|
Eh, I did it right, unit testing isn't 'right', they don't sit on a nice grid so it is extremly imprecise. Even if the ultralisk is attacking at slightly the wrong angle, it will distort the splash area to be very confusing, and ultras don't like to sit next to units and not attack them. You can tell from the images above that the results of this test are far more precise than any unit test could possibly be...
Anyway, I've already invested more than 4 hours into the testing of this ultra damage, actual in-game not counting this thread, some of which involved units, but I'm not going to share those results because they are so inaccurate.
My only desire is that Zelniq would post his results... no single test has any validity, no matter what variables you change, there is always observer bias & experimental errors.
|
If nothing else, this change makes more sense visually.
|
A couple posts i would like to respond to, but have forgotten where they are exactly. So I'll just post it as a block of text instead. First I would like to say that the zerg's ability to remax their army faster than Terran or protoss is balanced by the fact that a maxed Terran or Protoss army is stronger than a maxed Zerg army. Ling Ultra dies to marauder if you add in 2-3 blue flamed helions without infestor support. And even with it, it's quite ridiculous how much better marauders do than they should. Ultras really only counter thors (or vikings after a BL tech switch), which is really the only time ultras are successfully used against Terran. Against Protoss they have some use breaking down forcefields, but I think with the buffs to protoss air, less and less Ultras and BLs are going to be seen.
|
On December 07 2010 15:19 FabledIntegral wrote: Zerg can just take too many bases and rebuild its army too quickly. I'm talking about a late game scenario btw, not midgame or midlate game or any sort of "ultra rush."
EDIT: I should also clarify, a major point in Ultra strength relies in its ability to be mass produced. So if you throw mass ultra at the enemy, and were not even cost efficient, in the late game stages you can just rebuild all those ultra from all the larvae and stomp over the new army that will be much weaker, and just continue to stream those units in. And you absolutely have to include the fact that ultras are made from larvae as making it a much stronger asset to the Zerg arsenal.
Your argument is faulty, because you are saying Ultras are good because they can be mass produced; this is a race-specific advantage, not a unit-specific one. You can mass produce and streamline any unit, not just Ultras. If a Z goes mass Ultras and wins, it was a game the Z would've won anyways. Feels like you build it just because you can... What purpose does it serve?
|
Maybe they will be a little more manoeuvrable if they are indeed a little smaller. edit: Didn't notice the six extra pages apart from the front one, this might have been discussed >.<
|
Not that anybody really uses ultras anyway. They're like the mothership of zerg, except they are useful 2% of the time instead of 1%
|
On December 08 2010 06:54 GoldenH wrote:Show nested quote +On December 08 2010 06:16 Mystgun wrote: Feel free to flame me if I am misintrepeting something, but how does this test give any conclusive evidence that ultras have been nerfed if we are testing it on a 1x1 sensor tower? You are only getting a granularity of 1 grid and while the ultralisk splash range is 2.5. If the splash center was moved by 0.5 grid its possible that testing with sensor tower is not capturing the full effects of the changes Taken by itself the sensor tower experiment isn't enough, but combined with the pylon test it is. Pylon is 2x2, and two pylons in a row were hit by the same ultra attack. Yet Sensor Tower is 1x1, again, only two Sensor Towers in a row were hit by the same ultra attack. Also, the Sensor Towers in the corner were hit, but the Pylons in the corners were not. This demonstrates that the ultra damage range is exactly 2 in both versions.
The data is right, but your conclusion is just not logical. If a Sensor Tower were 1x1, and a Pylon 2x2, this means that the same space that is filled by a pylon is filled by 4 sensor towers, which in turn means that there would never be a situation where in a pylon-based grid there is damage done to a pylon where in a ST-based grid there would not be damage done to at least one of the STs.
However, as can be seen on the picture in the OP, in a straight line in front of the ultralisk, two pylons are hit, but also only 2 STs, as you stated. The only logical conclusion is that a pylon is larger than two STs, or at least the hitbox of the pylon is larger than the box made from the outmost edges of a 2x2 ST building area.
This means that the Hitbox of buildings is smaller than the surface area they block from building.
The area they block from units moving seems to be even another thing, if an Ultra before 1.2.0 could fit between two pylons, but not between two STs.
So, finally, this means that a building (and probably any other unit) actually has 3 different types of size, which are usually in a similar ballpark, but not necessarily the same. They have an area they block from building, which is and exact amount of 1x1 squares. They have an area they block from the moving of ground-units, and they have an area which must be hit by AoE effects for them to take damage. But one should not carelessly assume that all three of these areas are the same area that is blocked from moving. The data above shows that they most probably are not.
It is also important to note that the splash of the ultralisks obviously does not hit a rectangle area, since, again, only the middle one of the pylons in the second line got hit. It is possibly some kind of ellipse, but that can not be exactly told from this data.
