

Missile Buff Please?
#21
Posted 10 April 2013 - 06:53 AM
#22
Posted 12 April 2013 - 12:14 AM
MrMainiac, on 09 April 2013 - 06:35 PM, said:
I don't know if you caught what I meant by damage per missile. You said a LRM5 should do 13.5 damage. If you split that up between the 5 missiles it shoots, that totals out to be 2.7 damage for each missile. You also stated that LRM10's should do 17 damage. That totals out to 1.7 damage for each missile.
What makes this and issue is a LRM5 weighs 2 tons, and a LRM10 weighs 5 tons. I could take 2 LRM5's and not only do more damage but save 1 ton for other things on your mech like ammo. Plus, if I use the same ammo between each missile launcher, shouldn't it do the same damage no matter what launcher I use?
I previously stated we should increase damage per missile to give missiles a buff so they're still damaging if they hit, but isn't taking a pair of AC20's to the mech.
The problem is LRM5's will always be more efficient whether we buff them or not as long as damage is based on missile count alone and not the tonnage of the launcher. The problem we had before was that launchers across the board did 1.8 damage per missiles LRM20's 1.8*20=36, LRM15's 1.8*20=27, LRM10's 1.8*10=18, LRM5's=9. It's fine if you want to balance the LRM5 but it has to take its weight AND heat output into consideration, it can't just be about damage per missile. If you just say that LRM's across the board do X damage per missile, LRM5's will always be more efficient when it comes to damage if damage is only calculated based on missile count.
However, if we're going to look at damage per ton we should also look at the heat output scale. The trade off right now is that the LRM5 system is more efficient when it comes to damage per ton but it also produces more heat per ton compared to the larger missile systems.
If the logic behind the larger missile systems producing more heat is that they do so because they're firing more missiles, then we would have to increase the heat output of the LRM15 and the LRM20. If we also use weight to calculate heat then we'd also have to decrease the amount of heat produced by the LRM5.
The only way meet these requirements would be to make all the LRM systems produce 0.40 units of heat per missile. This would put the LRM5 at 1.6, the LRM15 at 5.6 and the LRM20 at 8 and keep the LRM10 at 4. If you just tried to lower the heat ouput of the LRM5, the LRM10 would still be at a disadvantage over the 15 and 20 and the LRM15 would still be at a disadvantage over the 20. If you left the LRM5 alone and raised the heat ouput of the others just based on tonnage, heat still wouldn't adhere to a linear scale if you're also taking the number of missiles into account.
To summarize, the adjusted damage scale with corrected heat output would look like this:
LRM20's @ 24.0 damage, 8.0 heat for 10 tons (1.20 damage and 0.4 heat per missile)
LRM15's @ 16.8 damage, 5.6 heat for 7 tons (1.12 damage and 0.4 heat per missile)
LRM10's @ 12.0 damage, 4.0 heat for 5 tons (1.20 damage and 0.4 heat per missile)
LRM5's @ 4.80 damage, 1.6 heat for 2 tons (0.96 damage and 0.4 heat per missile)
Again, just fixing the damage and heat scale wouldn't be the best answer. Without spreading the damage around we'd also be increasing the lethality because it would still all be directed at the center torso.
#23
Posted 12 April 2013 - 01:00 AM
Tyr4nt, on 12 April 2013 - 12:14 AM, said:
The problem is LRM5's will always be more efficient whether we buff them or not as long as damage is based on missile count alone and not the tonnage of the launcher. The problem we had before was that launchers across the board did 1.8 damage per missiles LRM20's 1.8*20=36, LRM15's 1.8*20=27, LRM10's 1.8*10=18, LRM5's=9. It's fine if you want to balance the LRM5 but it has to take its weight AND heat output into consideration, it can't just be about damage per missile. If you just say that LRM's across the board do X damage per missile, LRM5's will always be more efficient when it comes to damage if damage is only calculated based on missile count.
However, if we're going to look at damage per ton we should also look at the heat output scale. The trade off right now is that the LRM5 system is more efficient when it comes to damage per ton but it also produces more heat per ton compared to the larger missile systems.
If the logic behind the larger missile systems producing more heat is that they do so because they're firing more missiles, then we would have to increase the heat output of the LRM15 and the LRM20. If we also use weight to calculate heat then we'd also have to decrease the amount of heat produced by the LRM5.
The only way meet these requirements would be to make all the LRM systems produce 0.40 units of heat per missile. This would put the LRM5 at 1.6, the LRM15 at 5.6 and the LRM20 at 8 and keep the LRM10 at 4. If you just tried to lower the heat ouput of the LRM5, the LRM10 would still be at a disadvantage over the 15 and 20 and the LRM15 would still be at a disadvantage over the 20. If you left the LRM5 alone and raised the heat ouput of the others just based on tonnage, heat still wouldn't adhere to a linear scale if you're also taking the number of missiles into account.
