|
Since Phase 2 started, I've decided to dabble in the other two races, Protoss being the final one that I've yet to pick up.
Now that I'm starting, I always wondered how exactly the Colossi's attack works. I decided to investigate the matter and thought I'd share my findings with Team Liquid.
This might be super basic to those of you who have been playing Protoss for a while - I honestly have no idea. I'm hoping this will be vaguely helpful to someone out there. If anything, you get to see some pretty pictures of lasers.
The Test
Take a big group of marines and move them towards the Colossi, then have the Colossi fire on the moving group. I've repeated the test moving the marines in random directions after the beam fired and with range. Both support the findings below.
Step 1: Damage Calculation
![[image loading]](http://i26.tinypic.com/wjuo0l.jpg)
The Colossi's attack animation begins and damage is calculated. The units which will receive damage are chosen at this time.
Step 2: Damage Dealing (Phase 1)
![[image loading]](http://i26.tinypic.com/2d0e7nq.png)
The attack deals damage first to the units on the outside, but this leading damage only proceeds the rest of the damage by a few frames.
Step 3: Damage Dealing (Phase 2)
![[image loading]](http://i27.tinypic.com/zj8j60.png)
The rest of the units are fully damaged as the beams sweep together.
Some interesting things to note:
Damage isn't calculated by what the animation touches, it's calculated the moment the attack animation starts. The animation itself is just for show. If you notice in Step 3, the beams are clearly in the middle of the rest of the marines, but only that one line of marines has received damage.
Also, damage is done exactly perpendicular to the direction the colossi is facing and extends in a very narrow line. This means that to maximize damage, you should position your colossi such that they're not flanking the opponent's concave, but instead standing behind your row of units (presumably tanking). Of course, maximizing damage isn't always what your priority is - sometimes you'd rather position your colossi in safer locations.
|
You'd think it would not be terribly hard for Blizzard to align the animation with the line of units that get damaged.
|
It used to be that the collossi didn't even do any damage at all until the animation crossed beams. unnoted fix in a recent patch.
|
are the marines or colossus moving, or does the animation actually sweep forward like that
also is it any different with the range upgrade?
|
No difference with the range upgrade, the marines are moving towards the colossi. The same thing happens if the colossi beam begins and you afterwards move the marines in some random direction - that one line of marines will still get hit.
|
one thing i don't understand out of the pictures is the amount of marines taking dmg.
in picture 2 the top marine in the 3rd row takes dmg. and the bottom marine is unharmed.
but in picture 3 it's the other way around.
i assume these screenies aren't made in the same attack?
(cpu time for taking and putting away a screeny to the HD is most likely longer than the attack animation)
|
On July 17 2010 05:13 Schurk wrote:
one thing i don't understand out of the pictures is the amount of marines taking dmg.
in picture 2 the top marine in the 3rd row takes dmg. and the bottom marine is unharmed.
but in picture 3 it's the other way around.
i assume these screenies aren't made in the same attack?
(cpu time for taking and putting away a screeny to the HD is most likely longer than the attack animation)
Right, these are different attacks, but the mechanics still stand.
|
huh that's interesting. I really hope Blizzard fixes it so that the beams lock onto the target so it looks a bit less silly.
|
On July 17 2010 05:18 Chairman Ray wrote:
huh that's interesting. I really hope Blizzard fixes it so that the beams lock onto the target so it looks a bit less silly.
yea and they were all about that visual clarity for the spectator issue in this game too.
|
17067 Posts
This is very interesting. Have you considered adding these results to liquipedia?
|
On July 17 2010 05:22 Empyrean wrote:
This is very interesting. Have you considered adding these results to liquipedia?
You know to be honest, I've always been a bit scared to touch the liquipedia, after all who am I to be editing that monster?
But I guess that defeats the purpose of a community run wiki
|
17067 Posts
On July 17 2010 05:24 FC.Strike wrote:Show nested quote +On July 17 2010 05:22 Empyrean wrote:
This is very interesting. Have you considered adding these results to liquipedia? You know to be honest, I've always been a bit scared to touch the liquipedia, after all who am I to be editing that monster? But I guess that defeats the purpose of a community run wiki
Actually, everyone is very much encouraged to contribute to the wiki. Having the userbase contribute to a strategy page that many people will view is not only helpful to the community in that one sense, but also invites other users to contribute as well. Contributing to liquipedia is one of the best ways to help the community.
EDIT: The formatting can be a little bit daunting, though. But still, this is a really informative thread and the information would be great for other users to know about even after this thread has run its course.
|
On July 17 2010 05:28 Empyrean wrote:Show nested quote +On July 17 2010 05:24 FC.Strike wrote:On July 17 2010 05:22 Empyrean wrote:
This is very interesting. Have you considered adding these results to liquipedia? You know to be honest, I've always been a bit scared to touch the liquipedia, after all who am I to be editing that monster? But I guess that defeats the purpose of a community run wiki Actually, everyone is very much encouraged to contribute to the wiki. Having the userbase contribute to a strategy page that many people will view is not only helpful to the community in that one sense, but also invites other users to contribute as well. Contributing to liquipedia is one of the best ways to help the community. EDIT: The formatting can be a little bit daunting, though. But still, this is a really informative thread and the information would be great for other users to know about even after this thread has run its course.
