- If target locks allowed you to override convergence to your reticle placement (so weapons would converge at the range of your target rather than reticle), this would allow mechs with worse hardpoint placement to better combat fast moving targets or targets like poptarts, really any time you have to lead a target, hardpoints matter a lot because it impacts where your shots converge with poptarts probably gaining the biggest benefits just due to terrain in the background often being further than ground pounders.
- In soup queue, it allows you to share more enemy placements easier without having to call anything. Target info sharing is a very beneficial thing in soup queue given the lack of comms in these games so ECM being a deterrent for that just makes ECM a lot more powerful than it should be IMO.
- I think the big thing that would likely have to change though is ECM/BAP mechanics. My thought is that BAP should probably be limited much like ECM and offer something worth limiting like that. The thing that comes to mind is non-LOS radar similar to seismic, but with actual locks rather than the little radar pings you get with seismic currently. Range could stay short-ish, it doesn't need to be infinite like visual radar. ECM on the other hand protects you from detection of the BAP mech's wall hacks.


Infinite Radar Range?
#1
Posted 15 October 2023 - 07:36 AM
#2
Posted 15 October 2023 - 08:01 AM
#4
Posted 15 October 2023 - 11:14 AM
#5
Posted 15 October 2023 - 12:35 PM
KursedVixen, on 15 October 2023 - 11:14 AM, said:
Not really relevant to the discussion. Gauss Rifles and autocannons wouldn't be so short range nor would they probably fire that slow of projectiles so yes, no one really cares about how things work in real life unless it adds something interesting mechanically to the game.
Edited by Quicksilver Kalasa, 15 October 2023 - 12:36 PM.
#6
Posted 15 October 2023 - 02:23 PM
#7
Posted 15 October 2023 - 05:46 PM
ThreeStooges, on 15 October 2023 - 02:23 PM, said:
I wouldn't bother with the TC, it's boost to sensor range is pretty minimal. With just a BAP and skills the 3L has 1600m of radar range, which lets you essentially see everything worth seeing as long as it isn't under ECM. It even gives you 400m of sensors even under stealth, which is pretty nice to have as well.
#8
Posted 16 October 2023 - 09:12 AM
Quicksilver Kalasa, on 15 October 2023 - 07:36 AM, said:
- If target locks allowed you to override convergence to your reticle placement (so weapons would converge at the range of your target rather than reticle), this would allow mechs with worse hardpoint placement to better combat fast moving targets or targets like poptarts, really any time you have to lead a target, hardpoints matter a lot because it impacts where your shots converge with poptarts probably gaining the biggest benefits just due to terrain in the background often being further than ground pounders.
- In soup queue, it allows you to share more enemy placements easier without having to call anything. Target info sharing is a very beneficial thing in soup queue given the lack of comms in these games so ECM being a deterrent for that just makes ECM a lot more powerful than it should be IMO.
- I think the big thing that would likely have to change though is ECM/BAP mechanics. My thought is that BAP should probably be limited much like ECM and offer something worth limiting like that. The thing that comes to mind is non-LOS radar similar to seismic, but with actual locks rather than the little radar pings you get with seismic currently. Range could stay short-ish, it doesn't need to be infinite like visual radar. ECM on the other hand protects you from detection of the BAP mech's wall hacks.
Hmm. Im of two minds here.
On one hand, i completely agree that in the current game, ECM is excessively powerful, almost to the point of making any chassis that doesnt have it pointless, at least in QP / Pug games.
However, i think our sensors are way too strong in the first place, and ECM brings it to where i would like it to be for everyone by default. Having a sodding giant red dorito highlight you as soon as you're within LOS makes being sneaky effectively impossible. In my opinion, there does need to be some way to stop yourself being highlighted like that and, as it stands, the only functional option is ECM. Moving it to stealth armour only isn't OK, because thats IS only.