To summarize, the adjusted damage scale with corrected heat output would look like this:
LRM20's @ 24.0 damage, 8.0 heat for 10 tons (1.20 damage and 0.4 heat per missile)
LRM15's @ 16.8 damage, 5.6 heat for 7 tons (1.12 damage and 0.4 heat per missile)
LRM10's @ 12.0 damage, 4.0 heat for 5 tons (1.20 damage and 0.4 heat per missile)
LRM5's @ 4.80 damage, 1.6 heat for 2 tons (0.96 damage and 0.4 heat per missile)
Again, just fixing the damage and heat scale wouldn't be the best answer. Without spreading the damage around we'd also be increasing the lethality because it would still all be directed at the center torso.
Hm the Commando have a size of 6m in height and 3m width.
Means a splash radius 2m and with secundary hitzones...will affect the whole mech.
a single LRM will deal damage to every location
The Atlas is 12m in height and 6 in width?
The same splash radius will hardly effect the same count of hitzones.
If you consider that the commando is more difficult to hit it could be fine. Because if a Commando get a whole LRM 20 salvo into the chest it got serious problems...no matter if the LRM did 1.2 damage on hit without splash damage or 0.4dmg with splash damage or anything else.
I'm still thinking that the splash damage is the best idea to spread the damage over the whole target -> sanding of its armor.
Making pure LRM worthless because you have to remove all the armor of the target before dealing damage.
#24
Posted 12 April 2013 - 01:36 AM
Karl Streiger, on 12 April 2013 - 01:00 AM, said:
Hm the Commando have a size of 6m in height and 3m width.
Means a splash radius 2m and with secundary hitzones...will affect the whole mech.
a single LRM will deal damage to every location
The Atlas is 12m in height and 6 in width?
The same splash radius will hardly effect the same count of hitzones.
If you consider that the commando is more difficult to hit it could be fine. Because if a Commando get a whole LRM 20 salvo into the chest it got serious problems...no matter if the LRM did 1.2 damage on hit without splash damage or 0.4dmg with splash damage or anything else.
I'm still thinking that the splash damage is the best idea to spread the damage over the whole target -> sanding of its armor.
Making pure LRM worthless because you have to remove all the armor of the target before dealing damage.
That's why we were talking about spreading the damage to the arms in addition to the left and right torso. Even still, just adding the left and right torso and reducing the incoming damage to the center torso would still be better than having it all go into the CT no matter what mech you're piloting. The fact that right now all the damage is going into the engine, even if you're using a standard engine, is a problem unto itself.
#25
Posted 12 April 2013 - 08:17 AM
I prefer the idea of splash damage for missiles vs a straight damage increase as if properly implemented it would be more realistic in that people should be reluctant to fire 100's of LRM's at an enemy mech when its being circled by a teammate as they will probably kill their teammate as well.
#26
Posted 12 April 2013 - 08:42 AM
I agree that the nerf has ruined many missle platform mechs, my favorite is the AWS-8V, I haven't used it for weeks; there are no ballistics on it and not enough energy points to justify changing it, so it sits collecting dust. PGI needs to remove splash damage, period and to restore missle damage, but they won't because the grouping is also messed up. At least for LRM's, for SRM's this excuse makes no sense. SRM's should've been, and could've been fixed by now; and the LRM fix could've come later.
Now don't get me wrong, splash damage is a neat effect, but there is no real way to control it. Maybe one day there will be, but even then you will have to reduce the amount of damage to almost nothing. LRM's should get a splash of .15 damage, and SRM's should get a splash of .25 damage. The low amounts reduce the chances of splash mechanic abuse, and still light up the paper doll enough to scare the crap out of someone being hit by them. Just my two cents on the matter.
#27
Posted 12 April 2013 - 12:37 PM
I never understood the LRM's aren't primary weapons argument. They were always primary weapons of those mechs that used them with only a few exceptions (and even in those cases, the LRM racks used were always a smaller rack than what is normally used in that weight class). Considering their weight investment, they should be something to be feared.
But I'm also of the mindset that AMS should be buffed. It should be able to destroy more missiles than it currently does. I'm not sure what the mechanic is set to right now, but AMS should kick in at about 300m and shoot down probably about 10 missiles per second while it has the ammo for it. Drop the ammo per ton to 500 and call it a day.
#28
Posted 13 April 2013 - 10:57 AM
#29
Posted 13 April 2013 - 07:45 PM
I think that SRMs are at a pretty good point atm with damage. Not too much not too little (so it seems for me)
LRMs havent used them: Arent as scary but could probably use a bit of a buff. Maybe up to 1.0 dmg? (At what now 0.7?)