Alright, I'll look into it. I'm familiar with a few programming languages, and I'm sure the wiki can't be that hard to edit.
|
Something else to note about collosi it's possible to use them in a manner similiar to reavers, where you unload them they fire (only need to begin firing) and load them back up.
|
On July 17 2010 05:46 PrinceXizor wrote:
Something else to note about collosi it's possible to use them in a manner similiar to reavers, where you unload them they fire (only need to begin firing) and load them back up.
Does this mean you can do some sort of attack cancel move as soon as it does damage.
As in, attack, press s, attack again much faster? Or attack, move(skipping most of the attack animation), attack?
|
All of the damage is dealt by the time the beams first come together. At that point, you can cancel the animation if you want.
The damage is not actually entirely frontloaded like a reaver (as you can see in the photos).
|
Exellent post. I play toss at diamond level and always wondered but never investigated the full mechanics of the pew pew lazerz. I did notice that the animation didn't exactly line up with the units taking dmg (stim + speedlings its obvious) but this is good to know for micro purposes rather than "leading" for maximum dmg.
The only thing else I would like to know is where the attack starts, and its width/depth and how that applies to smaller (or bigger units). I have done some tests verse buildings and small ones like depos or pylons build next to each other will take full damage from the splash (correct me if I'm wrong). Since Collosus aren't great vs buildings selecting small buildings near other buildings is the most efficient path to destruction (unless you are targeting key buildings).
Maybe do a test against clumped units to see how many of X units would get hit and how much dmg it does
For example
Colossus hits up to five marines for 150dmg
Colossus hits up to three roaches for 90dmg
The units to include are obviously smaller ground units. This also brings the question to what to target, if you target a tank it will have less splash but if you target a marine next to a tank it can hit a lot more potentially.
I believe colossus take out supply/pylons faster or at the same speed as immortals, while having superior range and mobility. However sniping Nexi or bigger buildings will not be faster than immortals. Actually cost for cost immortals always win but with collo you can maximize dmg by doing this.
|
On July 17 2010 07:52 Ploppytheman wrote:
Exellent post. I play toss at diamond level and always wondered but never investigated the full mechanics of the pew pew lazerz. I did notice that the animation didn't exactly line up with the units taking dmg (stim + speedlings its obvious) but this is good to know for micro purposes rather than "leading" for maximum dmg.
The only thing else I would like to know is where the attack starts, and its width/depth and how that applies to smaller (or bigger units). I have done some tests verse buildings and small ones like depos or pylons build next to each other will take full damage from the splash (correct me if I'm wrong). Since Collosus aren't great vs buildings selecting small buildings near other buildings is the most efficient path to destruction (unless you are targeting key buildings).
Maybe do a test against clumped units to see how many of X units would get hit and how much dmg it does
For example
Colossus hits up to five marines for 150dmg
Colossus hits up to three roaches for 90dmg
The units to include are obviously smaller ground units. This also brings the question to what to target, if you target a tank it will have less splash but if you target a marine next to a tank it can hit a lot more potentially.
I believe colossus take out supply/pylons faster or at the same speed as immortals, while having superior range and mobility. However sniping Nexi or bigger buildings will not be faster than immortals. Actually cost for cost immortals always win but with collo you can maximize dmg by doing this.
That would take a lot of work, but is something to consider. I think for now the point to take from this post is to aim right in the middle of the fattest portion of the blob of stuff you want to kill.
Also colossi splash is indeed weird in that it's 100% splash in a straight line. I think that line is like a half unit wide, but that's only speculation.
|
Very helpful, thank you.
Blizzard should not / cannot "fix" the timing of the lasers. The reason is - it is meant to only affect one small line of units at a time. Once. Since the attack animation lasts some period of time, if it were to update which units it damages, it could affect much more than just a line, as long as those units move very fast through it. It also would require to spread the damage among all those moving units, and hence a single one would get less damage. It's not meant to be that weak; it must pick a line of units and then hit them very hard, even if they run away super fast. I guess they could update the animation, rather than the attack, and make the animation follow the median line of the units that were picked. Still won't be perfect, but would make it look a lot more realistic in the most cases. I see one problem with that - that it could pull the animation outside of its normal-looking range. Well, for this case they should limit the animation chasing, while still taking the damage nevertheless.
|
On July 17 2010 12:07 figq wrote:
Very helpful, thank you.
Blizzard should not / cannot "fix" the timing of the lasers. The reason is - it is meant to only affect one small line of units at a time. Once. Since the attack animation lasts some period of time, if it were to update which units it damages, it could affect much more than just a line, as long as those units move very fast through it. It also would require to spread the damage among all those moving units, and hence a single one would get less damage. It's not meant to be that weak; it must pick a line of units and then hit them very hard, even if they run away super fast. I guess they could update the animation, rather than the attack, and make the animation follow the median line of the units that were picked. Still won't be perfect, but would make it look a lot more realistic in the most cases. I see one problem with that - that it could pull the animation outside of its normal-looking range. Well, for this case they should limit the animation chasing, while still taking the damage nevertheless.
