Jump to content

Live Srm Damage Test Results


22 replies to this topic

#1 Amaris the Usurper

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 100 posts

Posted 15 March 2013 - 09:09 AM

Disclaimer: These numbers do not reflect how much damage an SRM always does when striking a given type of 'mech. The damage done, while in some cases very great, is highly dependent on the impact point. The problem is that missiles often do enormous damage when striking the torso and miniscule damage when striking the extremities (see the remarks in my earlier topic), when the vast majority of people would expect them to simply deal their indicated damage to the components struck, like other weapons.

Here is my original topic on Streak SRM damage testing: http://mwomercs.com/...results-inside/

Here, I present a collection of similar tests, but conducted in live matches.

I carried out all tests using SRM 2s while standing still at a range of 35 m. The test vehicles were a CPLT-A1 with six launchers, a RVN-4X with one launcher, and a RVN-3L with two launchers. All missiles appeared to hit the target. Generally, I fired missiles into what I perceived as the "center of area" (from the front perspective) of the torso section. (This is a bit of a judgement call, but most people will pick about the same point. Obviously, we can repeat the tests with more precise aimpoints if we really have to.) All deviations in this procedure are noted below.

Keep in mind that the tests in my earlier post were carried out with Streak SRM 2s, which select their own impact points. We can therefore expect these results to differ from the earlier ones, even if the mechanic for determining the damage inflicted per missile is the same.

COM-3A (96 armor, 95 structure, 191 total HP)
Test vehicle: CPLT-A1
6 missiles -- 61% -- 12 damage/missile
8 missiles -- 54% -- 11 damage/missile

JR7-D (128 armor, 125 structure, 253 total HP)
Test vehicle: CPLT-A1
6 missiles -- 59% -- 17 damage/missile

I fired at the grille above the head to avoid prematurely blowing the head off.

RVN-2X (224 armor, 125 structure, 349 HP)
Test vehicle: RVN-3L
12 missiles first fired at nose -- 98% -- 0.6 damage/missile
20 missiles then fired at "lower torso" -- 70% (28% change) -- 4.9 damage/missile

I first tried shooting at the tip of the raven's nose, but, seeing that this did almost no damage, I switched to shooting just below the the "beak" at the box to which the legs are connected.

RVN-3L (232 armor, 125 structure, 357 total HP)
Test vehicle CPLT-A1
18 missiles fired at "lower torso" -- 69% -- 6.1 damage/missile

This time, I fired all shots at the leg box, as above.

TBT-7M (240 armor, 175 structure, 415 total HP)
Test vehicle: RVN-3L
18 missiles -- 78% -- 5.1 damage/missile

CPLT-C1 (368 armor, 217 structure, 585 total HP)
Test vehicle: RVN-4X
6 missiles first fired at nose -- 98% -- 2 damage/missile
12 missiles then fired at point above cockpit, with the target's torso rotated downward -- 87% (11% change) -- 5.4 damage/missile

CTF-3D (352 armor, 223 structure, 575 total HP)
Test vehicle: CPLT-A1
34 missiles -- 76% -- 4.1 damage/missile

AS7-D (528 armor, 313 structure, 841 total HP)
Test vehicle: RVN-3L
86 missiles -- 84% -- 1.6 damage/missile

Evidently, I picked one of the worst possible aimpoints for this test. Missiles appeared to fly straight into the center torso, just below the square plates that look like reactive armor. In my earlier test (where I got 3.9 damage/missile), I was shooting between the top and bottom rows of square plates. Maybe this is a consistent pattern.

Credits: Much thanks to FactorlanP for allowing multiple withering barrages of missile fire to be directed at his person, and to the several other people who offered to help.

Edited by Amaris the Usurper, 15 March 2013 - 03:17 PM.


#2 FactorlanP

    Member

  • PipPipPipPipPipPipPipPip
  • 1,574 posts
  • LocationMy Wallet is closed until I see real progress.

Posted 15 March 2013 - 09:14 AM

Let me know if you want to re-try the Atlas test (or any others).

