Balance Discussion - Aug 2018 - Post Podcast Feedback
#61
Posted 07 August 2018 - 04:22 PM
-LRMs are somewhat overpowered, on this we agree, but I HATE them because they change a game of maneuver warfare into awkward standoffs and spending a lot of time not doing what you want to do, which is shoot mechs... and for even a premade team to have a built in hard counter to LRM a full lance is going to have to take mechs that are sub optimal at best and probably not what they wanted to build or play in the first place... Because of the LRM buffs I now only take mechs into solo queue that either have ecm, stealth armor, or enough mobility to get between points of cover, which excludes most of the mechs I enjoy. I've had to tinker to an uncomfortable degree and build mechs I never wanted to build now to deal with solo queue because of this one weapon system and how solo queue behaves.
-Clan Boogeymen... what about players who stand still while getting shot? what about the massive time to dissipate heat? what about how they become almost useless in a brawl? what about how slow they have to go?... and btw when so called overpowered clan mechs get nerfed into the ground with the rest of them thats the day i demand a refund and removal of all my clan mechs because THATS ALL CLANS HAVE LEFT... I have long been a clanner but that is quickly changing as i get more usefulness out of IS mechs' agility, cooldown, armor, quirks, duration, heat and so on... i look at the catalog of clan mechs i have in pure frustration now, and despite all the clan boogeymen out there, in most of my IS mechs I'm just not even worried about it, I know what to do.
-Skilltree issues... there are mechs that I just dont even use any more because it would require far more than 91 skill points to make it useful or even be expected to survive contact. Direwolves and Kodiaks are now giant "kick-me" signs, countering LRMs with skill nodes detracts from the purpose and role of MOST MECH BUILDS... Additionally I would like to see unskilled mechs to actually be usable because... well we gotta use em to skill em and a large number of skillpoints being practically mandatory just adds frustration to lack of flexibility.
-I strongly, STRONGLY suspect that what pgi wants is not balance but one of two things: either constantly shifting the meta to promote the purchase of different mechs, or catering to lower skilled players at the expense of true balance itself and the competative community overall.... "cheapskates" and "beer league" and all that.
That is all...
#62
Posted 07 August 2018 - 04:22 PM
kjubert, on 07 August 2018 - 04:17 PM, said:
U know, populists use this behaivor too. Why even bother making this thread if u go your way anyway?
It feels like it doesnt matter if several smart people disagree. I am dissapointed
Let me touch on this. From the original post in this thread:
Quote
Balance is a highly sensitive subject and we understand that. What is being redacted/removed are non-constructive or direct attacks as per the Code of Conduct. There is negative feedback where which we are engaging discussion and it's what we intend to do moving forward.
#63
Posted 07 August 2018 - 04:24 PM
Also engine desync is a terrible and nonsensical change, roll that back and game instantly handles better. IF you want to buff/nerf mech mobility do it on a case by case basis, you can have a % based change which diminishes the larger engine you go so that a mech can't twist faster than intended with larger engines.
LRMs need flatter trajectory. That way they can't loop over cover and that way maps like polar highlands which are terribly designed can become playable. Maybe can skip this if planned LRM/lock changes achieve decent enough improvement.
Bring back cSPL maybe not to OG stats although they were hardly OP considering the range. 4 damage is just laughable and wasted tonnage at this point when combined with 165 effective range.
Probably wouldn't even have to nerf clan lasers. Maybe just slightly buff IS laser ranges and/or adjust IS vs clan ghost heat limits. So allow IS to fire 8 medium lasers for 40 alpha vs the 6 clan ermlas for 42. Can either bring IS up or clan down doesn't matter to me. But this is a pretty damn easy thing to implement and could help alot.
Some might not agree with me but this creates a balanced and competitive game with the widest range of effective builds the game has ever seen. TTK may go down very slightly but this games TTK is insanely long rn anyway.
Edited by POOTYTANGASAUR, 07 August 2018 - 04:25 PM.
#64
Posted 07 August 2018 - 04:27 PM
process, on 29 June 2018 - 06:34 AM, said:
The proposed math, not including time between projectiles:
AC20: 20 damage, 4s cooldown = 5 DPS
UAC20, best case: 40 damage, 4s cooldown = 10 DPS
UAC20, worst case: 20 damage, 5s cooldown = 4 DPS
Net UAC20 performance = 0.25*best case + 0.75*worst case = 5.5 DPS
Compared with the current mechanics (15% jam chance, 7.5s penalty):
UAC20, best case: 40 damage, 4s cooldown = 10 DPS
UAC20, worst case: 20 damage, 11.5s cooldown = 1.74 DPS
Net UAC20 performance = 0.85*best case + 0.15*worst case = 8.75 DPS
Also buff IS AC20 projectile velocity!
