127 replies to this topic

[Sarlic]

Make sure to read all the pages!
TLDR?:
http://mwomercs.com/...ost__p__4762897
http://mwomercs.com/...ost__p__4764127
http://mwomercs.com/...ost__p__4767242

---

Allright. Let's do this folks!

Piloting a Orion semi-brawler. Not forgetting to mention: Low arms! I understand that. But look at it first before you write a comment please.


My ping is 25 ms (shown in pictures)
Edit: Adders ping is 30 ms.

Or videolink here

A single SRM 4 mount does 8.60 pts of damage. DPS is 2.87. You do the math.




First salvo


Breakdown, skipped a frame:


Target analysis:

---

You can clearly see that i got my both arms pointed at the target where my SRMs are located.

• If i skip by frame i see my missiles dissappearing in a black hole(?) Look at the impact behind the Adder. Impossible that many missiles could have missed the Adder!
  
• The initial first salvo doesnt show me NOT hitting the mointain. However the impact does, but its miles away from the target! It occurs. 0.08 and 0.09 seconds. If i count about right, 6 missiles do not hit the target and goes around the Adder.

You judge the rest. Including the last SRM salvo.





I suspect i get scewed by the overly present of a enviromental hitbox of that hill and the hitreg the client(?). But the funny thing is my ping is around 25!

Last and but not least:



My brain hurts...Enlighten me!

Edited by Sarlic, 17 October 2015 - 09:52 AM.

[KodiakGW]

Did you happen to note the PING of the Adder? There has been a problem with hit reg between players with vastly different PINGs. At least there has been ever since it changed from 8v8 to 12v12.

"Vastly" being a diffence of over 100ms.

[Sarlic]

KodiakGW, on 14 October 2015 - 04:16 AM, said:

Did you happen to note the PING of the Adder? There has been a problem with hit reg between players with vastly different PINGs. At least there has been ever since it changed from 8v8 to 12v12.

"Vastly" being a diffence of over 100ms.

Ping of Adder was 30 according last end screen.

This is in my opinion enough evidence how terrible SRMs stinks in hitreg by either client/sever problems.

Edited by Sarlic, 14 October 2015 - 04:55 AM.

[Juodas Varnas]

#JustSrmHitregThings

[Torgun]

There are something called SRMs in this game? I must have forgot, since they've been crap like forever compared to lasers and ACs.

[Sarlic]

Here's more analysis from my part.



So went in testing ground and took a look on daytime on roughtly the same coords.

Upon closer look a fired my weapons at roughly the same position horizotnal and vertical target.





I have no idea what's going on. But there's certiainly something wonky going on since ever they tried to attempt to fix it.

Now i hope it's something worth looking into. But man something(?) is royalled f*cked and i try to be polite.

Edited by Sarlic, 14 October 2015 - 06:00 AM.

[Voq]

Great demonstration of the problem that discourages use of SRMs.
I hope they end up fixed, because some of my favourite builds use SRMS... But they're not fun if they can't be used reliably.

[Mazzyplz]

and paul inouye thinks he can ignore this and focus on balancing the variants. LOL!

missile variants will be forever garbage if you don't fix the missiles.

it would have been best if they focused on balancing weapons before they attempted all the fancy stuff they want to do in PTS.

it's like adding superchargers to a car with 4 flat tyres; and never bothering to change them

Edited by Mazzyplz, 14 October 2015 - 06:00 AM.

[sycocys]

For me they were working correctly as of 2 patches ago. I was extensively running srm4 on mediums and had very little issue racking up 600+ games with them.

Started leveling Blackjacks and tried to go back to my med's a couple weeks later and around 1 in 4 missiles actually register regardless of how you are firing them. -Alpha or single launcher-

Not really sure what changed, shouldn't have been anything but it really shredded my 3 ton lbx8's.

[Paigan]

I read somewhere that due to limitations of the engine, every frame of the game (or game logic "tick") can only process a certain amount of exploding (target affecting) missiles. The rest just falls short.

So if at the very same moment, same frame/tick, missiles from a few other players impact somewhere, it might have been that your missiles happened to get the short end of the stick.


I don't know if it actually works like that.
I'm a software engineer (not PGI) and I tried around a little with another game engine once (they really hooked in the game logic into the render loop which is kind of ... strange from an engineering point of view).
Plus hearing of all the limitations and weird behavior of the Cry Engine (which I personally can only explain by lazy/bad code design) makes it sound at least plausible.


So, hehe, for me, everytime I fiddle around with equipping SRMs for DAT brawling power, I think of the render loop, make faces and throw them out.