#3 PapajIGC

    Member

  • PipPipPip
  • Bridesmaid
  • Bridesmaid
  • 60 posts

Posted 15 March 2013 - 10:11 AM

Anybody who is anybody has known about the Raven's hitbox issues for awhile now; and by for awhile now I mean like 3+ months, particularly after they fixed the lag shield crap. When you're sitting 75m from a legged Raven and putting 2 Gauss dead center mass doesnt even strip the armor off anything on the front...you start wondering things. Things like how did the Raven get legged in the first place, or why is my Gauss convergence complete dung.

Edited by PapajIGC, 15 March 2013 - 10:12 AM.


#4 Phaesphoros

    Member

  • PipPipPipPipPipPipPip
  • 513 posts

Posted 15 March 2013 - 10:21 AM

(There needs to be a post in General Discussion to get not only the devs, but also more player's attention.)

First of all - thank you very much for doing the test and publishing the results.

Reference of hitboxes

Thoughts on your data:
  • I thought of a comparison with your SSRM data, but due to different scenarios (orbiting at close range vs. "center of area" <- standing still??) this is not very useful.
  • There's not a single case where the missile does approx. 2.5 dmg. There's 1.6 real dmg < 2.5 stat dmg < 4.1 real dmg.
  • Raven obviously does have an issue with hitboxes. Speculation: the missile explodes not on the raven hitbox, but before, doing only splash damage to the RVN. Could perhaps test that with ML.
  • CTPLT, RVN and AS7 get less than 2.5 dmg per missile, which indicates there's not only an issue of excess splash damage (maybe all of them have hit box issues?). Maybe the shape has something to do with it, like the pointed nose (consider STK and DRG).
  • Compare dmg/missile with hit boxes, especially consider splash dmg to side torsos. There might be a relation between the area hit by splash dmg and missile excess damage.

Edited by Phaesphoros, 15 March 2013 - 10:29 AM.


#5 Gen Kumon

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 319 posts

Posted 15 March 2013 - 01:10 PM

Well...this certainly confirms why Splatcats are so dangerous. Against all but a handful of mechs, they do an average of 180 damage per volley, not 90. No wonder my Treb just kind melts from a single shot.

I do have one request, though. Is it possible we can get a live LRM test? Paul said they weren't effected, but...the damage here seems to mesh fairly well with the numbers seen in earlier tests, which would indicate the damage in the Testing Grounds isn't as far off from real damage as he said it is, and considering LRMs were doing between 3 and 6 damage per missile in your first test, even a bugged testing ground that doubled damage wouldn't account for the 6 damage per.

If it really does turn out that even LRMs are bugged, there's a lot of "It's balanced, learn to play!" people that will either have to eat a lot of crow or just stop posting for awhile.

#6 Monky

    Member

  • PipPipPipPipPipPipPipPipPip
  • Shredder
  • Shredder
  • 3,219 posts
  • LocationHypothetical Warrior

Posted 15 March 2013 - 01:47 PM

It seems to be related to how many subsections within an actual armor section are hit by the explosion. It is ray-tracing to find and calculate how much volume is being hit from each individual armored section, and applying a fraction or increase damage depending on what it calculates.

Thus, a Raven hit on its beak only encompasses a portion of the mechs CT and little/none of the side torsos without line of sight to an of those parts, same for the CPLT - but when struck from the top or bottom rather than head on, much higher damage.

That is really bizarre.

Edited by Monky, 15 March 2013 - 01:48 PM.


#7 Karenai

    Member

  • PipPipPipPipPipPip
  • Big Brother
  • Big Brother
  • 340 posts
  • LocationGermany

Posted 15 March 2013 - 02:28 PM

Running a LRM boat 80% of the time I can guarantee you that Awesomes and Commandos DO take additional damage from LRMs. They pop like ripe fruit.
Also back section does sometimes get obliterated with LRM to the FACE. Have seen it several times.

#8 Amaris the Usurper

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 100 posts

Posted 15 March 2013 - 03:06 PM

The trouble with doing live LRM tests (and Streak tests, for that matter) is that we cannot target friendly units. Thus, we would need to sync-drop with multiple 8-man teams in order to have a good chance of getting a controlled environment with "enemy" units that we can target. Obviously, this can be arranged. I just haven't gotten around to it yet.