Edited by process, 07 August 2018 - 04:27 PM.
#65
Posted 07 August 2018 - 04:29 PM
it was civil feedback
all i see so far is. u do your own things. mainly listen to the brown sea, ignore long time players (especial those who actually understand whats going on) and make the majority of players leave because they get tired. So whats left? a somehow dying game with many people leaving.
#66
Posted 07 August 2018 - 04:31 PM
INVIICTUS, on 07 August 2018 - 04:22 PM, said:
This is why Chris is making changes in the August patch. Also still open for discussion.
Quote
We are not closing the gap between Clan and IS to zero and it's never going to be our intent to do so. Clans will always keep their range/tonnage/space advantages and a slightly higher damage output. We're just trying to narrow that gap slightly so we can keep a better eye on things like agility/mobility and quirks in general.
Quote
This is why we're going back and looking at the baseline tuning of these chassis to give them back some of what they've lost.
Quote
This is far from the truth.
#67
Posted 07 August 2018 - 04:31 PM
There has Been so much good input over the past two years ive been here and its just automatically disregarded. It seems you're focusing on the "beer league" players not only for WC, but for the furure of the game. We were in such a good damn place before desync and the trees, if only the effort put in had went towards qol and new player experience, coupled with influx from battletech you would be in such a good place.
Look at the info provided by people like tarogato and navid and denn and ash and these people like this. They *know* the game. They know what decent plauers want. They actually play it it a lot. For a game that used the tagline "highest skill in a-sports", weve gone really far in the wrong direction.
We keep going this way and you'll lose even more people, definitely me.. Makes me sad since this has Been my only game and hobby for the past 2.5 years and I really enjoyed.
#68
Posted 07 August 2018 - 04:33 PM
i reccomend buffing torso agility for all mechs with a humanoid shape. commando, firestarter, wolfhound, warhammer, atlas, orion, banshee, all those "shaped" mechs. give less torso speed buffs to mechs who primarily use dakka, such as the kdk-3, the annihilator, the blackwidow, maulers, the victor that has 3 ballistic harpoints... mechs like that, as dakka is powerful, but doesnt need the twist speed to be "good" since its a facetanking setup anyways and its strength comes from "staring down" other mechs.. those other mechs that require a "shoot and move" type of playstle, which is big acs, srms, mrms, laservomit, gauss, ppcs.... those types of mechs need the torso twist more, as they need to get of their shot, then move to safety while they reload their weapons.
doing torso mobility buffs based off this sytstem would benefit your goal of increasing time to kill. you can then use acceleration and develeration buffs to rein in outliers as you currently seem to be doing, without hurting some of the torso twisting, or what i consider to be the "fun" of a lot of mechs.
#69
Posted 07 August 2018 - 04:35 PM
Paul Inouye, on 07 August 2018 - 04:22 PM, said:
Let me touch on this. From the original post in this thread:
Balance is a highly sensitive subject and we understand that. What is being redacted/removed are non-constructive or direct attacks as per the Code of Conduct. There is negative feedback where which we are engaging discussion and it's what we intend to do moving forward.
Actually his whole post was gone, which had valid criticism. It only came back redacted after my post had been made
#70
Posted 07 August 2018 - 04:36 PM
kjubert, on 07 August 2018 - 04:29 PM, said:
I don't know what happened in your previous thread.. but this thread is a discussion thread. If things are constructive, they stick around and more importantly become part of the discussion. I get it that there are angry people out there. I just ask that people keep it civil and let's talk it through.
#71
Posted 07 August 2018 - 04:41 PM
#72
Posted 07 August 2018 - 04:43 PM
Whatever you have been doing between now and the last set of Dane / Tarogato / NavidA1 response podcasts back in Feb isn't working. There were a few things in those podcast that never even happened, like JJ buffs and gauss-peep de-nerfs just to name a couple.
Lets try going a different direction sooner rather than later. I don't really think PTS experiments are the solution because most of what I have seen go to PTS over the years has died (infotek, energy draw, etc).
As far as Clan lasers go, lets drop HLL to 16 or 17d and ERML to 6.5d. That shouldn't need a PTS and it also doesn't de-fang Clan lasers. The 90 whatever alpha isn't a mech I play but I do play the Hellbringer and I think that is a fair nerf.
