647 replies to this topic

[Haitchpeasauce]

Thanks Paul for posting, it's quite a relief to know that we're onto something. And thank you again Amaris for starting this discussion!

I have some results from the Testing Grounds.
Will be testing this on Live as well later on on mechs such as the Raven.

Edits: Thanks to Thontor and Scop for spotting an error in my formula for total HP. The result, however, makes missiles appear to do even more damage than originally calculated.

The mlas damage is an average, calculated using the final % health and the known HP of the mech.

The srm and lrm damage is an average per individual missile, that has been normalised according to the laser results.

Expected damages:

• mlas = 5 damage per shot
  
• srm = 2.5 damage per missile
  
• lrm = 1.8 damage per missile

COM-1B (128 armor, total 223 HP)
mlas: 5.13 (103%)
srm2: 19.11 (764%, however components were destroyed in a single hit)
srm6: 17.39 (696%)
lrm5: 8.48 (471%)

JR7-D (128 armor, total 253 HP)
mlas: 5.14 (103%)
srm2: 13.06 (523%)
srm6: 13.73 (549%)
lrm5: 6.64 (369%)

CN9-A (272 armor, total 447 HP)
mlas: 5.36 (107%)
srm2: 10.38 (415%)
srm6: 10.42 (417%)
lrm5: 1.75 (97%)

CPLT-A1 (384 armor, total 601 HP)
mlas: 5.11 (102%)
srm2: 6.47 (259%)
srm6: 7.11 (284%)
lrm5: 4.08 (227%)

CTF-1X (352 armor, total 575 HP)
mlas: 5.18 (104%)
srm2: 6.53 (261%)
srm6: 7.18 (287%)
lrm5: 2.89 (160%)

AS7-D (612 armor, total 925 HP)
mlas: 5.32 (106%)
srm2: 6.38 (255%)
srm6: 5.56 (222%)
lrm5: 2.35 (130%)


Observations:
mlas appeared to be doing correct damage and is functioning as the control figure here (variances are due to rounding errors from the % health reading).

The Commando received around 750% of the expected srm damage. 6 locations were hit on average per shot.
The Jenner received about 530% of the expected srm damage. 5 locations were hit on average per shot.
The Centurion received about 410% of the expected srm damage. 4 locations were hit on average per shot.
The heavy/assault mechs received 260% of the expected srm damage. 3 locations were hit on average per shot.

The Commando received 470% of the expected lrm damage
The Jenner received 370% of the expected lrm damage.
The Centurion received 94% of the expected lrm damage. Correct damage values? Incorrect method?
The Catapult received 230% of the expected lrm damage. Higher than other mechs of this weight class.
The Cataphract and Atlas received about 140% of the expected lrm damage.


Method:
All mechs are assumed to have trial mech armor.

Total HP of the mech was calculated with the following formula:
[Base Armor] + (([Max Armor] - 18) / 2) + 15

mlas were fired 10 times at a range of 25m at different body parts as a control. The final health % was noted. Repeated this several times for six different mechs.

The mlas damage is assumed to be according to official ratings (i.e. 5 damage), and the discrepancy is used to normalise the missile damage. The discrepancy is due to rounding of the % health and is considered negligible.

Then, srm2 were fired at the upper centre torso at a range of 25m, aiming to avoid splashing the legs. This produces the "worst case scenario", and also the leg damage registration is known to be bugged. Where possible, multiple shots were fired to get better averages.

Srm2 were used over ssrm2 because they hit the target more reliably; ssrm2 can strike the legs. Remember, this is a test about splash damage creating a multiplying effect on damage.

srm6 were fired at 25m to test if the per-missile damage of srm6 is in line with srm2.

Finally, lrm5+Artemis were fired at a range of 200m. It was noted that one missile will miss the target from time to time.


Comments:
mlas appear to be dealing the correct damage,

Although the tests were conducted in the Testing Grounds, the relative difference of mlas damage to srm and lrm missile damage is very apparent.

Hitbox geometry, especially the size of these hitboxes, appears to strongly influence the total damage taken.

Edited by Haitchpeasauce, 14 March 2013 - 08:21 PM.

[Nightcrept]

Thontor, on 14 March 2013 - 05:16 PM, said:

You had me until this paragraph

As I understand it right now, missiles are already doing their listed damage to whatever location they hit, plus splash damage to other locations within the splash damage radius.

Removing splash damage will simply remove the extra damage, the damage to the location where the missiles hit should in fact be less if splash damage is removed.

Looks like he though the damage was evenly distributed.

[Hayashi]

There's no need to go about accusing people of trolling, just need to see everything from as many points of view as possible.