#30
Posted 14 April 2013 - 05:33 AM
based of said recode we can propose changes.
what i personally want to see
cloud formation - in all documents lrms are DUMB, so no advanced homing systems etc - just barely flying on target. no spirals hitting 1-2 specific location - reasonably large cloud hitting area.
change in lock on mechanics - i assume currently lrm homing system target center of the mech model. what i wanted to see is said system assuming as target specific location im currently targeting. ar- ok center of cloud is arm. with relocking after each salvo, so for example medium mech coming up front on ideal distance will have most hits on said arm, less hits on side torso\leg and way less hits on ct with rarely any on cabin, with large portion of missiles go MISS (30% or more) because center of the cloud is arm and target mech takes less than half of cloud area. same with ct - most ct, less side torsos, legs, way less arms\head, some misses. this will not result in head sniping since while targeting head a very big portion of cloud will just go to waste and cabin is too small to receive critical number of hits per salvo.
save the splash (partially for dumb fire so missiles hitting ground will damage legs) but make it reasonably small (no more than 2 additional sections) with nonlinear damage drop with random mechanics for ground (say 3m max splash) and mech (no more than 2 additional sections, maybe with fixed damage nerf for ct\cabin splash)
up the damage.
this way lrms will require way more skill. boating will be not so useful - with additional launchers cloud will just be bigger, and each salvo will target location under cross hair in moment of launch. and dedicated mechs like ctpl will be preferable - ability to launch 40 lrms at once with one lucky lock on on ctpl, or need to maintain lock on on specific location for all duration of said 40 rlms launching from 15 tubes of fatlas.
#31
Posted 14 April 2013 - 08:16 AM
Actually the LRMs and SRMs in MW games are more rockets than missiles. LRMs are a fire support weapon more like a Soviet/Russian BM-21 - the difference being that they have a rudimentary guidance system that takes them to the general vicinity of a target, not so pinpoint. I have had good results dumb-firing LRMs on targets (ala BM-21) and the targets don't get an alert because I am not locking onto them. SRMs are patently rockets since we just point and shoot them. SSRMs are missiles in the strictest sense, being a rocket which is guided.
I can't wait for SLRMs and SMRMs to come in and hear all the QQ.
Edited by Gremlich Johns, 14 April 2013 - 08:17 AM.
#32
Posted 15 April 2013 - 09:25 AM
However, I dont like the reduction in damage not coming with a reduction in LRM-counter items and also no increase in ammo.
Ideally:
1- Leave LRM damage as is but increase its splash damage.
2- Remove missile warning message.
3- Double ammo per ton. ***
*** This is because currently, to take down a single, static, 4LRM15+TAG+Artemis catapult you need to fire 3 tons worth of ammo at it. Its just silly to expect a missile support mech to waste SO much tonnage (weapon+ammo) to carry a weapon system that is barely good to whack a heavy mech. The damage is ok as it is.. its not insta kill (lrm super-damage patch) nor is landing two volleys causing severe hurt (how lrms were before the super-damage patch). The LRMs now do the equivalent damage of being hit by two LBX10's ...and the damage is spread all over the mech's torso. Its not good but its not bad. The range of the LRM makes up for it.
But as it is now, low ammo and the current low damage is countering the range advantage. Give us more ammo per ton. Easy, simple fix. Remove the missile warning message...if the volleys are not significantly damaging then the warning no longer serves the purpose it had before and instead actively nerfs the use of missiles as long range weapons. Added splash zone on impact would or should help the issue of LRM's not applying damage correctly when a mech is moving faster than 100kph or when it is moving in the vertical plane (dropping from cliff, climbing steep slope of JJetting up/down).
#33
Posted 18 April 2013 - 10:38 AM
#34
Posted 18 April 2013 - 12:01 PM
If they are dealing .7 damage per missile and the avg pug lands 25% of their missiles the potential outcome is very very low.
Lets say you carry 1000lrms.
25% of 1000 is 250.
.7 per missile x 250 = 175
So out of 1000 missiles the avg pug has a potential of 175 damage.
Even the best players only hit between 30 and 40 percent of the time for a avg potential of 210 to 280 damage.
Even if you hit with all 1000 missiles your max damage potential is only 700.
Combine the poor damage potential with the plethora of game mechanics further restricting lrms and they are near useless at the moment.
#35
Posted 18 April 2013 - 01:22 PM
After we've done that we can talk about speeding up LRMs or upping the damage to them. SRMs should always do ~2x LRM damage/missile, so if LRMs get a damage buff, then SRMs will as well.
#36
Posted 18 April 2013 - 01:39 PM

Weapons should be balanced by DMGperTONN.
So 1 tonn of LRM ≈ 1tonn of SRM ≈ 1tonn of AC/x ammo.
Edited by Mokou, 18 April 2013 - 01:43 PM.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users