Edited by Kin3ticX, 07 August 2018 - 04:45 PM.
#73
Posted 07 August 2018 - 04:45 PM
Panthros, on 07 August 2018 - 04:41 PM, said:
This won't really work out due to buckets and stuff. I don't know many people who think it needs to go that direction either. It can work without this. Source: imm3diately prior to pre skill tree/desync
#74
Posted 07 August 2018 - 04:47 PM
Panthros, on 07 August 2018 - 04:41 PM, said:
Moving the timeline to add the more advanced 'Mech technology will elilmitate a ton of 'Mechs currently in the game no matter how much buffing/quirking you do.
#75
Posted 07 August 2018 - 04:49 PM
Also engine desync is a terrible and nonsensical change, roll that back and game instantly handles better. IF you want to buff/nerf mech mobility do it on a case by case basis, you can have a % based change which diminishes the larger engine you go so that a mech can't twist faster than intended with larger engines.
LRMs need flatter trajectory. That way they can't loop over cover and that way maps like polar highlands which are terribly designed can become playable. Maybe can skip this if planned LRM/lock changes achieve decent enough improvement.
Bring back cSPL maybe not to OG stats although they were hardly OP considering the range. 4 damage is just laughable and wasted tonnage at this point when combined with 165 effective range.
Probably wouldn't even have to nerf clan lasers. Maybe just slightly buff IS laser ranges and/or adjust IS vs clan ghost heat limits. So allow IS to fire 8 medium lasers for 40 alpha vs the 6 clan ermlas for 42. Can either bring IS up or clan down doesn't matter to me. But this is a pretty damn easy thing to implement and could help alot.
Some might not agree with me but this creates a balanced and competitive game with the widest range of effective builds the game has ever seen. TTK may go down very slightly but this games TTK is insanely long rn anyway.
#77
Posted 07 August 2018 - 04:55 PM
Paul Inouye, on 07 August 2018 - 04:31 PM, said:
We are not closing the gap between Clan and IS to zero and it's never going to be our intent to do so. Clans will always keep their range/tonnage/space advantages and a slightly higher damage output. We're just trying to narrow that gap slightly so we can keep a better eye on things like agility/mobility and quirks in general.
What I'm saying is that even with clans having better space, heat syncs, weapon and ammo weight, range and damage... IS mechs often still perform better despite a perceived "inferiority"... play a 6 uac2 night gyr for a good amount of matches and then play a triple uac2 or triple rac2 dragon... half the guns, 15 tons less of mech, less armor, no JJs, but still usually twice the utility.
Edited by INVIICTUS, 07 August 2018 - 05:00 PM.
#78
Posted 07 August 2018 - 04:59 PM
IN GENERAL, engine desync (even with zero skill) provided buffs to these aspects of mech mobility (compared to old pre-desync era WITH mobility quirks + good engines):
- Acceleration
- mech turn rate
those two stats define how well a mech can maneuver in close quarters. It should be noted though that around 30% of chassis got nerf'd by engine desync in those two stats.. and not the deserving ones even.
On the other hand engine desync OBLITERATED the following:
- Torso twist speed
Torso twist speed is the key stat that defines how well a mech can spread incoming damage, and as most people who play the game know, engine desync did not do well in this regard.
There are mechs that have been hit in all three fronts though... like the atlas (that not only is nerf'd in all mobility aspects... but HAS THE TOP SPOT in having the greatest magnitude of nerfs in ALL three fronts)
I've been gathering and compiling data on mech mobility along with Tarogato for some time now... you'll see the results in a video soon.
Edited by Navid A1, 07 August 2018 - 11:59 PM.
#79
Posted 07 August 2018 - 05:02 PM
So, if our criticisms seem harsh, please keep in mind we only are looking out for the future of the game, and are just frustrated.
#80
Posted 07 August 2018 - 05:02 PM
as for LRMs, i feel that you're taking the wrong approach with your changes by nerfing active spotting and artemis. what i would like to see is something like this:
significantly reduce tracking strength and increase missile spread on targets that are not either NARC'd or being spotted with TAG. in return, their spread could be tightened to be even better than it is now (assuming target is being affected by NARC or TAG), or perhaps their velocity could be increased even further. the idea is to promote active gameplay, giving LRMs more of a risk vs. reward dynamic. if you're getting your own locks or working together with a spotter, then you reap greater rewards than you would currently. if you're just sitting in the back clicking on the red squares then you're only going to offer moderate suppression at best.
2 user(s) are reading this topic
0 members, 2 guests, 0 anonymous users