As long as the area of impact of missiles is the same, there will be a balance issue no matter whether splash happens or not. Lasers, ballistics don't have this issue as they deal effectively pinpoint damage no matter what according to player skill - smaller mechs are simply harder to hit at the same area as their hitboxes are smaller, larger mechs are easier to hit in a single area but carry more armour per area.

But missiles, being auto tracking, are rather different. Should the impact area be the size of a Commando, fixed, without splash, then the same grouping used on an Atlas will result in all missiles impacting the torsos, with zero hitting the arms. This makes LRMs more effective against Mechs the larger they get as it increases the chances they will impact the Centre Torso (and sides, if XLs are used); combined with the slower speed of larger mechs, this makes large mechs disproportionately weak towards guided missiles. Should the impact area be the size of an Atlas, fixed, your light mechs are going to be impossible to kill.

Should splash be taken into account, damage is spread should the system hard limit the damage to the maximum firepower allowed by the weapon (2.5 per SRM, 1.8 per LRM IIRC). This still leads to tighter grouping on Assaults, but not as much on Lights. This mitigates the chances of them being used as an I-win button, but not completely.

Ideally, the hit radius for LRMs and SRMs should vary according to the target they hit - the smaller the target is, the tighter the grouping they will have - and splash will be limited by maximum firepower. So mechs which are small, or if you aim in between two hitbox areas for SRMs, will end up doing more splash damage to other components, but less direct damage due to the direct limiting mechanism. I'm not sure if this is possible to implement, I'm not privy to the code they're using here. =P

p4r4g0n, on 14 March 2013 - 05:11 PM, said:

6 X SRM6 X 2.5/missile = 90 dmg without splash. With splash and from pointblank range, the damage done would be 90 base dmg + splash damage. I don't understand how removing splash makes it any easier to 1 shot any mech in the rear centre torso.

Could you clarify please?

AFAIK, splash =/= missile spread

By not removing splash, what I'm hoping to see here is that instead of being 90 dmg + additional splash, the splash mechanism is kept but the damage is dialed down to max at 90 if all missiles hit. Or to be more clear, to max at 2.5 for every missile, so that if 20 missiles hit, you'll do 50 damage. But even if all 20 missiles hit the centre and side torsos, some of this 50 damage will spread to the legs and arms, whereas if 50 damage was dealt by, say, ballistics on the same area, all 50 damage will be restricted to the areas where they hit only.

If the total damage is to be reduced, then it will be equally easy to one shot mechs as at present. In a way I'm talking about keeping damage to 90 (which involves reduction of damage from current values) and keeping splash as a damage allocation mechanism, not about keeping splash as an additional damage bonus.

Then once this whole damage allocation problem is fixed, we can start using the playerbase usage statistics and damage dealt statistics for LRMs, Streak SRMs and SRMs to determine whether the missiles as a class require a further buff or nerf accordingly. At the moment, since they're doing wildly different amounts of damage to different targets, there's no common ground on which to make any balancing change.

---

A lot of the confusion here stems from the problem that we're attempting to argue three different issues at the same time:

• 1. Total damage output of missiles
  
• 2. Damage allocation of missiles via missile spread (unique to missiles and LBX-AC/10)
  
• 3. Damage allocation of missiles via splash damage per missile impact (unique to missiles only)

As of current iteration, removing splash without compensation reduces #1, keeps #2 constant and zeroes #3, which is in effect a major nerf to all missiles. Removing splash with compensation to damage keeps #1 constant, increases #2 (because of compensation) and zeroes #3, which unfairly biases the missiles against larger mechs as an equal spread area is a far tighter grouping on a larger target than a smaller one. Hard limiting damage while keeping the splash mechanism reduces #1 to programmed values, and reduces #2 and #3 proportionately; if we upscale the total damage later it keeps all three in balance as at present, but reduces the absurd vulnerability of certain mechs like the Commando and increases the vulnerability of zombie mechs like the Centurion. Hard limiting damage and reworking splash boxes to be proportional to the targets they impact will result in keeping the balance somewhat, but in addition to changing the absurd vulnerabilities of certain chassis based on their hitboxes, will also reduce the vulnerability of larger mechs to SRM spreads relative to smaller Mechs. It's the last one I favour in terms of optimal final balance, though it's pretty clear that it is also the most difficult solution to code.

I do see that removing splash entirely and upping damage to compensate is a very fast bandaid solution, but if that is done it won't be long before Atlas users notice how fast they now start to die, and the light Mech domination will continue in the meantime.

Edited by Hayashi, 14 March 2013 - 05:53 PM.

