Jump to content

Explosion Velocity Of Lrm Is Stil...wrong


56 replies to this topic

#1 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 24 October 2013 - 10:05 AM

"As I sat perched over the water on the Crimson Strait, I could not help but be in awe of the sunset before me. The rays danced across the water in a symphony of color, glimmering each as if they were suns themselves. Despite all of our conflicts and strife, the sun would rise again where we would not.
It was a moment of clarity - a rare commodity these days - broken only by the sound of klaxons lamenting on the approach of missiles.

Peace would have to wait for another sunset."

-Oni


----
====

EDIT:

This has been confirmed by Dev as intentional. From Tom's response:

"At the moment of impact for each missile does a radius check. If the radius is within the component, damage is done. Damage is split between all touching components, % depending how much radius is within component. Components overlap.

Having said that, the explosion radius of missiles is like 10cm now and no more is needed. We made it this small because so much QQ about small mech taking damage everywhere compared to large mechs. So now the illusion is that only at the warheads point of impact damage is done. There is no explosion velocity, this isn't EVEOnline, instantaneous damage within the radius and all damage done on server.

As for damage each missile will deal a maximum of 2 damage no more, no less. Airstrike and Artillery use large radii with the same kind of checks normal missiles do. "


So there you have it. The explosion radius (not the velocity as I had hoped) is throttled down. This produces a hot impact zone with very low overall splash (as suspected). As such, lights are able to skirt out of "bubble" and take little to no effect. In the end, the result is the same - slow burn, large radius vs. fast (instant) burn, small radius. I went down the explosion velocity theorycrafting route since the results are inconstistant in game - ala movement direction, hit registration, etc.

Now, this brings up even more questions on hit detection... and a huge sadface that lights get a kitchen pass on LRM :)


====

<SNIP>

Edited by Oni Ralas, 25 October 2013 - 09:21 AM.


#2 Cycleboy

    Member

  • PipPipPipPipPip
  • Bad Company
  • 183 posts
  • LocationWisconsin

Posted 24 October 2013 - 10:49 AM

Paraphrasing... so you think that the missile impact and explosion that deals damage are figured separately by the server? And that by the time the 2-3ms goes by for the calculation the 140+kph boogers are outside the damage detection zone? Hmm... hard to test by us externally. PGI would have to confirm this is not the case within house. Easy test though. Have Spider at 400m run straight away as you fire LRMs at it. If back not shredded, then problem.

#3 AC

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,161 posts

Posted 24 October 2013 - 11:21 AM

View PostCycleboy, on 24 October 2013 - 10:49 AM, said:

Paraphrasing... so you think that the missile impact and explosion that deals damage are figured separately by the server? And that by the time the 2-3ms goes by for the calculation the 140+kph boogers are outside the damage detection zone? Hmm... hard to test by us externally. PGI would have to confirm this is not the case within house. Easy test though. Have Spider at 400m run straight away as you fire LRMs at it. If back not shredded, then problem.



Actually no.... What I believe he is talking about is that missiles travel 1000m total. Not 1000m from the mech that fired it. So a light mechs best bet is to run sideways to the firing mech and force the missiles to arc around and follow them. By time the missiles make a big turn and get back to following the light mech, they often time have travelled their 1000m and self detonate.

The lag portion is that missiles try and slam into mechs from up in the air traveling in a downward arc. I have personally watched missiles (hundreds of them), slam into the ground just behind spiders (and Jenners to an extent). I assume this is due to lag issues. Let's use some real world math... don't freak I am an engineer and am licensed to wield such power.

So a Spider is flying along at 150kph. Missiles are chasing it and are arcing down on it, but slam into the ground. What happened? Let's assume a modest ping of 100ms. At 150kph, the spider is going roughly 42 meters/second. That means the spider is 4.2 meters ahead of where the missiles think it is. This is significant! (And quite honestly looks to be about what I witness in game.)

#4 Cycleboy

    Member

  • PipPipPipPipPip
  • Bad Company
  • 183 posts
  • LocationWisconsin

Posted 24 October 2013 - 11:27 AM

A/C, I can see what you are saying, and that IS what happens when lights dodge LRM fire. I'm going off the statment

Quote

[color=#959595]They do however have to make an additional damage mechanic check - area of effect based on an explosion radius. In order to calculate the affected area of an explosion, you need to know a few things -- how big the explosion is (radius) and how fast it expands (velocity).[/color]


So I'm thinking the OP wants to know if the HSR or whatever detection mechanism freezes time from the point of "LRM touches solid object" to figure out what is in that blast zone, or does it all happen real time, in which a light could theoretically run faster than the blast goes off.

