12 replies to this topic

[Spades Kincaid]

In light of the recent changes to LRM's, and the back and forth debate (to characterize it nicely) over whether they are OP, UP, still splashing too much, coring in 1-2 salvos, etc. I decided to do a series of tests in the Training Grounds.

Before anyone jumps on that, I'm well aware of the limitations of what that means in relation to actual gameplay. Doesn't mean it can't be useful in some ways.

The setup:

I took my 2xLRM15 C1 Catapult and ran through the 8 target mechs. I did a 'short' range test (250-300m) and a 'long' range test (550-600m). For my purposes 550-600 was long enough. At that range LRM's achieve a higher arcing flightpath than they do under 300m. The path at even longer ranges doesn't change noticeably from it.

I did it with Artemis, with Artemis and TAG, and then stripped the Artemis from my mech and did it with unmodified LRM's. Then I decided to do a single runthrough at short range (250-300m) from the side, just for comparison's sake. Values are salvo's fired. 30 missiles each.

Oh, and stock armor for the variants in the Training Grounds, since I doubt most people know what they all are off the top of their head. Because some are quite short of fully armored and some are near max. Which is important to some of the noteworthy results.

COM-1B : 128 (178)
JEN-D : 128 (238)
CIC-2A : 128 (274)
CEN-A : 272 (338)
CAT-A1 : 384 (422)
PHR-1X : 352 (434)
AWE-8Q : 480 (494)
ATL-D : 608 (614)

The results:



I found it quite interesting and I think there's some useful things to take from the numbers.


1) Don't face-eat LRM's. (Duh)

Included for those who've spouted that twisting doesn't matter/help. Hogwash. It may well depend how near directly sideways to them you get... but it can help quite a lot in most cases. Noted that side-on is not exactly the same as twisting. Since the 'hips' are part of the CT and those would not be turned by just twisting.

I don't think the difference would be that major, since most of the missiles impact above that anyway. Might vary with chassis, but certainly twisting -can- make a significant difference. If you can get your torso close enough to 90 degrees to the flightpath. How close? Who knows. Probably depends on the mech. Someone else can run all the firing tests required for that if they want =P


2) Still some coring oddities appearing.

Why is an Atlas D (4 pts short of max armor stock) getting cored in 3 salvos with Art+TAG? That's only 90ish damage if all missles hit CT. I don't recall what the stock ratio is, but it should have 90-ish front CT armor. So, I jumped back in with my Cataphract X and put 2 Gauss + 2 MLaser on it.

Took 5 shots of 2x Gauss and a finishing 2 MLaser burst to core it out. 150+ dmg. So it should be taking about 5 2xLRM15 salvos even under optimal Artemis+TAG. It's taking 3.

Similarly for the stock Cat A1. At it's stock 384 armor, (More than the stock Cataphract that averaged about an extra salvo to kill) I can't see how it should be possible to core it out with 60 LRM's. Took me the equivalent of 3 2xGauss shots. (Took 2 dual Gauss and 3 bursts of 2xMLaser) to core the A1 Cat. = 90 dmg

So if splash is as small as it supposedly is now then something else strange is going on that has nothing to do with splash. Or it's broken in some way still that they aren't aware of. I thought they were going to reduce splash -damage- directly. But from what I've read it's the radius they drastically changed. Well apparently that's still not exactly working out well if the LRM's strike you head-on.


3) Shape Matters.

The Jenner is naturally missile avoidant head-on. Look at the numbers above. It has the most irregular behavior and trends to taking more salvos to bring down. And that's with less than full armor. Almost half-armor. Same 128 as the stock Commando had. Less than half what the stock Cent had. Wow.

Why? Because of it's shape it would seem. It's torso is relatively short vertically, but elongated and narrow. Like a forward neck. What seems to be happening is the missiles are aiming for it's neck, or the head, and many either passing left/right of it and hitting the legs, or missing entirely.

Conversely, it took fewer salvo's to kill it from the side, than it did for the slender profile humanoid Commando that gets cored in one salvo head-on.

The Cicada shares this to a more minimal extent. The neck being far shorter. It still took more to kill it than the Commando, even with the same amount of armor.

The Cataphract has the most potential for twisting to protect itself. Due to it's prominent shoulder ridges and maybe some other oddity in flightpath alignment with mechs. For whatever reason, probably something in my positioning, the trial with Artemis alone took -27- salvos to kill it from the side.



---------

So the Jenner gives LRM's odd fits. The Cataphract can be exceptional at twist protection. The Catapult and Atlas love LRM's so much they gather them to their bosoms in joyous death.

LRM's as a whole, still seem to have some kind of damage issue. The Cat and Atlas at least, still show something strange going on with coring out more quickly than they should be. And frontal coring in general still seems a bit too quick in almost all cases. While side impacts -can- make them look UP, as on the Cataphract. Even ignoring the 27 result. 16 x 30 at .9 is 432 direct LRM damage (no splash at all). That's 80 pts above the -total- armor value of the mech. Trying to kill it from -one- side.