[Nightcrept]

Warrax the Chaos Warrior, on 14 March 2013 - 05:23 PM, said:

I think the miscommunication here is that your definition of a "bug" is different from what everyone else's is.

The software is doing something that the developers didn't expect it to do= bug. The fact that there may or may not have been a flaw in the actual code is irrelevant, it's still a "bug".

I think your right. I think they knew about it and either choose to ignore it like ssrms and ecm or don't know how to fix it. Thus calling it a bug is a bit nice considering their past balancing.

[Acid Phase]

Prosperity Park, on 14 March 2013 - 03:54 PM, said:

You're saying that PGI is catering to people who say that SRM damage is fine as it is? Um... I disagree.

Let's see what PGI has to say about SRMs, before determining who they are catering to.

Not on SRMs. I'm talking about catering to those who have ridiculous builds that are capable of instant kills. Case in point why do you think the time limit is what it is in these matches? Because mostly everyone wants the quick and easy. Quick and easy kills, quick and easy caps. I'm betting they are also quick to defend PGI's errors....because they want to stay with the upper hand. Little do they know is that in BT, mech battles are actually long lasting battles.

All I see out there is SRM intensive builds. I can guarantee that if the SRMs get tweaked to what they should be, how many out there will still be using it. Since the closed beta, I've seen no changes to the things that matter the most. Hence the OPs point figuring how these weapon balances are all out of wack.

[Nightcrept]

Hayashi, on 14 March 2013 - 05:26 PM, said:

There's no need to go about accusing people of trolling, just need to see everything from as many points of view as possible.

As long as the area of impact of missiles is the same, there will be a balance issue no matter whether splash happens or not. Lasers, ballistics don't have this issue as they deal effectively pinpoint damage no matter what according to player skill - smaller mechs are simply harder to hit at the same area as their hitboxes are smaller, larger mechs are easier to hit in a single area but carry more armour per area.

But missiles, being auto tracking, are rather different. Should the impact area be the size of a Commando, fixed, without splash, then the same grouping used on an Atlas will result in all missiles impacting the torsos, with zero hitting the arms. This makes LRMs more effective against Mechs the larger they get as it increases the chances they will impact the Centre Torso (and sides, if XLs are used); combined with the slower speed of larger mechs, this makes large mechs disproportionately weak towards guided missiles. Should the impact area be the size of an Atlas, fixed, your light mechs are going to be impossible to kill.

Should splash be taken into account, damage is spread should the system hard limit the damage to the maximum firepower allowed by the weapon (2.5 per SRM, 1.8 per LRM IIRC). This still leads to tighter grouping on Assaults, but not as much on Lights. This mitigates the chances of them being used as an I-win button, but not completely.

Ideally, the hit radius for LRMs and SRMs should vary according to the target they hit - the smaller the target is, the tighter the grouping they will have - and splash will be limited by maximum firepower. So mechs which are small, or if you aim in between two hitbox areas for SRMs, will end up doing more splash damage to other components, but less direct damage due to the direct limiting mechanism. I'm not sure if this is possible to implement, I'm not privy to the code they're using here. =P

I was curious if they could do what you suggest in the last part of your post. I am not a programmer to know and the one posting didn't respond.

[IrrelevantFish]

Nightcrept, on 14 March 2013 - 05:04 PM, said:

I understand the numbers balance players are seeking but I also see the in game effects of removing it entirely without counter balancing.

The problem is from my pov that the whole system is in parts intertwined with splash damage.

I agree with you, somewhat. The cries for the immediate elimination of splash are both unreasonable (coding is hard) and foolish. Just turning off splash will reduce missiles to a fraction of their current effectiveness, and while missiles are OP right now, they're not that OP.

But that's in the short-term. Long-term, I think splash will continue to be a problem. At the least, it will require tuning and testing every new chassis's vulnerability to missiles.

Though I shudder to think at how much time the devs had to spend putting splash damage in, I think they should completely remove it, but only after extensive testing/tinkering. I'm sure they'd need to adjust cool-down time, damage, and spread to provide true balance, and I want them to take the time to do it right.

[DrSecretStache]

Here's my take on splash:

They can still basically keep splash damage by both adjusting the spread of the missiles so they aren't in such a tight arc, as well as increasing the damage by some margin. This achieves the same effect as a "splash," with very real, easy to balance numbers. I'd much rather they have splash in this way. It may not actually be splash as we think of it, but it would achieve the same effect. (Of course, increased damage is a possiblity, I would much rather they stay at 2.5 )

Glad the devs have reported on this. Thanks Paul.

Edited by Zoughtbaj, 14 March 2013 - 05:37 PM.

