


Jan 2024 Patch Leaks And Rumors
#221
Posted 09 January 2024 - 11:29 AM

#223
Posted 09 January 2024 - 11:57 AM
Shineplasma, on 09 January 2024 - 10:49 AM, said:
There are comp divisions for players of all skill and commitment levels. Yes, from T5 to the tippity top. Folks on the forums here just don't know or interact with them for the most part. There are over 1200 members on the MWOComp discord.
It doesn't matter how many people are on the Discord channel, it matters how many people have a seat at the table, the actual Cauldron members who vote. I was aware there are non-competitive members, but in effect those members are outvoted.
And I do think the roster of said group skews toward the "tippity top" as you phrase it, rather than the T5 end. That doesn't help their image much, but I'm not concerned about that as I _DO_ want players who understand the game to give the feedback, I'm concerned about the patch content. And the effect of the patch content is most often toward a "reward the higher skill" mindset, as we've heard already in this thread.
#224
Posted 09 January 2024 - 12:13 PM
ScrapIron Prime, on 09 January 2024 - 11:57 AM, said:
I mean from the outward responses from Cauldron members, even some of the competitive players understand that QP experiences should matter when talking about balance for just the overall health of the game. Why else would they bother trying to adjust LRM's consistency in the first place if that weren't true? Why not just nerf them straight into the ground by giving them the velocity they started this game out with, and reducing the NARC duration to something goofy like 1s? If they were the conspiratorial group that some make them out to be, Gauss/PPC would have been unlinked in ghost heat already

#225
Posted 09 January 2024 - 12:33 PM
Quicksilver Aberration, on 09 January 2024 - 12:13 PM, said:

Fair point, yes. That's the reason I'm making some noise on this, as I want to make sure they hear the concern. If I didn't think they'd listen I'd have shut up ages ago.

