24 replies to this topic

[Bagheera]

Before we begin, a couple of caveats:

1. This is completely unscientific, but perhaps we can get a few other test results in and see what is actually what. I do plan repeating the test a few more times on multiple targets at different ranges. I wanted to start the thread though, because I know most of you are better at this sort of empirical stuff than I am.

2. Please, let's avoid arguing the current state of dmg per missile. The goal of my test was based on a hunch: that an A-SRM4 was more efficient than an A-SRM6 in putting down effective dmg. I believe that due to flight tragectories I would see the same result at any dmg/missile number.

3. The splash mechanic comes into play here, but in ways I am not entirely clear upon. Anyone with more knowledge on that please feel free to chime in.

4. There was an expected result, and an unexpected result. I'll try to cover both.

5. As there seems to be a level of randomness in the flight pattern, many runs would be needed to really see what is happening here. I'm hoping a few others might be interested enough to compare various SRM racks in a similar fashion and post results here.

6. Again, this is not a dmg/missile thread, and it is not a thread to compare past and present SRM damage. This is a thread solely for the comparison of SRMs racks to other SRM racks and their relative efficiency in terms of TTK and APK (time to kill, ammo per kill).

The test:

Went into testing grounds with a TBT-5J. First run was with an A-SRM4, second run was with an A-SRM6. No other damage was applied to the test target. Test target was a stock Centurion on Frozen City. Luckily drew that map twice in a row. Range: 100m (common srm engagement range, though others should be tested) I targeted CT, slightly off center and high in each test. FWIW, I was slightly down-hill from the target with an un-obstructed LoF, and there might have been a slight off angle in facing.

A-SRM4 took 6 volleys to kill: 24 rounds.

A-SRM6 took 6 volleys to kill: 36 rounds.

This, unscientifically, implies that the A-SRM4 is a better choice than the A-SRM6: TTK is identical, and APK is better with the SRM4.

This suggests that there might be some oddness in the flight path that makes smaller SRM racks more effective than larger.

Discuss?

[DarkDevilDancer]

You know weapons dont work quite right in the trial maps quiaff?

[Lefty Lucy]

DarkDevilDancer, on 16 April 2013 - 09:34 AM, said:

You know weapons dont work quite right in the trial maps quiaff?

I'm pretty sure this isn't true.

[Viper69]

I often wondered if it would be better to put 3 SRM 4 with artemis in my Atlas than have 3 SRM 6s. Not having the funds to try it out doesnt help but I see standard SRM 4s seem to have a tighter group over their flight path than the 6s.

[StalaggtIKE]

Nice test. 4SRM have a tighter flight pattern than the 6SRM, so this isn't unexpected. 4SRM and 6SRM are comparable to a shotgun and a shotgun with a choke.

Viper69, on 16 April 2013 - 09:37 AM, said:

I often wondered if it would be better to put 3 SRM 4 with artemis in my Atlas than have 3 SRM 6s. Not having the funds to try it out doesnt help but I see standard SRM 4s seem to have a tighter group over their flight path than the 6s.

I highly recommend against artemis w/ 4SRM.

Edited by StalaggtIKE, 16 April 2013 - 09:41 AM.

[Kattspya]

Someone did a large thread about the different configurations on 6's, 4's and artemis on splatcats. He did it live with a decent number of runs and his conclusion was more or less that the difference between 4's and 6's was preference according to his live damage statistics. SRM6 wins out heavily on paper though and probably in practice. I am talking about pre nerf SRMs.

As to your tests the largest problem is that testing grounds code apparently has little to do with live code (thanks PGI for a belated testing and trial mode that doesn't work right for either purpose). But assuming the flight pattern matches up with live there are still problems. You are doing bench testing from a set distance. A distance large enough that 4's won out. Decrease the distance and you will find a breaking point where 6's are more efficient. I know I'd rather have more punch at knife fighting range than further away. But that's down to my style. Currently the only thing I have 4's in is my 2x because the arm only has eight tubes and I found that missed way to much with the second wave and would rather have more armor or heatsinks or whatever I crammed in instead.

EDIT: I looked for the thread but couldn't find it.

Edited by Kattspya, 16 April 2013 - 09:55 AM.

[Bagheera]

DarkDevilDancer, on 16 April 2013 - 09:34 AM, said:

You know weapons dont work quite right in the trial maps quiaff?