[Sparks Murphey]

Kell Draygo, on 14 March 2013 - 01:04 PM, said:

This testing should've been thoroughly done before the terrible ECM was added into the game to balance streaks.

ECM was never introduced simply to balance Streaks. It's true that the devs delayed applying balancing measures until ECM was in, but ECM was always a planned feature of the game, and as a thing likely to massively affect Streaks wasn't in the game yet, there was no point tweaking them until after it arrived and the whole equation could be observed.

As for arguing it should have been done thoroughly before in general...it sounds like it was. Back when they implemented SRMs. They had a number, checked that the code had that number, took it out into the field and tested it. At the time, though, they only had the 4 'Mechs to test it on, and with hindsight we now find that that was a skewed experiment, but when it was first conducted, no evidence indicated a need for change. Now that there is evidence, they're addressing the issue.

Acid Phase, on 14 March 2013 - 05:30 PM, said:

Not on SRMs. I'm talking about catering to those who have ridiculous builds that are capable of instant kills. Case in point why do you think the time limit is what it is in these matches? Because mostly everyone wants the quick and easy. Quick and easy kills, quick and easy caps. I'm betting they are also quick to defend PGI's errors....because they want to stay with the upper hand. Little do they know is that in BT, mech battles are actually long lasting battles.

Long lasting battles? It might take hours of standing around the table, but a long, drawn out, 20 turn game ("can I go home yet?") of BattleTech represents just over 3 minutes of fighting. If you finish a MWO game in 3 minutes, people are generally pretty astounded at the sheer speed with which they were taken down.

Though, admittedly, it might be a bit quicker if PGI hadn't doubled armour values on MWO 'Mechs. Because doubling armour values is all about the quick and easy kills.

[Scop]

Haitchpeasauce, on 14 March 2013 - 05:24 PM, said:

COM-1B (128 armor, total 198 HP)

I fear your figures may be off, at least for the COM. Check the image on the right from Paul's post below.

Paul Inouye, on 14 March 2013 - 12:53 PM, said:

You can add the values shown for 98 IS and 108 armor, totaling 203 overall HP. So this may just be an assist on Thontor's post, http://mwomercs.com/forums/topic/105185-streak-srm-damage-is-much-higher-than-expected-test-results-inside/page__st__440__p__2061407#entry2061407. Hope that helps your figures!

[N0MAD]

Im thinking the devs need to make a decision on what type of Amo a misile is, is it a HE round or is it an AP round, it shouldnt be both. If its a HE round it will cause splash damage to a wide area BUT have no efect on armor, as in it will not be efective until armor has been depleted in a certain area. If its a AP round it will penetrate or damage armor BUT only at the point of impact.
To have it be both HE and AP is just not posible or fair.

[Nightcrept]

Sparks Murphey, on 14 March 2013 - 05:37 PM, said:

As for arguing it should have been done thoroughly before in general...it sounds like it was. Back when they implemented SRMs. They had a number, checked that the code had that number, took it out into the field and tested it. At the time, though, they only had the 4 'Mechs to test it on, and with hindsight we now find that that was a skewed experiment, but when it was first conducted, no evidence indicated a need for change. Now that there is evidence, they're addressing the issue.

And you just hit on why I hate calling it a bug.

Splash damage has been here. It isn't new. So when they tested the new mechs the new mech were taking splash damage. So
new mechs are balanced against the damage dealt out by the splash damage modifiers.

So while the devs for some reason failed to look at what was causing the damage the effects the damage had on the survival of the mech and therefore the end results were very visible.

[IrrelevantFish]

Voidsinger, on 14 March 2013 - 05:20 PM, said:

Core damage is applied to the location hit. Splash damage would be applied to the percentage of surface area in the sphere covered by the explosive radius (even if this is the same location as core).

Unfortunately, this would require some truly nasty calculations, and is a completely impractical solution when you have to perform those calculations for 100+ simultaneous missile impacts.

Which brings me to my true problem with splash damage: in the context of MWO, I don't think there's any way to implement it properly without using copious amounts of computing power.

[Deathlike]

Removing splash altogether will immediately nerf LRMs, and since LRM OP threads are the current norm, you will now get a surge of LRM are UP threads as a consequence. Guys, moving the targets is not an option. This has to be dealt with and fixed. Splash damage has always meant to be proportional damage, so they have to figure out the #s here and go from there. Removing splash will have unfortunate and unintended consequences.