Personally, I think the whole "area of effect from missiles" is dumb. Should be shape charges. Touch=damage. No AoE. No shrapnel. None of that. LRM touches arm, arm damaged. Next touches RT, RT damaged.

#5 jtyotJOTJIPAEFVJ

    Member

  • PipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 206 posts

Posted 24 October 2013 - 11:42 AM

It would seem to me that OP has no idea about what he's talking about. AC/20 damage calculation is as follows: damage = 20. There's no formula. And neither are missile explosions computed as expanding spheres. They expand instantly to their maximum radius and apply damage to whatever is inside the sphere. There is no explosion velocity.

#6 Captain Stiffy

    Member

  • PipPipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 2,234 posts

Posted 24 October 2013 - 11:59 AM

Try an A1 with 6 LRM5's and chain them at a light running around. It's clown-shoes farking silly as hell how the missiles fail to track.

I want my twisty LRM's back. They looked so cool.

#7 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 24 October 2013 - 12:06 PM

View PostPiipu, on 24 October 2013 - 11:42 AM, said:

It would seem to me that OP has no idea about what he's talking about. AC/20 damage calculation is as follows: damage = 20. There's no formula. And neither are missile explosions computed as expanding spheres. They expand instantly to their maximum radius and apply damage to whatever is inside the sphere. There is no explosion velocity.


Then you would be wrong. Let me elaborate on your failure...

An AC20's (all ballistics) calculation does apply damage based on distance, not velocity. This is why, Piipu, when you fire at max range you hit for a % of what you would under your effective range. I use velocity as a comparative, not as a derivative.

Missiles, I hate to tell you, do in fact have an explosion velocity and radius calculation (expansion rate = explosion velocity there kiddo..which you just said is what is applied). If they didn't, then any light mech would be struck with the same concussive force if they were in motion vs. standing still. If you are suggesting a light mech in motion takes the same amount as one stationary, then you have never played MWO.


Think of it like this:

At any given moment, I have some constants - the size of a mech, the speed of a missile, and the amount (rads) an explosion will encompass. Ideally, I also know how long an explosion will last (there is a decay rate of explosive, but I doubt it's being used to apply negative modifiers) but we'll assume it's also a fixed point.

Now imagine if you will I have an explosive charge that has a very fast expansion time, but a relatively small diameter of effect. If I was to strike say...an atlas, than the physical size of the atlas ensures all of the radius of explosion is covered, but it also means the damaged areas will be confined to a much smaller area (remember: Explosions happen from a central point, release most of their energy from 0 to 50% of their lifespan, then begin to rapidly fade after that to being null at 100% lifespan). The total damage taken, raw number wise, we'll say is 10. Distribution wise, the impact site took the most damage, with outlying components getting a relatively small amount (by scale).

Now that same missile fired at a spider will have a different effect. When stationary (again, we'll assume LoS to keep it simple), that same missile hits a smaller surface area. Now, since the surface area is smaller, you'd think the total damage would be lower - when in fact it's higher. Why? Because the explosion velocities are fast, radius is slow. They're confined charges. The spider, with no other modifiers, takes that hit and the damage is distributed across an area. Problem is, the area is significantly smaller than say the Atlas. A sphere of effect that would have hit the ct/rt/arm would envelop pretty much all of the spider. Ah Ha! but Oni, that would mean there are *more* parts to hit, and thus the damage to each *component* is lower right!? Well... kinda. That would be true if the calculation was raw damage / number of components, but we all know it isn't. There is a curve in their calculations of splash (modified heavily since the SRM bug was first discovered). As such, even with a larger number of total components, the splash element doesn't really change the dynamic much. What does hurt the spider is the relatively small explosion radius. Remember how I said explosions aren't linear? That is, they do an energy dump up front, and dissipate quickly as they expand? Well our poor spider has the bulk of it's components covered in that high damage range of 0-50% (rate of expansion is fixed).

A smaller mech should take more punishment from a missile than a larger one...but they don't. Why? Transversal velocity I'm guessing. The idea here is that a moving body will not remain in a fixed point in space, and anything that happens during a slice of time must take into effect our movement through said space. Let's go back to the above:

Now my atlas is moving full speed at a whopping 52kph. Even though it is moving, it is not moving fast enough where its total components exit the impact radius in a timely fashion (ie: enough to not get as hurt by staying in the fire). We see no real difference in damage.

Our spider, by contrast, is moving at 140kph. The explosion, however, is not (at the point of origin). So, as my sphere beings to expand, the spider is moving - ever so rapidly - away from that point in space. Even with a relatively fast rate of expansion, the total affected area remains constant. The spider is able to move itself out of the main impact area and takes little (to no, depending on HSR) damage.