#9
Posted 16 October 2023 - 10:26 AM
battletech universe is like buying an old radio from an antiques store. something that may have been state of the art in its heyday but is now broken and malfunctioning due to decades of neglect and disrepair. so you end up with this fuzzy hissing noise box if it works at all. but you use it because its the only thing you can get.
Edited by LordNothing, 16 October 2023 - 10:34 AM.
#10
Posted 16 October 2023 - 10:58 AM
Widowmaker1981, on 16 October 2023 - 09:12 AM, said:
I'm of the mind that the counter play that relies on that sort of thing is not well regarded these days, partially why modern shooters avoid those kinds of scenarios. Counterstrike tries to ensure parts of the map are well lit so you don't feel cheated by pixel peeking or corners that you can hold that make it hard to see, but they also try to use lighting to avoid shadows giving away your approach as well (Inferno in CS2 got changed a bit specifically so that shadows didn't give things away). The Rose skin that made you blend in really well with the shadows and was extremely contentious in CoD is another example. So while yes, it should be possible to catch people by surprised, doing it by blending in too well either your surroundings is typically unfun for most players in PvP games. I actually like the glow effect that Overwatch had at one point (not sure if they still have it though) around enemies as well but maybe that's overly ambitious to ask to be added.
To your point though, in CS flanks are possible because of the lack of info but TBH given that we have LOS radar not really sure what is honestly that different between the two games other than the fact the number of players per side in this game is much much higher (which matters for getting info on enemy positioning). There is also a lack of utility to help combat that either (smokes, flashbangs, or just flat out area denial like molotovs). However I think providing "scouts" with a better seismic sensor effectively with BAP (and denying other mechs seismic) along with offering ECM the potential to counter that also opens up a better opportunity to scouts without JJs. Scouts without JJs are typically held back because of how important JJs are for maneuvering around maps to get angles and also jump scout. Offering them a way to scout without JJs and putting themselves in an overly risky position is pretty important.
Widowmaker1981, on 16 October 2023 - 09:12 AM, said:
I mean, lore-wise it is

Edited by Quicksilver Kalasa, 16 October 2023 - 11:36 AM.
#11
Posted 16 October 2023 - 11:54 AM
Quicksilver Kalasa, on 16 October 2023 - 10:58 AM, said:
To your point though, in CS flanks are possible because of the lack of info but TBH given that we have LOS radar not really sure what is honestly that different between the two games other than the fact the number of players per side in this game is much much higher (which matters for getting info on enemy positioning). There is also a lack of utility to help combat that either (smokes, flashbangs, or just flat out area denial like molotovs). However I think providing "scouts" with a better seismic sensor effectively with BAP (and denying other mechs seismic) along with offering ECM the potential to counter that also opens up a better opportunity to scouts without JJs. Scouts without JJs are typically held back because of how important JJs are for maneuvering around maps to get angles and also jump scout. Offering them a way to scout without JJs and putting themselves in an overly risky position is pretty important.
Edit: Ive never played CS, so correct me if im wrong, but i was under the impression that TTK is vastly lower than MWO (even currently), and there is no limitation on reaction time. That should change things a lot when considering how sneaky its ok for enemies to be. No one likes to die instantly without seeing their enemy, but thats a lot less likely to happen in MWO, and there is a need for smaller mechs to be able to deal damage without taking return fire in MWO that doesnt have a counterpart issue in CS.
I think there is a middle ground between it being hard to spot enemies even when you're looking right at them, and a red targeting triangle that makes it impossible not to notice them even in peripheral vision (and that being shared with their entire team immediately as well) Im fine with making mechs stand out more, visually.
Quicksilver Kalasa, on 16 October 2023 - 10:58 AM, said:

Eh, i guess ill admit that biased in that i prefer the clan mech feel (faster, more agile, squishier) and wouldnt like to have to play IS to avoid the dorito.
Edited by Widowmaker1981, 16 October 2023 - 12:01 PM.