[HRR Insanity]

Deathlike, on 14 March 2013 - 06:16 PM, said:

Removing splash altogether will immediately nerf LRMs, and since LRM OP threads are the current norm, you will now get a surge of LRM are UP threads as a consequence. Guys, moving the targets is not an option. This has to be dealt with and fixed. Splash damage has always meant to be proportional damage, so they have to figure out the #s here and go from there. Removing splash will have unfortunate and unintended consequences.

You can't fix it without removing splash. Since there are hitboxes, smart/good pilots will always aim for hitbox interfaces and do double or more damage. If you tune for those pilots, then poor/dumb pilots will think SRMs are underpowered.

Remove splash.

[Tekadept]

From what I read and understand wouldn't a simpler solution be if you want to keep splash damage change it so that you allocate each missile a finite possible amount of damage it can inflict? If this is already how it works then I merely don't understand how this currently problem is happening.

A Missile can do "X" Damage to a direct hit location, and it can do "Y" Damage as splash Damage.
When a missile hit is recorded it does "X" Damage to the location it hit, then the system works out all the possible splash locations and "Y" damage is divided up and allocated to those locations and deducted from the amount of damage left available.

So Once all that damage is used up that's it for that missile. So it isn't possible for a Missile to do more then the "X" + "Y Damage amount. This factors in that a Missile will only have a Finite amount of destructive force.

So if a missile hits the Center Torso it will do "X" Damage to that location. The system then determines all the possible locations for splash damage, then it divvies up the "Y" damage amongst those locations. Then even with complex geometry for any future mech designs their wont be an overlap problem and this wouldn't recur again as a Missile can then only do a maximum damage of "X" + "Y".

Complex Geometry will be still Bork Splash Damage, but the side effect will be the splash damage is just spread too thin rather then be amplified.

Of course the simplest solution is to just remove it.

Edited by Tekadept, 14 March 2013 - 06:55 PM.

[p4r4g0n]

Hayashi, on 14 March 2013 - 05:26 PM, said:

-snip-

Thank you for your clarification. If I understood it correctly, you are saying that the damage done inclusive of splash damage should be limited to the listed damage per missile. I think no one would have an issue with this as long as the amount of splash damage is a known fixed number rather than a variable one.

Given the probable difficulty of implementing the optimal solution which you pointed out requires a variable damage effect based on chassis, I'm thinking this is unlikely. Even if possible, how could this work if chassis identification is impossible with ECM (in its present form) is in effect?

[HRR Insanity]

Tekadept, on 14 March 2013 - 06:20 PM, said:

From what I read and understand wouldn't a simpler solution be if you want to keep splash damage change it so that you allocate each missile a finite possible amount of damage it can inflict? If this is already how it works then I merely don't understand how this currently problem is happening.

"X" Damage to the hit location, and then allocate another "Y" Damage to splash damage. When the missile hits, the possible splash locations are worked out , then the damage allocated for splash is split between the possible splashes, and deducted from the amount of damage left available, Once all that damage is used up that's it. So it isnt possible for a Missile to do more then Point Damage. This factors in that a Missile will only have a Finite amount of destructive force.

So if a missile hits the Center Torso it will do "X" Damage to that location. The system then determines all the possible locations for splash damage, then it divvies up the "Y" damage amongst those locations. Then even with complex geometry for any future mech designs their wont be an overlap problem and this wouldn't recur again as a Missile can then only do a maximum damage of "X" + "Y".

Still broken. Then if you hit an arm (for example), X is done to the arm and Y is done to the side torso. This would be very different than a CT hit, where X is done to the CT and Y is split between 3-4 other panels.

Just make it hit the spot is hits. Splash damage isn't needed. That's why we have multiple missiles in an LRM/SRM launcher. To do damage to multiple locations.

[Tekadept]

HRR Insanity, on 14 March 2013 - 06:26 PM, said:

Still broken. Then if you hit an arm (for example), X is done to the arm and Y is done to the side torso. This would be very different than a CT hit, where X is done to the CT and Y is split between 3-4 other panels.

Just make it hit the spot is hits. Splash damage isn't needed. That's why we have multiple missiles in an LRM/SRM launcher. To do damage to multiple locations.

That is true and I Agree it would still be broken. But at least the damage a Missile does would be a known set amount, rather then how it seems to be "Amplified" as it is now

I advocate for it to be completely removed as i agree with your statement re mutiple tube launchers But I Wont hold my breath on them removing it..

Edited by Tekadept, 14 March 2013 - 06:51 PM.

[HRR Insanity]

Tekadept, on 14 March 2013 - 06:35 PM, said:

That is true and I Agree it would still be broken. But at least the damage a Missile does would be a known set amount, rather then how it seems to be "Amplified" as it is now.

Or... you could just remove splash damage and solve the problem immediately.