So there is always a trade off between explosion velocity and radius. Big missiles make big booms, but they also are slow to deliver all that energy outward. Small missiles make small booms, but are very efficient at delivering it quickly. You can't really have a large radius with a super fast expansion velocity, else it pretty much throws off the mechanics of other systems.

----

So, with all that being said - there is a problem with how the explosions are being calculated. I highly suspect that forward momentum (which we did NOT take into account above to illustrate) of the missile is not added. That is, we're simply flipping a bit to say we've exploded and not taking into account that the missile was traveling forward at the time. That is why a smaller mech is able to outrun the actual missile -- it's not projecting the damage along the same trajectory, it's almost like popping in mid air based on the collision detection.

Edited by Oni Ralas, 24 October 2013 - 12:37 PM.


#8 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 24 October 2013 - 12:36 PM

View PostCaptain Stiffy, on 24 October 2013 - 11:59 AM, said:

Try an A1 with 6 LRM5's and chain them at a light running around. It's clown-shoes farking silly as hell how the missiles fail to track.

I want my twisty LRM's back. They looked so cool.


That is actually what lead me to write the post. I run the 6LRM5 cat for lulz all the time but have all but stopped targeting lights. They cannot be hit.

#9 Cycleboy

    Member

  • PipPipPipPipPip
  • Bad Company
  • 183 posts
  • LocationWisconsin

Posted 24 October 2013 - 12:42 PM

View PostOni Ralas, on 24 October 2013 - 12:06 PM, said:

So, with all that being said - there is a problem with how the explosions are being calculated. I highly suspect that forward momentum (which we did NOT take into account above to illustrate) of the missile is not added. That is, we're simply flipping a bit to say we've exploded and not taking into account that the missile was traveling forward at the time. That is why a smaller mech is able to outrun the actual missile -- it's not projecting the damage along the same trajectory, it's almost like popping in mid air based on the collision detection.


Hellllooooo, Mr. Fancypants. ;) So... was my paraphrase accurate? You believe the explosion sphere used to calculate damage caused is being applied after collision detection of the missile, but that lights move so fast they exit that sphere before the server can register the damage? That even the low number of missiles that do hit a fast moving light do even less damage because of this phenomenon?

#10 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 24 October 2013 - 12:52 PM

View PostCycleboy, on 24 October 2013 - 12:42 PM, said:



Hellllooooo, Mr. Fancypants. :P So... was my paraphrase accurate? You believe the explosion sphere used to calculate damage caused is being applied after collision detection of the missile, but that lights move so fast they exit that sphere before the server can register the damage? That even the low number of missiles that do hit a fast moving light do even less damage because of this phenomenon?


Aye, pretty much ;) . I think it all boils down to a combination of two factors:

1) Missile speed is not added to the explosion numbers (ie: the missile doesn't just "stop" when it hits something, it (and the explosion) continue moving forward)

and

2) The change to the splash mechanics narrowed the explosion radius down quite a bit without changing the explosion velocity. Too narrow while also being too slow.

#11 Cycleboy

    Member

  • PipPipPipPipPip
  • Bad Company
  • 183 posts
  • LocationWisconsin

Posted 24 October 2013 - 01:03 PM

Gotcha. I still go back to my previous statement that "splash = stupid". Just have a location get hit and apply damage. Done. Next shooter please. Why over-complicate it.

#12 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 24 October 2013 - 01:05 PM

View PostCycleboy, on 24 October 2013 - 01:03 PM, said:

Gotcha. I still go back to my previous statement that &quot;splash = stupid&quot;. Just have a location get hit and apply damage. Done. Next shooter please. Why over-complicate it.


We used to kinda have that back in the day actually. I personally dislike the current implementation of splash - not that splash itself is wrong, but they way they are doing it is.

Artemis changes the game a bit, makes it seek CT on LoS. Funny thing though, it still doesn't hit the CT while in motion :|

#13 wintersborn

    Member

  • PipPipPipPipPipPip
  • 412 posts

Posted 24 October 2013 - 01:41 PM

Simply double the speed of missiles.

I can out run most all LRM's and if I get hit is will only be one maybe 2 missiles out of that 15-20.

Most LRM users don't even bother with lights since it does not work vs their run speed.

#14 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 24 October 2013 - 01:47 PM

View Postwintersborn, on 24 October 2013 - 01:41 PM, said:

Simply double the speed of missiles.

I can out run most all LRM's and if I get hit is will only be one maybe 2 missiles out of that 15-20.

