without a scan except on maps like secret spring and catallena. even so, a poke at the front to reveals almost everything
Another important detail to note is that refineries from 6 to 9 o’ clock and 12 to 3 o’ clock position give more gas than refineries on other positions.
Good point! In fact, I had considered this too. But from the perspective of both players, it seems like it is difficult to prevent the first gas timing to be scouted unless you wall off early on purpose. Then, the question is, is it worthy to make the scouting scv to travel a bit more to see it. Most of the time, I will say no. This is because if it is 12 Barracks into Marine, the first Marine will either deny the scv from entering in the first place or kill the scv if it comes in. The trade off of the second Refinery is relatively higher.