Citation please. I haven't seen any reports of this, and haven't noticed it in my own use of the Testing Grounds.

StalaggtIKE, on 16 April 2013 - 09:40 AM, said:

I highly recommend against artemis w/ 4SRM.

Ideally I'd like to do this with 2/4/6 Artemis and non-Artemis, but I doubt I have enough time between now and May to create a scientifically significant data set on my own. Impending tweaks to missiles in May could render this moot.

Also, I forget to include that target size is probably going to affect the results.

@Katt - multiple ranges are definitely needed for any weapons test to be viable, I may have forgotten to mention that in the OP. 100m was an arbitrary choice, results will certainly differ at 50m, 150m, 200m, and 250m-270m.

Edited by Bagheera, 16 April 2013 - 09:50 AM.

[Kattspya]

Bagheera, on 16 April 2013 - 09:48 AM, said:

Citation please. I haven't seen any reports of this, and haven't noticed it in my own use of the Testing Grounds.

I can confirm that I have seen several developer posts that claim that test and live differs.

Most notably in this thread: http://mwomercs.com/...ted-2013-03-15/

There is a dev post in there somewhere towards the middle but I'm not going to look for with broken search and a bad memory. Can't even remember which dev but it might've been garth.

[Sybreed]

LRMs deal a lot more damage on training ground then live... so I think those saying weapons act differently are right.

[LordBraxton]

6's are the only SRMs you want artemis on, 4s already have a nice tight grouping. Nobody brings 2s...

[Nihilo9]

I have always been under the assumption SRM6 was more preferable for the chance to hit on a moving target, less for the extra damage of two missiles.

As an aside - you're going to need far more than one test case to attempt to prove a point. I'd hesitate at 50.

[Jace Lancer]

Try it in a7m with a srm in the 1 tube missile slot. see if there is a spread difference. I swear it's more accurate out of there

[OneEyed Jack]

Everyone already knows that smaller launchers have tighter spread. So you'll naturally land around the same number in a given area, with the extra missiles being the outlyers. I'd wager that, while it took you the same number of volleys to kill the mech, you had put significantly more damage on other locations with the 6-packs than the 4-packs. SRMs are almost entirely point-blank weapons, though, so the information is mostly academic, since the larger launchers simply do more damage at the range they are typically used at, with spread being minimal.

[OneEyed Jack]

Jace Lancer, on 16 April 2013 - 10:07 AM, said:

Try it in a7m with a srm in the 1 tube missile slot. see if there is a spread difference. I swear it's more accurate out of there

It is. Single missiles fire in either an straight line or close enough to make little difference.

[Lefty Lucy]

Kattspya, on 16 April 2013 - 10:00 AM, said:

I can confirm that I have seen several developer posts that claim that test and live differs.

Most notably in this thread: http://mwomercs.com/...ted-2013-03-15/

There is a dev post in there somewhere towards the middle but I'm not going to look for with broken search and a bad memory. Can't even remember which dev but it might've been garth.

Yeah, they said that in that thread, but then it was later found that splash damage was incorrectly multiplying missile damage.

A dev also said in my thread about PPC min-range damage drop-off that testing grounds are not accurate. So I repeated a test in a live server and got *exactly the same results* as in testing grounds.

I'm more convinced that ever that testing grounds damage is spot-on.

Just remember, "dev" does not always mean "programmer." Also, even if a developer *is* someone who writes code, no single person is intimately familiar with the entirety of the code of the game. Game code is just way too complicated for that to be possible.

[Agent KI7KO]

Assuming this is all true, don't forget SRM-4s have .25s refire rate advantage. 4s versus 3.75s.

TTK of 20s versus 18.75s

Edited by Afoxi, 16 April 2013 - 10:18 AM.

[Kattspya]

Lefty Lucy, on 16 April 2013 - 10:14 AM, said:

Yeah, they said that in that thread, but then it was later found that splash damage was incorrectly multiplying missile damage.

A dev also said in my thread about PPC min-range damage drop-off that testing grounds are not accurate. So I repeated a test in a live server and got *exactly the same results* as in testing grounds.

I'm more convinced that ever that testing grounds damage is spot-on.

Just remember, "dev" does not always mean "programmer." Also, even if a developer *is* someone who writes code, no single person is intimately familiar with the entirety of the code of the game. Game code is just way too complicated for that to be possible.