Edited by Paigan, 14 October 2015 - 06:07 AM.

[Sarlic]

Mazzyplz, on 14 October 2015 - 05:59 AM, said:

and paul inouye thinks he can ignore this and focus on balancing the variants. LOL!

missile variants will be forever garbage if you don't fix the missiles.

it would have been best if they focused on balancing weapons before they attempted all the fancy stuff they want to do in PTS.

it's like adding superchargers to a car with 4 flat tyres; and never bothering to change them

I think the main problem is the spread of these things. Combined with possible hitreg problems (and limitations of particles?!). As discussed earlier, the huge spread of those thing is rediculous.

You should get in testing grounds and find the SRM fire patterns. As far i can tell there are roughly three different fire patterns. Atleast i dont think it's that random as they claim to be. But i am not so sure about this as i havent seen a other test before.

Edited by Sarlic, 14 October 2015 - 06:07 AM.

[Mazzyplz]

Sarlic, on 14 October 2015 - 06:07 AM, said:

I think the main problem is the spread of these things. Combined with possible hitreg problems (and limitations of particles?!). As discussed earlier, the huge spread of those thing is rediculous.

You should get in testing grounds and find the SRM fire patterns. As far i can tell there are roughly three different fire patterns. Atleast i dont think it's that random as they claim to be. But i am not so sure about this as i havent seen a other test before.

ok we can test if this is the case by comparing LBX rounds; they have just as many particles right?

that theory also doesn't explain why the missiles worked better 3-4 months ago (i can tell you which patch exactly, it was the last variable geometry pass for inner sphere)

it was may or june

my bad it was july:
here is someone commenting on the patch visual changes - what nobody really picked up on... was that missile damage had dropped to only 60% out of 100 that patch.
http://mwomercs.com/...-geometry-pass/


i did notice but i assumed it was because my launchers had been spread out (which is a fair assumption since they had nothing on SRM written on the patch notes)

http://mwomercs.com/...__fromsearch__1

here's the thread i created to complain.

from this patch onward; the damage i did with SRM had a big chunk missing per match.
where as before emptying my ammo with some ppc shots on the enemy could net me something like 1000dmg (assuming i live to shoot all my ammo) now it only did 600. that's 40% missing, give or take (not factoring out extra erppc shots that count toward that sum - but still - it certainly doesn't explain this)

Edited by Mazzyplz, 14 October 2015 - 06:28 AM.

[sycocys]

Paigan - I tested it over a dozen matches and the results were all about the same - roughly 1 in 4 missiles counted for damage.

Was pretty frustrated by that point so I gave up on it and equipped streaks, bap and pulse lasers.

The one thing that theory doesn't explain is why they'd work properly 2 weeks ago and prior.

- and yes, LBX, LRM, lasers ticking should all have the same issue if its a problem with it dropping out damage counts per tick.

Edited by sycocys, 14 October 2015 - 06:13 AM.

[CDLord HHGD]

You got my vote Sarlic, I've noticed this too. Get a third person off to the side who can view the whole trajectory. Should be for some interesting new screenshots to further prove your point.

I get that some missiles will miss it being a spread weapon, but the wonky detonations and black hole thing is super weird.

Also test with Artemis too, see if there's a difference (and maybe backdoor way to entice people to use artemis too).

Edited by cdlord, 14 October 2015 - 06:15 AM.

[Sarlic]

Mazzyplz, on 14 October 2015 - 06:11 AM, said:

ok we can test if this is the case by comparing LBX rounds; they have just as many particles right?

that theory also doesn't explain why the missiles worked better 3-4 months ago (i can tell you which patch exactly, it was the last variable geometry pass for inner sphere)

it was may or june

I have no idea, but i did know when i first dropped on the fresh EU servers (It was after the HSR fix i believe, might be wrong!) everything felt much more fresh regarding hitting mechs.

But it's terrible frustating for sure. Each match i praise to the SRM god to drop with me out that same dropship.

[Paigan]

sycocys, on 14 October 2015 - 06:11 AM, said:

Paigan - I tested it over a dozen matches and the results were all about the same - roughly 1 in 4 missiles counted for damage.

Was pretty frustrated by that point so I gave up on it and equipped streaks, bap and pulse lasers.

The one thing that theory doesn't explain is why they'd work properly 2 weeks ago and prior.

- and yes, LBX, LRM, lasers ticking should all have the same issue if its a problem with it dropping out damage counts per tick.

Hm, interesting. Thank you.
As I said, I wasn't sure if what I wrote was accurate. I just read it somewhere and it sounds technically plausible.
Maybe it's that plus another problem/bug.
Or it's complete rubbish .

