Edited by IsaAurinkoinen, 22 September 2015 - 12:37 AM.
#21
Posted 22 September 2015 - 12:34 AM
#22
Posted 24 September 2015 - 05:40 AM
#23
Posted 24 September 2015 - 06:14 AM
IsaAurinkoinen, on 22 September 2015 - 12:34 AM, said:
Most players already and nearly every competetive player has armlock off anyways. Doesn't change a thing.
#24
Posted 26 September 2015 - 09:03 PM
Mercier, on 21 September 2015 - 09:28 AM, said:
( with, ironically, the exception of some arms - the one weapon mount that COULD converge as such )
is the heart of the TTK and meta problems...
It's not only bad for game play and game balance, it's physically impossible:
https://en.wikipedia...n_harmonisation
Canonically mech weapons have actuators that adjust the convergence point dynamically so in our case it is possible.
Though I agree otherwise, convergence (combined with large alphas) is definitely the main issue currently.
#25
Posted 01 October 2015 - 10:23 AM
#26
Posted 02 October 2015 - 08:31 AM
It has a 95 point heat scale. The only penalty for going over 95 points of heat is a momentary shutdown.
It also carries a magic a box that prevents all locks outside of 200 meters as well as missile locks inside 180 meters.
MWO has trully become a mockery of Battletech.
#27
Posted 02 October 2015 - 01:56 PM
Group-Fire is core gameplay. Alpha-Strike (not group-fire) can't all hit the same spot if you are moving. The solution is to make MWO's mechs tougher to keep up with the large weapons all recharging in 4 seconds or less. Something like that. However, I am not usually losing mech sections until after a good deal of combat has occured. I check for Banshees.
#28
Posted 02 October 2015 - 03:29 PM
#29
Posted 02 October 2015 - 06:40 PM
Also, reducing the sheer number of duplicate hardpoints on some mechs (cap it to stock build weapon numbers for the most part), or when that isn't possible, negative quirking mechs with tons of duplicate points, like what was done with timby and storm crow.
Hardpoint sizes and number reduction would reduce the number of heavy weapons that a mech can carry, so that they can be balanced to where, say, a single AC/10, AC/5 or PPC could be useful. Hardpoint number reductions would cut down on the medium laser spam. Both together would increase TTK, and stop the power creep in alpha strikes.
It also wouldn't hurt to make SRMs, AC/10, LBX, etc. less of a waste of tonnage.
Edited by Postumus, 02 October 2015 - 06:45 PM.
#30
Posted 02 October 2015 - 08:48 PM
Twist your torso to spread damage, you'll live longer.
There are a number of things that this game needs to improve on, weapon convergence to placate the TTers is not one of them.
#31
Posted 03 October 2015 - 03:50 PM
Quote
Except reducing the heat cap wont fix the problem.
It will just force everyone into using dual Gauss builds which generate virtually no heat. And then you'll end up with the same exact problem of high pinpoint alphas all over again.
The problem isnt heat. Anyone who thinks the problem is heat doesnt understand the game at all. The problem is pinpoint convergence and the lack of damage spreading mechanics in the game. Making heat more punitive doesnt fix convergence and convergence is what needs to be fixed or at least lessened with damage distribution mechanics.
Quote
Except mechs boating heavy weapons isnt the problem. Ghost heat and the 2 gauss limit already prevents mechs from firing more than 1-2 heavy weapons. Mechs boating LIGHT weapons like CERMLs is the problem. And hardpoint sizes do NOTHING to solve that problem.
Hardpoint sizes, just like a lower heat cap, completely fails to address the real problem of convergence.
Quote
duplicate hardpoints arnt the problem either. convergence is.
a mech firing 6 lasers would be perfectly balanced as long as those lasers dont all hit the same location.
again the best solution is to change the way damage is applied by not allowing it to all go into one location.
Edited by Khobai, 03 October 2015 - 04:25 PM.
#32
Posted 03 October 2015 - 05:40 PM
Khobai, on 03 October 2015 - 03:50 PM, said:
Except reducing the heat cap wont fix the problem.
It will just force everyone into using dual Gauss builds which generate virtually no heat. And then you'll end up with the same exact problem of high pinpoint alphas all over again.
The problem isnt heat. Anyone who thinks the problem is heat doesnt understand the game at all. The problem is pinpoint convergence and the lack of damage spreading mechanics in the game. Making heat more punitive doesnt fix convergence and convergence is what needs to be fixed or at least lessened with damage distribution mechanics.
This exactly. Spooky ghost heat does not prevent my robot from getting blown to hell in one single attack from one single opponent. The only thing ghost heat does is stop the same guy from likewise blowing up the robot next to me in exactly the same fashion immediately thereafter, ghost heat makes them wait a few seconds first.
Problem not solved.
The only things that will reduce TTK are:
reducing or eliminating weapon convergence (by no means adding random spread values, please no)
Hardpoint normalization: giving every mech equal potential for alpha strikes (why does a firestarter have 8 guns and a cicada 4? wut?)
Universal weapon damage debuffs
Universal armor buffs for everything (locust could use a little love here)
[Could take a look at the whole Gauss not generating any heat thing too, I've still got people asking me why the AC/15 gun doesn't make heat, the answer being "it explodes when hit sometimes" which is usually responded to with "That's ********"]
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users