Make it so it only affects the mech that has it.
The mech with ECM is invisible on enemy radar.
The mech with ECM cant have missiles locked onto it unless its being countered or tagged.
The mech with ECM is more difficult to get detailed target information from.
ECM does NOT affect allies appearing on enemy radar and does not affect enemy radar being able to detect members of their own team.
I honestly believe this is the most simple and elegant solution to the problem.


My Best Solution To Balancing Ecm
Started by Space Odin, Mar 30 2013 10:09 AM
4 replies to this topic
#1
Posted 30 March 2013 - 10:09 AM
#2
Posted 30 March 2013 - 12:31 PM
I think this is a good way forward. If not removing the radius entirely it needs to be WAY WAY WAY smaller. Like 40-50 meters. Standing that close to a mech that'll be a fire-magnet in about ten seconds if it gets noticed can be a liability, and still work within both fluff, and stated designer parameters for how they want ECM to function. I'm interested to see what they Devs are going to do to 'adjust it'. I wish they'd start an open development dialogue with us. 
But yeah, this is definitely the most simple way to balance ECM, but they're probably not going to do the most simple thing.

But yeah, this is definitely the most simple way to balance ECM, but they're probably not going to do the most simple thing.
#3
Posted 30 March 2013 - 01:50 PM
I second this suggestion, It still makes ECM good and viable for those that have it without making it overpowered (like it is now)
#4
Posted 30 March 2013 - 02:04 PM
http://mwomercs.com/...m-guardian-ecm/
This is a far better idea, and based not only in TT, but how it worked in other MW games when people didn't complain it was totally broken.
This is a far better idea, and based not only in TT, but how it worked in other MW games when people didn't complain it was totally broken.
#5
Posted 30 March 2013 - 03:41 PM
CancR, on 30 March 2013 - 02:04 PM, said:
http://mwomercs.com/...m-guardian-ecm/
This is a far better idea, and based not only in TT, but how it worked in other MW games when people didn't complain it was totally broken.
This is a far better idea, and based not only in TT, but how it worked in other MW games when people didn't complain it was totally broken.
But it wasn't any good either. ECM was only marginally useful in any other Mechwarrior title. I think PGI has obviously decided to move away from that. If they make ECM mechs targetable, with just a lengthened missile lock time, it will represent a gigantic shift in the meta. More than likely any adjustment to ECM will be some rebalance of existing ECM capabilities.
I also dispute the relevance of the tabletop to MWO's missile system. There's a lot of overlap between tabletop and MWO, surely, but one place where there is none, other than names, is how missiles work. TT utilizes dice, and only dice, to determine how many missiles hit. That makes ECM's flat missile hit reduction possible. It however, does not lend itself well to current Mechwarrior, since it'd require the design, testing, and implementation of several new missile paths. (Something along the lines of a new path for NTg-NAr, Tg-NA NT-Ar Tg-Ar would be required at least.) Losing Locks at Random Intervals is also a lame way to implement that. It arbitrarily ***** you over, which, I'd argue, is worse than requiring a TAG. Decreasing detection range is interesting, but it just sounds kinda like what's already implemented. I'd argue making the shroud way smaller and/or non-existent would limit ECM's strength more than that fix.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users