What I hear, one thing remains constant: the CryEngine and/or PGI's way of using it has a LOT of problems.




edit:
one thing, though:

sycocys, on 14 October 2015 - 06:11 AM, said:

[...]
- and yes, LBX, LRM, lasers ticking should all have the same issue if its a problem with it dropping out damage counts per tick.

Not necessarily.
People not familiar with a certain code (game or otherwise, developer or not) assume that every logic is written nice and clean, with everything similar handled by the same piece of code in a general way.
90% of the time, this is not the case.
Developers improvise, hack, rush, make mistakes, etc. and in the end every lousy little detail very often gets its own copy&pasted special casing to just "get the job done" (= maintenance nightmare. I just say "We don't touch ammo code any more because the guy who wrote it quit a long time ago")

Meaning: It could very well be that missiles have their own unique logic or even abused engine side effect to function. Maybe even SRMs have another one than LRMs. Maybe even IS SRMs have another one than C SRMs.
The possibilities for dirty hacking are endless and they all exist somewhere in some program.

Edited by Paigan, 14 October 2015 - 06:26 AM.

[Mazzyplz]

Paigan, on 14 October 2015 - 06:17 AM, said:

Not necessarily.
People not familiar with a certain code (game or otherwise, developer or not) assume that every logic is written nice and clean, with everything similar handled by the same piece of code in a general way.
90% of the time, this is not the case.
Developers improvise, hack, rush, make mistakes, etc. and in the end every lousy little details very often gets its own copy&pasted special casing (= maintenance nightmare. I just say "We don't touch ammo code any more because the guy who wrote it quit a long time ago")

Meaning: It could very well be that missiles have their own unique logic or even abused engine side effect to function. Maybe even SRMs have another one than LRMs. Maybe even IS SRMs have another one than C SRMs.
The possibilities for dirty hacking are endless and they all exist somewhere in some program.

if this is the case then throw out the SRM code and swap it with lbx code and tweak the graphics/dmg.

it was all the splash tweaks they did since open beta that keep making this weapon a laughing stock - instead of giving it code similar to lbx - they did a filthy hackjob by making the radius on the splash for each missile infinitesimally small, it's like one of those string theory extra dimensions -

could this sloppy programming trick explain many of the SRM woes? I SAY YES

Edited by Mazzyplz, 14 October 2015 - 06:30 AM.

[Paigan]

Mazzyplz, on 14 October 2015 - 06:25 AM, said:

if this is the case then throw out the SRM code and swap it with lbx code and tweak the graphics/dmg.

it was all the splash tweaks they did since open beta that keep making this weapon a laughing stock

What if SRMs have some slight difference to LBX that can't be done with the LBX logic? E.g. attaching a graphical effect for the missile.
Players would complain endlessly about how crappy the missiles look.

E.g. in Mass Effect 3, weapons firing ballistic missiles (e.g. grenade launcher) don't have a shooting effect at all but every shot triggers the creation of the grenade out of thin air (hack) or something like that. Because whenever it's laggy, the grenade launcher sometimes fails to fire its grenade (lag eats the hack).

As far as I know, PGI doesn't have the possibility or the knowledge or the man power (or maybe all at once) to dig deep enough into the engine to code the proper logic they actually need.

Edited by Paigan, 14 October 2015 - 06:31 AM.

[Mazzyplz]

Paigan, on 14 October 2015 - 06:30 AM, said:

What if SRMs have some slight difference to LBX that can't be done with the LBX logic? E.g. attaching a graphical effect for the missile.
Players would complain endlessly about how crappy the missiles look.

E.g. in Mass Effect 3, weapons firing ballistic missiles (e.g. grenade launcher) don't have a shooting effect at all but every shot triggers the creation of the grenade out of thin air (hack) or something like that. Because whenever it's laggy, the grenade launcher sometimes fails to fire its grenade (lag eats the hack).

As far as I know, PGI doesn't have the possibibilty or the knowledge or the man power (or maybe all at once) to dig deep enough into the engine to code the proper logic they actually need.

like - who cares if SRMS lose their smoke trail - so long as they work

[Paigan]

Mazzyplz, on 14 October 2015 - 06:31 AM, said:

like - who cares if SRMS lose their smoke trail - so long as they work

I'd tend to agree.
... until hell breaks loose in the forums about the broken missile graphics .

(also, it was just a simple, plastic example. Could also be a much more abstract, logic-level thing)

Edited by Paigan, 14 October 2015 - 06:34 AM.