Nonetheless, the data in this thread indicates strongly that the area of effect is indead smaller after the patch than it was before, since with similar experiments, areas that were hit before are now not hit anymore, but no area that was not hit before was hit after. Theoretically, this could be because the margin of error by chance only affected the area where there would be a gain on surface, but that is not very likely, since one can safely assume that the area cowered by a tower is not that different from its hitbox.
|
Would you mind testing with missile turrets? There was that additional bug regarding ultras attacking multiple sensor towers.
|
On December 07 2010 19:07 Everlong wrote:Show nested quote +On December 07 2010 19:02 diegonolan wrote: Do any of you morons understand what people are saying? If a zerg can go BLs to ultralisks he's already won. He could have made any unit and won. Its like saying well why don't terran go mass BCs to mass thor. If i go bc is he has to make corrupters and then those corrupters can't even attack my thors, Its an ingenious strat! If you are able to do that then you already one. The tech switch between BL and Ultra is not what wins the game. Im sorry, but you are totally wrong.. Please, dont post if you have nothing to cover your silly arguments with. Even your god (Idra?) says Ultras are good for tech switches between Broods, wake up..
First off don't double post. Edit your posts so that they include what you have to say in 1 post. Secondly, you really need to stop flaming people. I've read your posts and every single one of them in this thread has been you flaming someone saying you're wrong cuz of this, you're stupid because of this... seriously just stop. Say what you have to say and don't tag an insult behind it.
Final thing is that you didn't read his post when you responded. He never said that ultras are not good for tech switches between brood lords. His post simply highlights the fact that if you get enough money to switch between brood lords and ultralisks you've already won the match. If you can loose 10 ultralisks (3000/2000) then immediately switch to 6 brood lords and 12 corrupters (2700/2100). Then tell me you haven't already won. Where do you get the resources to immediately spend 2700 minerals and 2100 gas right off the bat? You must have like 4+ bases by then. If you can do that tech switch, micro properly, and still loose the game then I'd say there's something really wrong with terran and that terran needs to be nerfed seriously. In any case it doesn't happen because he's right. When you reach that level of economic excess then most of the time zerg does just win the game.
|
On December 10 2010 06:16 Simberto wrote:Show nested quote +On December 08 2010 06:54 GoldenH wrote:On December 08 2010 06:16 Mystgun wrote: Feel free to flame me if I am misintrepeting something, but how does this test give any conclusive evidence that ultras have been nerfed if we are testing it on a 1x1 sensor tower? You are only getting a granularity of 1 grid and while the ultralisk splash range is 2.5. If the splash center was moved by 0.5 grid its possible that testing with sensor tower is not capturing the full effects of the changes Taken by itself the sensor tower experiment isn't enough, but combined with the pylon test it is. Pylon is 2x2, and two pylons in a row were hit by the same ultra attack. Yet Sensor Tower is 1x1, again, only two Sensor Towers in a row were hit by the same ultra attack. Also, the Sensor Towers in the corner were hit, but the Pylons in the corners were not. This demonstrates that the ultra damage range is exactly 2 in both versions. The data is right, but your conclusion is just not logical. If a Sensor Tower were 1x1, and a Pylon 2x2, this means that the same space that is filled by a pylon is filled by 4 sensor towers, which in turn means that there would never be a situation where in a pylon-based grid there is damage done to a pylon where in a ST-based grid there would not be damage done to at least one of the STs. However, as can be seen on the picture in the OP, in a straight line in front of the ultralisk, two pylons are hit, but also only 2 STs, as you stated. The only logical conclusion is that a pylon is larger than two STs, or at least the hitbox of the pylon is larger than the box made from the outmost edges of a 2x2 ST building area. This means that the Hitbox of buildings is smaller than the surface area they block from building. The area they block from units moving seems to be even another thing, if an Ultra before 1.2.0 could fit between two pylons, but not between two STs. So, finally, this means that a building (and probably any other unit) actually has 3 different types of size, which are usually in a similar ballpark, but not necessarily the same. They have an area they block from building, which is and exact amount of 1x1 squares. They have an area they block from the moving of ground-units, and they have an area which must be hit by AoE effects for them to take damage. But one should not carelessly assume that all three of these areas are the same area that is blocked from moving. The data above shows that they most probably are not. It is also important to note that the splash of the ultralisks obviously does not hit a rectangle area, since, again, only the middle one of the pylons in the second line got hit. It is possibly some kind of ellipse, but that can not be exactly told from this data. Nonetheless, the data in this thread indicates strongly that the area of effect is indead smaller after the patch than it was before, since with similar experiments, areas that were hit before are now not hit anymore, but no area that was not hit before was hit after. Theoretically, this could be because the margin of error by chance only affected the area where there would be a gain on surface, but that is not very likely, since one can safely assume that the area cowered by a tower is not that different from its hitbox.
Even if true, would not the margin of error remain the same?