#226
Posted 09 January 2024 - 01:19 PM
- LRMs are strong against bad players and weak against good players
- LRMs are easy to use and hit their skill ceiling pretty fast. Afterwards, they only get better with improved positioning, which is a separate skill
- LRMs are very feast or famine depending on the makeup of a match (how much ECM, AMS, etc.)
- Due to coding limitations, the Cauldron has very few options for adjusting LRMs. They can't make any changes to the core lockon system.
- They can't take any risks with LRM power because it could easily kill the game if another lurmpocalypse happens. So they are erring on the side of caution.
- At the moment, LRMs are useable and strong in lower tiers, but not in higher (unless used in a coordinated group, like with NARC). This is about the best they can be balanced without overhauls to the LRM system. But they are too inconsistent.
- So, the best that can be done right now is to try to make them a bit more consistent by nerfing ECM and Radar Dep. But if that is done without taking something away from LRMs themselves, then they'll get too strong.
- So LRMs have to lose something. And that something is velocity.
Edited by Heavy Money, 09 January 2024 - 01:35 PM.
#227
Posted 09 January 2024 - 01:22 PM
ScrapIron Prime, on 09 January 2024 - 11:57 AM, said:
It doesn't matter how many people are on the Discord channel, it matters how many people have a seat at the table, the actual Cauldron members who vote. I was aware there are non-competitive members, but in effect those members are outvoted.
And I do think the roster of said group skews toward the "tippity top" as you phrase it, rather than the T5 end. That doesn't help their image much, but I'm not concerned about that as I _DO_ want players who understand the game to give the feedback, I'm concerned about the patch content. And the effect of the patch content is most often toward a "reward the higher skill" mindset, as we've heard already in this thread.
Are they outvoted, or is the patch content you're referring to *already* directed towards making the game better for players of all skill levels?
You seem to assume a lot about a group of people you know very little about.
Edited by Shineplasma, 09 January 2024 - 01:28 PM.
#229
Posted 09 January 2024 - 02:03 PM
ScrapIron Prime, on 09 January 2024 - 01:49 PM, said:
He was giving you the benefit of the doubt because what you're saying about them is so obviously wrong.
#230
Posted 09 January 2024 - 02:48 PM
Frost_Byte, on 07 January 2024 - 09:12 PM, said:
HAGs roll on the cluster munitions table. So then, by your own logic, spread over splash it is. Excellent, thanks!
#231
Posted 09 January 2024 - 02:57 PM
MechMaster059, on 08 January 2024 - 04:51 AM, said:
I thought it was disgusting watching Sneaky Snek kill 3 players in a row on Hellbore Outpost by backstabbing them with a Scaleshot before the recent missile spread quirk nerfs.
https://youtu.be/JSm...7G2FmCJlDN&t=15
I thought it was disgusting watching Sneaky Snek slowly waddle his Gausszilla into a firing position on Polar Highlands and rack up 2,100+ damage spamming HAG30s. Standing in one spot and blasting stuff at range with HAGs doesn't take much skill either.
https://youtu.be/5oD...3GmvbiN_M&t=563
Where is this fight taking place? In an open plain? Because I can tell you it would be very difficult on Tourmaline Desert to turn 180 degrees without facing into huge spires that block anything from getting past them. Free Worlds Colosseum had gigantic mushrooms installed that block LRMs from nearly every angle, and on and on...
The circumstance you describe seems to be a best case scenario type situation.
Zam
#232
Posted 09 January 2024 - 02:59 PM
That can easily be countered by limiting the ammo via reducing the ammo per ton...
It will have next to none affect to people who see the LRM as complementary weapon for certain mech setup but greatly reduce the ability to boat the weapon as LRM boats would run out of ammo quite qickly if you i.e. reduce the ammo per ton by 50% something.
Making LRM even slower as they alredy are - well, you can put a lame in a Segway up for racing LRMs and he would win!
Stop making up excuses for keeping LRM useless!
#233
Posted 09 January 2024 - 03:12 PM
Thorqemada, on 09 January 2024 - 02:59 PM, said:
It will have next to none affect to people who see the LRM as complementary weapon for certain mech setup but greatly reduce the ability to boat the weapon as LRM boats would run out of ammo quite qickly if you i.e. reduce the ammo per ton by 50% something.
Not really, reducing the ammo per ton hurts all builds because it just means you need more tonnage to run LRMs whether you are boating them or not, and typically mixed builds need the tonnage just as much as any other build.
Ammo per ton is just a build tax that increases the tonnage tax to use a weapon.
Edited by Quicksilver Aberration, 09 January 2024 - 03:13 PM.
#234
Posted 09 January 2024 - 03:49 PM
Heavy Money, on 09 January 2024 - 01:19 PM, said:
- LRMs are strong against bad players and weak against good players
- LRMs are easy to use and hit their skill ceiling pretty fast. Afterwards, they only get better with improved positioning, which is a separate skill
- LRMs are very feast or famine depending on the makeup of a match (how much ECM, AMS, etc.)
- Due to coding limitations, the Cauldron has very few options for adjusting LRMs. They can't make any changes to the core lockon system.
- They can't take any risks with LRM power because it could easily kill the game if another lurmpocalypse happens. So they are erring on the side of caution.
- At the moment, LRMs are useable and strong in lower tiers, but not in higher (unless used in a coordinated group, like with NARC). This is about the best they can be balanced without overhauls to the LRM system. But they are too inconsistent.
- So, the best that can be done right now is to try to make them a bit more consistent by nerfing ECM and Radar Dep. But if that is done without taking something away from LRMs themselves, then they'll get too strong.
- So LRMs have to lose something. And that something is velocity.
Overlooking other points, which are debatable, how lurms are going to be more consistent and usable if velocity and attack angle get nerfed?
Even if they nerf raderp (the very major point to start with) to make it perform at the typical levels before the skill tree change, when raderp was not perceived so damn OP. LRMs have been already velocity nerfed to the ground, taking 4X to 7X the time to hit, once launched, than any other direct-fire-long-range weapon. This is granting several seconds to reach cover (loads of them in every map) and skip the volley.
But they want another velocity nerf? And a flatter trajectory? Seriously? Even if raderp bonus were 0%, how can anyone, genuinely, believe that any of that is going to make lurms more usable and will hit targets more consistently?
Shineplasma, on 09 January 2024 - 01:22 PM, said:
You seem to assume a lot about a group of people you know very little about.
I guess that we'll see in a few weeks if they make the game "better" or if we just get a quite predictable major LRM nerf that anyone with a brain can expect, on the basis of prior experiences.
For example, their 1% per node raderp nerf, allegedly to make missiles more viable (lol). For example, maps reworks, like Polar or planting shitloads of metal mushrooms in Steiner Colloseum. For example, the lastest map is a cave. For example, the lastest velocity nerf because ... reasons. For example, the recent ECM range reduction that changed actually nothing for the usability of guided missiles.