#9 Scop

    Member

  • PipPipPipPipPip
  • 110 posts

Posted 15 March 2013 - 03:13 PM

View PostAmaris the Usurper, on 15 March 2013 - 03:06 PM, said:

The trouble with doing live LRM tests (and Streak tests, for that matter) is that we cannot target friendly units. Thus, we would need to sync-drop with multiple 8-man teams in order to have a good chance of getting a controlled environment with "enemy" units that we can target. Obviously, this can be arranged. I just haven't gotten around to it yet.


I have a method for getting some info on LRM splash. http://mwomercs.com/...18#entry2062018 It's not as good as an 8-man simul-drop with your test subjects, obviously, but hopefully worth something.

#10 Sagamore

    Member

  • PipPipPipPipPipPipPip
  • The Blood Bound
  • The Blood Bound
  • 930 posts
  • LocationCanada

Posted 15 March 2013 - 03:20 PM

I find it hilarious that the Raven can shrug off almost all the damage if you hit it in the nose. Seeing as they are the ones inflicting lots of the splash damage, they take very little.

#11 Monky

    Member

  • PipPipPipPipPipPipPipPipPip
  • Shredder
  • Shredder
  • 3,219 posts
  • LocationHypothetical Warrior

Posted 15 March 2013 - 03:41 PM

View PostThontor, on 15 March 2013 - 03:30 PM, said:

couldnt you just dumb fire them?


LRMs yes, but I don't think they would benefit from tag/artemis, so they would have a wide spread. total damage done would still be calculable, but you wouldn't be able to detect specific spots that are good/bad to aim at. Not that this is a big issue for LRM's anyways as they mainly just go center of mass.

Edited by Monky, 15 March 2013 - 03:41 PM.


#12 shabowie

    Member

  • PipPipPipPipPipPipPip
  • 877 posts

Posted 15 March 2013 - 03:53 PM

Regarding some mechs taking lower splash damage than expected... Could it be that the damage drops off in a linear way the further you get away from the explosion within the splash radius? Then if there is a slight "dead area" for lack of a better term on a hitbox where a missile will detonate at a slightly further distance but not do damage right at the point of impact, like the very tip of the Raven's beak, it won't do the expected damage due to this?

Edited by shabowie, 15 March 2013 - 03:59 PM.


#13 Xelchon

    Member

  • PipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 118 posts
  • LocationJust Outside Sensor Range

Posted 15 March 2013 - 03:54 PM

View PostKarenai, on 15 March 2013 - 02:28 PM, said:

Running a LRM boat 80% of the time I can guarantee you that Awesomes and Commandos DO take additional damage from LRMs. They pop like ripe fruit.
Also back section does sometimes get obliterated with LRM to the FACE. Have seen it several times.


I true that about Awesome. I play both as the LRM user, and as a user of the AWS mech. I true that VERY MUCH. I think that is one of the main reasons Awesome are seen far less resilient than Stalkers while they should have been about the same. Theres also the massive CT hitbox (which i also think is a bit much) I know, but its simply more than that.

Edited by Xelchon, 15 March 2013 - 03:56 PM.


#14 Acid Phase

    Member

  • PipPipPipPipPipPipPip
  • Big Brother
  • 553 posts
  • Twitter: Link
  • LocationNew Jersey

Posted 15 March 2013 - 04:16 PM

Unbelievable. Completely and utterly game breaking revelation. Talk about priorities. This issue is number one. While PGI is milking money out of your pockets for cockpit items.....this is happening. Naked to the eyes of those gullible enough thinking they are king sh*t pilots.....please.

#15 Amaris the Usurper

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 100 posts

Posted 15 March 2013 - 04:39 PM

View Postshabowie, on 15 March 2013 - 03:53 PM, said:

Regarding some mechs taking lower splash damage than expected... Could it be that the damage drops off in a linear way the further you get away from the explosion within the splash radius?

I would not be at all surprised to find that the falloff is linear, but even if it is some other decreasing function, the principle is the same.

View Postshabowie, on 15 March 2013 - 03:53 PM, said:

Then if there is a slight "dead area" for lack of a better term on a hitbox where a missile will detonate at a slightly further distance but not do damage right at the point of impact, like the very tip of the Raven's beak, it won't do the expected damage due to this?