I think we can both agree that the ultra can get closer to a sensor tower than it can to a pylon. And that a pylon's hitbox is larger than a sensor tower's hitbox. Therefore if the ultra is getting closer to the sensor tower, it is also getting closer to the three-deep sensor tower. You'll notice I put in a bit of yellow on the other side of the boxes. Since I centered the ultras on top of each other, it does not matter if the building grids are exactly lined up, only the farthest hitboxes of the undamaged buildings matter for determining range. I am in fact measuring from the closest the ultra could get to each building to the farthest hitbox its attack could possibly effect.
BTW I tried for a while to extract the SC2Mod data from the PTR MPQs, but I couldn't get it to load, if anyone has had success let me know.
Finally as far as further tests go, if you want more tests you will have to do them yourself. I recommend that you have a friend clump together burrowed zerglings and attack the one in the middle a couple times. Then draw a shape of the hit zerglings and post it here. This should be more accurate than any test with buildings or other unit (marines for example). It is important that people other than me post results of tests. It will make any conclusions stronger.
|
On December 09 2010 12:11 kckkryptonite wrote:Show nested quote +On December 07 2010 15:19 FabledIntegral wrote: Zerg can just take too many bases and rebuild its army too quickly. I'm talking about a late game scenario btw, not midgame or midlate game or any sort of "ultra rush."
EDIT: I should also clarify, a major point in Ultra strength relies in its ability to be mass produced. So if you throw mass ultra at the enemy, and were not even cost efficient, in the late game stages you can just rebuild all those ultra from all the larvae and stomp over the new army that will be much weaker, and just continue to stream those units in. And you absolutely have to include the fact that ultras are made from larvae as making it a much stronger asset to the Zerg arsenal. Your argument is faulty, because you are saying Ultras are good because they can be mass produced; this is a race-specific advantage, not a unit-specific one. You can mass produce and streamline any unit, not just Ultras. If a Z goes mass Ultras and wins, it was a game the Z would've won anyways. Feels like you build it just because you can... What purpose does it serve?
It doesn't make my argument faulty whatsoever when it has relevance to balance, lmao. Maybe because mass produced units doesn't prove to be a problem until it happens with ultralisks? Jesus christ use your head.
I don't even understand what you're saying "it was a game the Z would've won anyways." Entirely untrue and blanket statement.
|
Testing on small units like marines and zerglings is completely valid. pylons are 2x2 in size, marines would be 1x1. It is entirely possible that two rows of marines maybe hit in front. Id test it out but im away from my gaming computer for a week yet. Dont be to quick to mention the ban hammer.
Thanks for the hard work, Adun.
|
Please post your results, I look forward to them 
I have been testing other things... most recently, I came across the idea that the Vortex change so that it ignored missiles might mean that Interceptors could fly through it now, making carrier+mothership viable, but no such luck... I'll continue pondering the patch notes and post if I find anything exciting.
|
carrier+mothership viable
Try making either of those 2 units viable, those are easily the 2 most exotic units in the game, just like dark archon in sc1. Carriers have only been build once in 3 seasons of GSL. I would love to see more of those 2 units, but for now i think its just wishful thinking.
Thanks for the post, Adun
|
|
Ultras are supposed to be one of the most powerful units and they're reduced to this in SC 2. There's just too many lower tier units that crush ultras.
|
^^ I think thats what blizzard wants. They said somethign similar when they nerfed the siege tanks. Same with thors, carriers, and, colossi(without support), They all just die so fast, to even lower tier units.
|
If the sensor tower portion is true, isn't that a buff? It's hitting towers behind other towers?
|
Looks like ultras were slightly buffed in the final 1.2 patch.
![[image loading]](http://hammerand.com/goldenh/sc2/ultra1.2final.jpg)
First image, the same damage radius as pre-1.2 but showing more damage underneath the ultra.
![[image loading]](http://hammerand.com/goldenh/sc2/ultra1.2final-burrowedzerglins.jpg)
Damage radius on burrowed zerglings for those who obsess over unit size.
![[image loading]](http://hammerand.com/goldenh/sc2/ultra1.2final-burrowedzerglinsandbuildingsize.jpg)
Testing a building to see if damage is radiated from a building. answer is no. Damage seems to be radiated from attack point.
![[image loading]](http://hammerand.com/goldenh/sc2/ultra1.2final-burrowedzerglinsandunitsize.jpg)
Testing on a unit to see if units and buildings are handled differently. Answer is no.
conclusion: Ultralisk damage now does a circular damage area, instead of an arc shaped damage area. This could help against burrowed roaches or banelings but otherwise, its kind of pointless, but at least it's not a nerf. Good to know my testing resulted in a positive result.
|
So if I'm reading you correctly you are saying that in live 1.2 the splash range of the ultra is about the same as before, but it now does damage below the front of itself as well?
|
|
|
|