Edited by Tarteso, 09 January 2024 - 03:50 PM.
#235
Posted 09 January 2024 - 04:11 PM
Tarteso, on 09 January 2024 - 03:49 PM, said:
Overlooking other points, which are debatable, how lurms are going to be more consistent and usable if velocity and attack angle get nerfed?
Even if they nerf raderp (the very major point to start with) to make it perform at the typical levels before the skill tree change, when raderp was not perceived so damn OP. LRMs have been already velocity nerfed to the ground, taking 4X to 7X the time to hit, once launched, than any other direct-fire-long-range weapon. This is granting several seconds to reach cover (loads of them in every map) and skip the volley.
But they want another velocity nerf? And a flatter trajectory? Seriously? Even if raderp bonus were 0%, how can anyone, genuinely, believe that any of that is going to make lurms more usable and will hit targets more consistently?
They will be more consistent because locks will last longer, and because 100% radar dep mechs won't instantly break locks. Losing some velo is a fair price to pay for that.
LRMs miss when 1) People get cover physically blocking them or 2) The lock breaks and they target is moving. Making radar dep weaker will make locks last longer, which means #2 has less impact. Reducing velo compensates for this by giving them more time to break the lock or get cover. So that's a zero sum change on its own (depending on how much dep and velo are changed, of course.)
Currently, it is very frustrating to be playing against fast mechs with 100% dep and lose your lock because they ran behind a pebble for a fraction of a second. 100% dep causes that. Even if its changed to just 90%, then it will cause split second blips to not break locks. That's a huge deal and worth losing some velo for.
Less velo makes almost no difference when shooting at exposed targets, especially if you have your own LOS to it. They may even only reduce velo on indirect fire, which would make this an indirect buff to direct fire (which nobody is complaining about. All the issues with LRMs being oppressive come from the indirect fire.)
So overall, having 100% dep not be possible makes a huge difference to quality of life and ease of use, especially against small fast moving targets. Losing some velo makes almost no difference against optimal targets like slow assaults in the open, and should balance out in other situations.
Of course, all of this depends on what the changed amounts of each actually are.
Edited by Heavy Money, 09 January 2024 - 04:11 PM.
#236
Posted 09 January 2024 - 05:34 PM
Tarteso, on 09 January 2024 - 03:49 PM, said:
For example, their 1% per node raderp nerf, allegedly to make missiles more viable (lol). For example, maps reworks, like Polar or planting shitloads of metal mushrooms in Steiner Colloseum. For example, the lastest map is a cave. For example, the lastest velocity nerf because ... reasons. For example, the recent ECM range reduction that changed actually nothing for the usability of guided missiles.
1% dep nerf was only to cut back on losing locks because a mech crossed behind something like a pole or rock for a split second. Was never intended as a major balance change.
And cauldron has little to do with map design.
#237
Posted 09 January 2024 - 05:49 PM
dario03, on 09 January 2024 - 05:34 PM, said:
1% dep nerf was only to cut back on losing locks because a mech crossed behind something like a pole or rock for a split second. Was never intended as a major balance change.
And cauldron has little to do with map design.
95% raderp leaves 0.29 s + spotting time (if any) to lock in LOS if you equip 100% target decay, which is far less than you need just to start moving your reticle over the target. So, pointless change.
And, yeah! all the anti-lurming map design was coincidence.
Edited by Tarteso, 09 January 2024 - 05:51 PM.
#238
Posted 09 January 2024 - 05:59 PM
Heavy Money, on 09 January 2024 - 04:11 PM, said:
They will be more consistent because locks will last longer, and because 100% radar dep mechs won't instantly break locks. Losing some velo is a fair price to pay for that.
LRMs miss when 1) People get cover physically blocking them or 2) The lock breaks and they target is moving. Making radar dep weaker will make locks last longer, which means #2 has less impact. Reducing velo compensates for this by giving them more time to break the lock or get cover. So that's a zero sum change on its own (depending on how much dep and velo are changed, of course.)
Currently, it is very frustrating to be playing against fast mechs with 100% dep and lose your lock because they ran behind a pebble for a fraction of a second. 100% dep causes that. Even if its changed to just 90%, then it will cause split second blips to not break locks. That's a huge deal and worth losing some velo for.
Less velo makes almost no difference when shooting at exposed targets, especially if you have your own LOS to it. They may even only reduce velo on indirect fire, which would make this an indirect buff to direct fire (which nobody is complaining about. All the issues with LRMs being oppressive come from the indirect fire.)
So overall, having 100% dep not be possible makes a huge difference to quality of life and ease of use, especially against small fast moving targets. Losing some velo makes almost no difference against optimal targets like slow assaults in the open, and should balance out in other situations.
Of course, all of this depends on what the changed amounts of each actually are.
All the oppressive thing (lurmaggedons?) about indirect fire comes from group coordinated tactics. So, the question is: are they going to "balance" missiles around the idea of "3 LRM90 mechs assisted by a narc light in every match"?
#239
Posted 09 January 2024 - 06:06 PM
Tarteso, on 09 January 2024 - 05:49 PM, said:
95% raderp leaves 0.29 s + spotting time (if any) to lock in LOS if you equip 100% target decay, which is far less than you need just to start moving your reticle over the target. So, pointless change.
Its a huge change in that it means locks don't break if your aimer is on target and a light runs behind a lamp post. It is a change that fixes this one specific issue, which is a significant one.
Tarteso, on 09 January 2024 - 05:59 PM, said:
What happens if they don't and then people play that? Or just multiple LRM boats end up on the same team that aren't grouped? And if this becomes possible, then there's a huge incentive for people to play it. Which is exactly what happened the other times LRMs nearly destroyed the game.
Are you suggesting that we just ignore the possibility of multiple LRM boats being on the same team and trust to players to just be honorable and not play that? lol.
#240
Posted 09 January 2024 - 06:22 PM
Tarteso, on 09 January 2024 - 05:49 PM, said:
95% raderp leaves 0.29 s + spotting time (if any) to lock in LOS if you equip 100% target decay, which is far less than you need just to start moving your reticle over the target. So, pointless change.
And, yeah! all the anti-lurming map design was coincidence.
It wasn't to get lock, it was to keep lock. As in you have a lock, enemy runs in the open except for 0.05 seconds they pass by a rock or pole or something. With 100% dep you would lose lock, 95% you don't.
5 user(s) are reading this topic
0 members, 5 guests, 0 anonymous users