I'm not even sure how you read my post or what parts of it you are responding to. I think I used an overabundance of qualifiers and I don't particularly need anymore but sure add them if you got 'em.

The first dev post in that thread said that yes, damage is high in test but also in live only less so. What do you remember them as saying? By the way the lower but still too high damage on live was confirmed by the community so in that case the devs are right. As for PPC dropoff I can't really take your word for it over the devs. But the devs are incompetents so you have a small edge. I'd need independent confirmation to believe you fully.

I wrote dev because I didn't remember which kind of dev. Not because I think a graphical artist can speak authoritatively on damage code. Since this is PGI we are talking about it is probably best not to take anyone's word on anything just to be safe. Incompetents... *mutter* *grumble etc etc

[Bagheera]

Afoxi, on 16 April 2013 - 10:18 AM, said:

Assuming this is all true, don't forget SRM-4s have .25s refire rate advantage. 4s versus 3.75s.

TTK of 20s versus 18.75s

Actually, I totally forgot about that. Thanks.

@General Responses - Yes, this is not even close to scientific. Pretty much the first thing I said in the thread.

I used the 5J as it seemed like a good way to test without tube count coming into play, for now. Against a stationary target the single tube launcher will put down missiles onto a single section. On a moving target, not so much.

@Katt - I'll swing through that thread later today. I didn't even know there was a discussion of live vs testing grounds results.

OneEyed Jack, on 16 April 2013 - 10:10 AM, said:

Everyone already knows that smaller launchers have tighter spread. So you'll naturally land around the same number in a given area, with the extra missiles being the outlyers. I'd wager that, while it took you the same number of volleys to kill the mech, you had put significantly more damage on other locations with the 6-packs than the 4-packs. SRMs are almost entirely point-blank weapons, though, so the information is mostly academic, since the larger launchers simply do more damage at the range they are typically used at, with spread being minimal.

Fair enough. I had a hunch though that the tighter spread of smaller racks might make them more effective than larger ones - admittedly there are a TON of variables for which my quick and dirty test doesn't account. Particularly tube count and play-style. I try to avoid getting closer 100m of anything when piloting a fast medium with srms. Sure, if I get closer I get a tighter pattern, but I also open myself up to more damage or getting caught up on another mech - something I like to avoid.

While the 6, in my scenario did likely put more ancillary damage on the target, I am more interested in effective damage. If I am focusing the CT, damage that goes to legs and arms is effectively wasted damage. At least academically. In practice that damage should get added to other ancillary damage to target from your team, or could be hitting sections already damaged, or give a greater chance of hitting a fast moving target moving laterally from the shooter's perspective.

Edited by Bagheera, 16 April 2013 - 10:55 AM.

[Lefty Lucy]

Kattspya, on 16 April 2013 - 10:25 AM, said:

I'm not even sure how you read my post or what parts of it you are responding to. I think I used an overabundance of qualifiers and I don't particularly need anymore but sure add them if you got 'em.

The first dev post in that thread said that yes, damage is high in test but also in live only less so. What do you remember them as saying? By the way the lower but still too high damage on live was confirmed by the community so in that case the devs are right. As for PPC dropoff I can't really take your word for it over the devs. But the devs are incompetents so you have a small edge. I'd need independent confirmation to believe you fully.

I wrote dev because I didn't remember which kind of dev. Not because I think a graphical artist can speak authoritatively on damage code. Since this is PGI we are talking about it is probably best not to take anyone's word on anything just to be safe. Incompetents... *mutter* *grumble etc etc

Another player actually graphed results and got the same results I did. Damage drop-off within 90 m is 10*(x/90)^2, where x is the range in meters.

[Wildstreak]

From using SRM6 and SRM4 in game, I know this.

I have used SRM6 on my AWS-8R, 4 of them. Cannot remember the ASRM6 pattern but the SRM6 pattern, from chain firing each launcher individually, they launch in 2 up and down rows of 3 with the left 3 drifting left and the right 3 drifting right for their spread.

Now SRM4 (I forget if I had Artemis when I saw this) on my HBK-4SP, they have an odd pattern. Picture 3 missiles in a triangle, the fourth is out a bit from one side of the triangle and the whole thing spins around the triangle's center.