Most LRM users don't even bother with lights since it does not work vs their run speed.


Speed is good, but if speed is negated by the explosion issue - then it won't make any difference.

I do think a speed increase would allow for a better positioning dynamic. There is soooo much bloody cover these days... coupled with the (what I feel) is a hit issue. Average engagement range for a good LRM pilot is 300-400m. A slightly faster missile speed would allow a slight increase of engagement range without just hitting the dirt all the time.

#15 jtyotJOTJIPAEFVJ

    Member

  • PipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 206 posts

Posted 25 October 2013 - 01:55 AM

View PostOni Ralas, on 24 October 2013 - 12:06 PM, said:

Useless wall of text

You're confusing real life physics with game physics. Why LRMs don't damage moving lights is due to hit detection issues, it has nothing to do with what you're talking about... kiddo. There's no point in making explosives behave in such a manner in the game simulation when you can achieve a similiar-looking effect with instant area of effect damage. That's why pretty much any game uses that method. It looks almost the same and is a lot lighter to compute. You might know something about real physics but you clearly have no idea how games work.

Edited by Piipu, 25 October 2013 - 02:01 AM.


#16 Oni Ralas

    Member

  • PipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 762 posts

Posted 25 October 2013 - 04:28 AM

View PostPiipu, on 25 October 2013 - 01:55 AM, said:


You're confusing real life physics with game physics. Why LRMs don't damage moving lights is due to hit detection issues, it has nothing to do with what you're talking about... kiddo. There's no point in making explosives behave in such a manner in the game simulation when you can achieve a similiar-looking effect with instant area of effect damage. That's why pretty much any game uses that method. It looks almost the same and is a lot lighter to compute. You might know something about real physics but you clearly have no idea how games work.


*cough* EVE *cough*

Ahem, sorry about that...something in my *cough* HAWKEN *cough* throat.

Weird.

Now where were we? Ah yes, the explosion calculations for CRY3....

#17 OneEyed Jack

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,500 posts

Posted 25 October 2013 - 05:50 AM

LRMs DO damage Lights, but most of them miss.

Against a moving Light, when they dive, they don't turn to follow much or at all right at the end, so the mech moves out of the way and only a few will clip it. Even against a stationary Light, smaller body size means larger missile volleys are going to cover more area than the mech does. But they certainly do hit.

I kill/leg Lights on a pretty regular basis when running LRMs, particularly when I can chain small clusters. It may not be great for just launching at a Light that's randomly running along in the open at 800 meters, but it's huge when they're engaged with another friendly, particularly another Light that they can't easily outrun.

#18 Karl Streiger

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • Wrath
  • Wrath
  • 20,369 posts
  • LocationBlack Dot in a Sea of Blue

Posted 25 October 2013 - 05:57 AM

View PostAC, on 24 October 2013 - 11:21 AM, said:

The lag portion is that missiles try and slam into mechs from up in the air traveling in a downward arc. I have personally watched missiles (hundreds of them), slam into the ground just behind spiders (and Jenners to an extent). I assume this is due to lag issues. Let's use some real world math... don't freak I am an engineer and am licensed to wield such power.

No not lag issue allone:

You can try some test runs on testing grounds:
Had a Highlander and a single LRM 20...target: Awesome: front - dead after 6th or 8th volley - can't remember.
After that i attacked a Catapract - from the flank - target standing still - 10 or 12 volley and hardly any damage- most volley simple hit the ground behind the target - other try - i attacked from the front and it was soon over

#19 Satan n stuff

    Member

  • PipPipPipPipPipPipPipPipPip
  • Urban Commando
  • Urban Commando
  • 3,508 posts
  • LocationLooking right at you, lining up my shot.

Posted 25 October 2013 - 06:55 AM

View PostOni Ralas, on 24 October 2013 - 12:06 PM, said:

Then you would be wrong. Let me elaborate on your failure...

Or... a light in motion takes less damage because fewer missiles actually hit it? Please don't invent mechanics and claim that's how the game works.

#20 Krivvan

    Member

  • PipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 4,318 posts
  • LocationUSA/Canada

Posted 25 October 2013 - 07:04 AM

View PostAC, on 24 October 2013 - 11:21 AM, said:

So a light mechs best bet is to run sideways to the firing mech and force the missiles to arc around and follow them. By time the missiles make a big turn and get back to following the light mech, they often time have travelled their 1000m and self detonate.


I can confirm that one way us light pilots deal with LRMs in the open is to run perpendicular to where they were initially coming from. Maybe followed by a hard turn or jump jet turn when they're just about to hit.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users