OP -and- UP.

No wonder player views on them can be so varied, even personal prejudices for/against aside.


If you got through all of that, congrats. =P

Just thought I'd share the results I got. Comments welcome. But if you're going to criticize the sample-size, etc; don't bother. I'm well aware. It's not a scientific assessment.

I spent hours just to get what I did. And a lot more runs may have narrowed out a couple outliers more clearly (like the 27 salvo Cataphract kill), but not much beyond that I think. What trends there are seem fairly consistant.

Edited by Spades Kincaid, 25 May 2013 - 07:07 PM.

[Nightcrept]

I applaud your efforts.

Problem is that the devs have said the training grounds are completely borked.
But your findings are fascinating,

[Victor Morson]

Nightcrept, on 25 May 2013 - 07:18 PM, said:

I applaud your efforts.

Problem is that the devs have said the training grounds are completely borked.
But your findings are fascinating,

Pretty much what he said.

Training Grounds is totally not representing actual in game behavior right now.

[Renthrak]

Victor Morson, on 25 May 2013 - 07:34 PM, said:

Training Grounds is totally not representing actual in game behavior right now.

That is a problem. Since we can't choose who we end up with in a game to any reliable degree, people getting together to test game mechanics rather than just fight as normal is nearly impossible. You could shoot your own team mates, but since you can't lock LRMs on friendlies, there really isn't much that can be done to get good test data.

The best that we can hope for is that behavior in the Testing Grounds will at least be consistent inside the Testing Grounds. That is, even if it takes 50 missiles to kill something in the TG and 35 in-game or something, as long as the missiles inflict damage the same way against every 'Mech in the TG, we can use the proportional differences for reference.

Let's say damage is doubled in the TG compared to the regular game. We can still determine that it takes twice as many shots to kill something from the side compared to straight-on. We should also be able to show that it takes 50% more hits to kill one 'Mech compared to another, or that splash damage kills things faster than direct damage, for example.

Of course, if the Training Ground is badly bugged, all bets are off. Then again, since we lack other means of collecting this sort of data ourselves, we'll just have to make do.

I've recently started a thread on changing missile targeting to distribute damage without splash. For those who are interested, details are here:

http://mwomercs.com/...eeking-to-lrms/

[Praehotec8]

Thanks for the thorough work, although it's concerning about the testing grounds as others have stated. I think they ought to just remove splash damage right now (if splash is warranted) do something like randomly assigning one missile to hit 2-3 components and do full damage to one and x-amount (splash) to the other 1-2 components (as opposed to making splash based on a radius from impact site).

[Sephlock]

Victor Morson, on 25 May 2013 - 07:34 PM, said:

Pretty much what he said.

Training Grounds is totally not representing actual in game behavior right now.

Then how come training grounds QQing got LRMs unnecessarily nerfed in the first place?

[Deathlike]

There's nothing wrong with LRMs exactly.. but if you are not a mobile target and/or are being spotted with TAG, LRMs will ruin your day very quickly.

Movement in combination with cover reduces your damage intake significantly.

[Nicholas Carlyle]

Sephlock, on 25 May 2013 - 10:31 PM, said:

Then how come training grounds QQing got LRMs unnecessarily nerfed in the first place?

If you'll notice, it was when people noticed splash, and PGI started messing with it, is when LRM's really went into the *******.

Same with SRM's.

Training Grounds have really screwed up the game.

[PeekaBoo I C Ju]

LRM's are useless again, my LRM mechs are parked once again, IMO 2 LRM 20's should command Similar fear as the gaus or ac 20's which just is not true..not worth their weight

Edited by Peekaboo I C JU, 26 May 2013 - 12:48 AM.

[One Medic Army]

Sephlock, on 25 May 2013 - 10:31 PM, said:

Then how come training grounds QQing got LRMs unnecessarily nerfed in the first place?

Because the devs first confirmed it in actual gameplay.

[Kellea]

Thanks for your efforts. Bugged or not bugged traininggrounds, I agree with your assessment that the cat cores awfully quick.

Nevertheless I still think that lrm are UNDERpowered. Maybe the damage application is flawed BUT if my mech features 2 lrm 20 with artemis and a BAP I expect more than just an average of 200 dmg per match considering all the disadvantages lrm have (los, warning, ecm, flight time, etc.). Why do lrm only 0.9 dmg? Why not 1.0 as they should? Compared to especially AC/Gauss lrm still suck.

[Nightcrept]

One Medic Army, on 26 May 2013 - 12:57 AM, said:

Because the devs first confirmed it in actual gameplay.

The devs confirmed that splash damage was broken and made worse after the patch a few months ago. The problem is that the QQers wanted splash removed completely and lrms nerfed until they where fixed instead of the patch simply reverted to what we had been at least semi happy with. Which I said would lead us to even more game imbalances and make it harder to re-introduce lrms without QQing.

[Revorn]

Atm, iam not impressed by the LRMs, the Enemy seems to think the same.