The system I have been imagining is that every hit location has a reference point (possibly center of volume) assigned for purposes of determining the splash damage. First you would determine the impact point, which is actually located on or near the surface of the 'mech, depending on how well the hitboxes correspond to the 3-D model. Then you would determine the distance from the impact point to the reference point. Finally, the damage dealt to each component would be computed according to

[damage] = 2.5*(1-[distance]/[splash radius]),

provided that [distance] < [splash radius].

This is pretty much what you were saying, but in mathematical terms. Obviously, it can't be the whole story. For example, how do we determine whether the front or the rear torso armor location takes the damage from a given impact? We might decide this by looking at what part of the hitbox for a given location a ray traced from the impact point to the reference point passes through. But that's beyond the point.

The point is that lots of reference points close to the impact point = lots of damage, and this happens more often on small 'mechs. Tellingly, if we consider the worst case scenario of full damage dealt to 8 locations (2 arm, 2 leg, 3 torso, 1 head), we get 20 damage per missile; the most damage per missile that we have seen is 19.

Is this the exact system used in MWO? Probably not. But it seems to match observations fairly well, and it is useful for illustrating what is going on.

Edited by Amaris the Usurper, 15 March 2013 - 04:40 PM.


#16 valkyrie

    Member

  • PipPipPipPipPipPipPip
  • 508 posts

Posted 15 March 2013 - 05:05 PM

You're doing God's work, Amaris.

Frankly, PGI needs to be paying you for this.

#17 Eddy Hawkins

    Member

  • PipPipPipPipPip
  • Knight Errant
  • Knight Errant
  • 154 posts

Posted 15 March 2013 - 05:59 PM

View PostPhaesphoros, on 15 March 2013 - 10:21 AM, said:

(There needs to be a post in General Discussion to get not only the devs, but also more player's attention.)

First of all - thank you very much for doing the test and publishing the results.

Reference of hitboxes

Thoughts on your data:
  • I thought of a comparison with your SSRM data, but due to different scenarios (orbiting at close range vs. "center of area" <- standing still??) this is not very useful.
  • There's not a single case where the missile does approx. 2.5 dmg. There's 1.6 real dmg < 2.5 stat dmg < 4.1 real dmg.
  • Raven obviously does have an issue with hitboxes. Speculation: the missile explodes not on the raven hitbox, but before, doing only splash damage to the RVN. Could perhaps test that with ML.
  • CTPLT, RVN and AS7 get less than 2.5 dmg per missile, which indicates there's not only an issue of excess splash damage (maybe all of them have hit box issues?). Maybe the shape has something to do with it, like the pointed nose (consider STK and DRG).
  • Compare dmg/missile with hit boxes, especially consider splash dmg to side torsos. There might be a relation between the area hit by splash dmg and missile excess damage.



i did have two posts in general, one got locked, the other deleted, with the reason being that there is already a thread covering the topic in another fourm. i think the moderators are just trigger happy to close threads.

Edited by Eddy Hawkins, 15 March 2013 - 06:06 PM.


#18 PoLaR

    Member

  • PipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 620 posts
  • LocationEast Bay

Posted 16 March 2013 - 05:33 AM

Would like to see more stats with the 8 man test groups.

Above info Is very interesting indeed.

I feel like we are the whistle blowers trying to expose the NWO lol Everyone who tries to expose the truth gets assassinated : P

#19 TexAce

    Member

  • PipPipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 2,861 posts
  • LocationGermany

Posted 18 March 2013 - 02:08 AM

See boyz this is the reason why we can't have nice things.

Look at the raven stats, 0.6 damage. What if this means it absorbs 70-80% or ALL damage done to it, this could explain why the damn chicks last longer than atlai.

AND WHY WE DONT HAVE RAVENS IN TRAINING GROUNDS!

Concerning the SRMs, I experienced this a lot of times. sometimes I can dump 12 salvos of 4x SRM6 into a stalker and not stip its armor of and sometimes I can core an atlas in 3 salvos. Totally messed up.

Edited by TexAss, 18 March 2013 - 02:09 AM.


#20 Red squirrel

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,626 posts

Posted 18 March 2013 - 01:03 PM

nice work





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users