Hi everyone!
I am trying to understand Overlords’ pillars placements and interactions.
Considering there’s no air vision provided, I would like to understand:
- Every pillar in the game is good? Or there are some pillars that are fake (i.e. they don’t hide the Overlord) And how to tell a pillar is good or is fake?
- Pillars prevent all ground units to spot the Overlord or are there ground units that can spot it anyway? I ask this point because I just read an old Reddit post in which some pillars were said to be good against Marines, but not against Stalkers (i.e. Stalkers can see the Overlord anyway). Is that possibile?
- Is it true that, in order to hide a Overlord above a pillar, you have to place it exactly at the centre?
- If yes, why not making all the visible surface hide the Overlord, instead of a tiny area? I feel this would make difficult to hide it especially on asymmetrical pillars like rocks.
- Is there any infographic about the current ladder map showing the “good” pillars’ locations?
I’m curious to hear from you
Thank you for your time and have a nice day/night!
Marco
Winter’s already answered most of the other questions, but to go a little further into this, it has to do with the way maps are made specifically, and the unit’s ability to path to certain areas.
Additionally, as mentioned, aesthetic rocks are not considered pillars; pillars are a function of the terrain creation, rather than doodad creation or structure generation such as for collapsible debris. As a result, there is a distinction between high-ground and low-ground for pillars, but generally not for other things.
As far as I know, nobody has created an infographic for good pillar locations; most players tend to go off general memory and game knowledge.
However, there are a couple things you can consider:
-
does the pillar provide you vision over attack paths or areas you might want to watch over?
-
Does the overlord placement give you access to any information you might not have otherwise had (Pillar over the natural expansion, for example) - or the ability to get that information (such as sending an overlord into the main base)?
1 Like
The best way to tell if it’s good is to watch replay and switch to your opponent view. If their units close to the pillar can see overlord despite it being above the pillar - then the pillar does not block vision of an air unit.
Overlords safe spot are by the way imbalanced because with them zerg gets too much vision on the map for free (until air units kill ovies) and this is more than enough to spot most attacks/pressure from opponent in advance. If map has plenty of safe pillars providing protection for overlords - this map is considered zerg favoured and should be removed from map pool/vetoed.
It’s probably a line intersection test. If the line from the observing unit to the unit being observed intersects a game entity marked as a line of sight blocker before intersecting the collision radius of the observed unit, then the unit is not observed and remains hidden. This is a logical or operation – if any ray from any observing unit under your control intersects the observed unit first, then the unit is visible. The scenario can be complicated due to the positioning and size of the line of sight blockers and the exact methods used for the intersection test. For example, if the shape of the line of sight blocker is a sphere, a rectangle, a line with zero or non zero width (and uncapped or capped). It’s complicated.
The unreliable vision is probably due to small area surrounded by line of sight blockers but that isn’t large enough to completely contain the unit’s collision radius – whether the line segment between the observer and the observed intersects the line of sight blocker first or not depends on the angle of line of sight and just how much of the observed unit’s collision area is protruding. Units that are fully contained can be totally hidden but units that are too large will always protrude from some angle and whether it is visible or not depends on the line of sight.
They probably precompute by breaking the map’s area into a grid and doing line of sight intersections to neighboring grid points. Then, the line intersection test isn’t needed because you can simply lookup the observer’s location in a table and it will say true or false that you can see that location from this location. This makes it even more unpredictable because the behavior could “snap” as a unit’s center crosses a boundary between grid points.
free, you kidding, it cost 100 minerals to make an overlord, on top of that they are your supply, so if they die your now down 8 supply, possibly supply blocking, and forcing to make another ovi at another 100 a pop. buddy it aint free, its very expensive for our vision, unlike scans sensor towers and free halucinations and or obsvers costing less than an ovi as well. and usually ovis are only good till a certain point in the game where most people start searching them out and picking them off. that means early game mostly, and when we lose them early thats even bigger hurt.
I’m only talking about maps with abundance of such spots. If there is only one close to your opponent base - that’s fine. But some maps have plenty of spots (not necessairly pillars) where overlord can be hidden. On top of that zerg has creep which is the best vision mechanic on the map. Combine these two and good luck surprising zerg with any attack.