#12
Posted 16 October 2023 - 12:39 PM
Widowmaker1981, on 16 October 2023 - 11:54 AM, said:
So TTK is lower definitely, but there are som common themes, such as Info gathering is important to both games because you simply don't have wall hack radar and also don't want to get punished for a bad peek. You can survive bad peeks more so in MWO than you can in CS:GO but they still value cover a lot and trades are still very much a thing, MWO's limited mobility also means though that bad pokes are harder to recover from which is worth mentioning (a bad poke is slower than in CS for example). This game is just a bit different in that ranged trades tend to be A lot less dangerous but at the same time, the maps are also a lot less claustrophobic so you have more room to approach the enemy safely (CS has three lanes effectively without an ability to rotate between the two outside of dedicated connectors), but that hasn't always been true (Gauss/PPCs come to mind).
Widowmaker1981, on 16 October 2023 - 11:54 AM, said:
At that point I don't really see necessarily the problem with providing a doritio because its rare that something like that is missed and not comm'd in a normal situation.
The info share portion comes into play more in QP currently than it does in comp because the dorito isn't enough info by itself (I mean it helps but meh) but that is also because radar range is limited (and ECM also impacts that) so that eyeballs are more important. For example thermal on any of the hot maps is waaaaaay more useful than any sort of radar currently.
FWIW I prefer playing Clan as well, but I also don't find it problematic that each tech base does not have the same things to play around with. I'm totally okay with stealth being locked in to the IS. It would be interesting to see Clans get some sort of equipment as well to have for a select few mechs outside of just having the better tech all around though.
#13
Posted 16 October 2023 - 01:45 PM
So infinite sensor range would break that.
#15
Posted 17 October 2023 - 01:34 AM
Target lock would enhance convergence to the range of your locked target.
So skill nodes like target retention that keeps an enemy radar and target locked within a small radius (brawl range) I think the nodes give that quirk out to 240m tops.
When you lose target lock, convergence switches back to the crosshair and the distance of what you are pointing at gets set to.
Now on one side, you may initially think OMG that would just instagib lights and mediums..
Keep this in mind.
Lasers still have burn times, all ballistics and missiles have travel time, the only weapons that travel fast enough to disregard a change like this, is Gauss, ERPPC's, or overquirked AC's.
You still have to aim, and adjust for leading targets.
Something like this would probably be extremely valuable for weapons that you have to lead your shots on, like slower velocity weapons, generally speaking, heavier autocannons, MRM, SRM, standard PPC's,
While making use of the Retention nodes and anti-radar dep nodes valuable, as well as BAP and targeting computers.
The only obstacle, which will be a guaranteed knee jerk "WHOA! STOP" is having to face the truth that ECM needs to be overhauled.. Seriously.
Why? Because so many abuse it, because it is quite literally a crutch that instantly increases survivability, because of information denial to other mechs, or even dedicated scouts.
Until ECM gets properly balanced, and players that rely on it have the balls to admit it is exactly that.. a crutch.
Nothing is going to happen to move intelligent ideas like the OP has, forward until that truth is faced and dealt with.
And the second half, is reworking LRM and lock-on weapons tracking mechanic in line with the change, but the argument is about 90% about ECM balance, and about 10% of all of the hate against lock-on weapons.
#16
Posted 19 October 2023 - 02:14 PM
#17
Posted 21 October 2023 - 06:38 PM
By my analysis, stealth tech has continued to evolve, so it's hard get a lock on a mech past ~800M, "Mark One Eye Ball" for the win.
ETA: or did you think it was a coincidence that radar absorbent https://www.sarna.net/wiki/Myomer was chosen for propulsion?
ETA2: also weird that it was written about "before" it was officially "discovered" by scientists... #tinfoilhat
Edited by kalashnikity, 21 October 2023 - 06:43 PM.
2 user(s) are reading this topic
0 members, 2 guests, 0 anonymous users