I actually think it would be awesome if the laser animation would chase the units away, sorta the way valkyrie missiles would follow units for 10 miles in SC:BW.
Could probably get annoying though 
|
On July 17 2010 11:51 FC.Strike wrote:Show nested quote +On July 17 2010 07:52 Ploppytheman wrote:
Exellent post. I play toss at diamond level and always wondered but never investigated the full mechanics of the pew pew lazerz. I did notice that the animation didn't exactly line up with the units taking dmg (stim + speedlings its obvious) but this is good to know for micro purposes rather than "leading" for maximum dmg.
The only thing else I would like to know is where the attack starts, and its width/depth and how that applies to smaller (or bigger units). I have done some tests verse buildings and small ones like depos or pylons build next to each other will take full damage from the splash (correct me if I'm wrong). Since Collosus aren't great vs buildings selecting small buildings near other buildings is the most efficient path to destruction (unless you are targeting key buildings).
Maybe do a test against clumped units to see how many of X units would get hit and how much dmg it does
For example
Colossus hits up to five marines for 150dmg
Colossus hits up to three roaches for 90dmg
The units to include are obviously smaller ground units. This also brings the question to what to target, if you target a tank it will have less splash but if you target a marine next to a tank it can hit a lot more potentially.
I believe colossus take out supply/pylons faster or at the same speed as immortals, while having superior range and mobility. However sniping Nexi or bigger buildings will not be faster than immortals. Actually cost for cost immortals always win but with collo you can maximize dmg by doing this. That would take a lot of work, but is something to consider. I think for now the point to take from this post is to aim right in the middle of the fattest portion of the blob of stuff you want to kill. Also colossi splash is indeed weird in that it's 100% splash in a straight line. I think that line is like a half unit wide, but that's only speculation.
I'd do it myself but I don't have and SC2 friends and you can't test it on your own units heh.
|
Hmm... this is probably why my roaches get slaughtered by colossus. Thanks for the analysis! 
|
Sweet information!
If the attack is calculated at the start of the animation, if you run stimmed marines past a col does this mean any unit caught in the "line" of fire gets damaged, even if they aren't in that "line" by the end of the shooting animation? I don't like that one bit.
|
i have played all the races in hundreds of games. colossus is the unit i have used the least because i dont know how it works... this kinda helps..but still...wtf colossus?
|
On July 17 2010 12:07 figq wrote:
Very helpful, thank you.
Blizzard should not / cannot "fix" the timing of the lasers. The reason is - it is meant to only affect one small line of units at a time. Once. Since the attack animation lasts some period of time, if it were to update which units it damages, it could affect much more than just a line, as long as those units move very fast through it. It also would require to spread the damage among all those moving units, and hence a single one would get less damage. It's not meant to be that weak; it must pick a line of units and then hit them very hard, even if they run away super fast. I guess they could update the animation, rather than the attack, and make the animation follow the median line of the units that were picked. Still won't be perfect, but would make it look a lot more realistic in the most cases. I see one problem with that - that it could pull the animation outside of its normal-looking range. Well, for this case they should limit the animation chasing, while still taking the damage nevertheless.
Actually I believe the animation is an after effect added on after the calculation has already been made. The calculation is not being made by following the animation, so there shouldn't be a problem with having the animation follow the correct units
|
On July 17 2010 05:18 Chairman Ray wrote:
huh that's interesting. I really hope Blizzard fixes it so that the beams lock onto the target so it looks a bit less silly.
It would be just as silly, since units can go in different directions, so the beams will jump around between each one of them and pass through other units that wouldn't be damaged, e.t.c.
|
On July 17 2010 16:48 Ozx wrote:
Actually I believe the animation is an after effect added on after the calculation has already been made. -Agreed. The calculation is not being made by following the animation, so there shouldn't be a problem with having the animation follow the correct units Yeah, except it's technical. The calculation made in the first moment is easy (because it's only made once), the calculation needed for the line to "chase" them is not that easy (will have to be repeated), and could still fail to look convincing.
For example: 12121 get affected; the 1s run up, the 2s run down. The line chases - who? So the line needs to re-calculate the median line of those units at each moment, depending on all their movements, and still could look as if it doesn't hit any of them.
I guess they could just add 1sec-stun, then we would see the effect we crave (I'd enjoy it too), but then this weapon could become too powerful. Especially since it will block groups from moving forward through the temporary stunned units.
|
On July 17 2010 16:00 zomgtossrush wrote:
Sweet information!
If the attack is calculated at the start of the animation, if you run stimmed marines past a col does this mean any unit caught in the "line" of fire gets damaged, even if they aren't in that "line" by the end of the shooting animation? I don't like that one bit.
Right, that's what it means. Once the attack animation starts, as long as the Colossi is still alive and still firing, the units calculated to take damage will take damage regardless of which direction they run and how fast.
|
Hm so this means colossi can deal a maximum of 30 x 5 damage against smaller units, and much less so against larger units that take much more space. Interesting and good to know, thanks!
|
|
|
|
|
|
EPT
