Jump to content

Pgi Is Nerfing Underperformers! (Pts3)


196 replies to this topic

Poll: Taking away existing quirks is a bad thing? (220 member(s) have cast votes)

Taking away existing quirks to balance an universal change is a bad thing?

  1. Yes (163 votes [74.09%])

    Percentage of vote: 74.09%

  2. No (48 votes [21.82%])

    Percentage of vote: 21.82%

  3. Other (Please post why) (9 votes [4.09%])

    Percentage of vote: 4.09%

Vote Guests cannot vote

#181 Koshirou

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 827 posts

Posted 13 March 2017 - 05:52 AM

View PostBud Crue, on 12 March 2017 - 07:28 AM, said:

Why skill out an objectively inferior mech that has no quirks to help it be competitve? Sure some of us have the XP to do it, so what the hell, but what is going to get played in the actual game? Just like now we will see a limited number of those mechs that are the best; but I think to an even greater degree, as the loss of quirks will make that "best" vs everything else distinction even more crystal clear for several chassis.

Precisely. For a player who knows what they are doing, most variants are useless.

But the reason I think it would be cleaner to get rid of them altogether is that the useless variants can still become traps for new players who might buy variants that are objectively worse than others. Now this could happen before, and some players who did so asked me for advice...
My response under the current system: "Eh, that's actually a subpar variant if you ask me... but hey, just play it to level up all the basics, which you need to do anyway if you want to master the better variants. It doesn't really cost you anything and maybe you'll even find you do well with it."
My response under the new system would have to be: "Eh, that's actually a subpar variant if you ask me... best get rid of it ASAP and write off the time and Cbs you spent on it so far as lost."

Under the new system, suboptimal variants are just traps for inexperienced players. I do not think that traps for inexperienced players are a good element of any game system.

#182 Monkey Lover

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 7,918 posts
  • LocationWazan

Posted 13 March 2017 - 02:42 PM

#1 thread on the test server forum but they're still talking about cbills.......

Im hoping the new post coming out here soon per Russ will have something.....

Edited by Monkey Lover, 13 March 2017 - 02:43 PM.


#183 Monkey Lover

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 7,918 posts
  • LocationWazan

Posted 13 March 2017 - 11:23 PM



Quote

Russ Bullock@russ_bullock 2h2 hours ago

@BishopSteiner I am fine disagreeing with the community on some aspects of balance as we have shown that is going to happen




Ok well, i got my answer right there.

Edited by Monkey Lover, 13 March 2017 - 11:25 PM.


#184 WolvesX

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Machete
  • The Machete
  • 2,072 posts

Posted 14 March 2017 - 05:51 AM

View PostMonkey Lover, on 13 March 2017 - 11:23 PM, said:







Ok well, i got my answer right there.

Wow! Does he really mean it?

#185 Athom83

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Death Wish
  • The Death Wish
  • 2,529 posts
  • LocationTFS Aurora, 1000km up.

Posted 14 March 2017 - 03:32 PM

I see a lot of "why did you vote no" post so here is why; I play IS only. I play the "under-performing" and underused mechs (Orion, Dragon, Wolfhound, etc) and I do fine in them.Yes, some of the quirks being taken away does suck. However, when voting I thought of the other upcoming changes like large modules getting a big HP buff while smaller ones get a HP nerf. This massively buffs IS mechs' survivability and nerfs most Clan builds at the same time. I was also thinking about the Time Jump where IS and Clan tech become less separated, meaning you don't need as many quirk bonuses to start having some "balance".

#186 Taxxian

    Member

  • PipPipPipPipPipPip
  • 227 posts
  • LocationLeipzig

Posted 14 March 2017 - 11:40 PM

While some IS "underperformers" are getting a slight nerf, many Clan and some IS overperformers (Locust, MadCat, Kodiak) get a huge nerf with the engine decoupling and crit changes, overall balance would have been better... would have been...

Edited by Taxxian, 14 March 2017 - 11:40 PM.


#187 Bud Crue

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 9,946 posts
  • LocationOn the farm in central Minnesota

Posted 15 March 2017 - 06:05 PM

View PostTaxxian, on 14 March 2017 - 11:40 PM, said:

While some IS "underperformers" are getting a slight nerf, many Clan and some IS overperformers (Locust, MadCat, Kodiak) get a huge nerf with the engine decoupling and crit changes, overall balance would have been better... would have been...


No. "Nerfing" the best mechs while also -with the same mechanism- nerfing the worst on a totally equal basis, and then piling on top of that "baseline" additional nerfs -however "slight"- does not make balance better. It makes the distinction between the best and the worst even more distinct.

#188 Widowmaker1981

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Widow Maker
  • The Widow Maker
  • 5,031 posts
  • LocationAt the other end of the pretty lights.

Posted 16 March 2017 - 04:55 AM

View PostBud Crue, on 15 March 2017 - 06:05 PM, said:


No. "Nerfing" the best mechs while also -with the same mechanism- nerfing the worst on a totally equal basis, and then piling on top of that "baseline" additional nerfs -however "slight"- does not make balance better. It makes the distinction between the best and the worst even more distinct.


What you dont seem to understand is that engine decoupling DOESNT affect all mechs on an equal basis. It allows PGI to balance mechs by agility attributes where before they couldnt do that (well, only by changing engine cap, which they dont want to mess with). In other words, the best clan mechs get rubbish base mobility despite having large engine caps, as a balancing feature.

Now, not all agility stats were correct, for example IS 100 tonners need to be more agile than Dires or Kodiaks. Also, IS mechs losing weapon quirks they needed to be balanced with their competition should be compensated in some way (either with more base agility, or with armour/structure).. but those are details not related to the concept and can be changed.

Edited by Widowmaker1981, 16 March 2017 - 04:55 AM.


#189 Athom83

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Death Wish
  • The Death Wish
  • 2,529 posts
  • LocationTFS Aurora, 1000km up.

Posted 16 March 2017 - 06:10 AM

View PostWidowmaker1981, on 16 March 2017 - 04:55 AM, said:

Also, IS mechs losing weapon quirks they needed to be balanced with their competition should be compensated in some way (either with more base agility, or with armour/structure).. but those are details not related to the concept and can be changed.

Perhaps by adding variants of base weaponry. The reason for the quirks was to somewhat simulate the different variants of the weapons. Take the Blackjack's main AC/2s. They are the Whirlwind-2 Autocannons with an longer range and faster RoF than say a Mydron or ZeusBolt AC/2. Or the AC/20s, some are 185mm short barrels, while others are 155mms with a longer barrel (balanced out to both have same tonnage and damage). Hence why some AC/20s can fire further than others, or why others still can fire faster or with less heat. Take a look at this page; http://www.sarna.net...eavy_Autocannon They are both AC/20s but vary quite massively in stats.

#190 Bud Crue

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 9,946 posts
  • LocationOn the farm in central Minnesota

Posted 16 March 2017 - 06:26 AM

View PostWidowmaker1981, on 16 March 2017 - 04:55 AM, said:


What you dont seem to understand is that engine decoupling DOESNT affect all mechs on an equal basis. It allows PGI to balance mechs by agility attributes where before they couldnt do that (well, only by changing engine cap, which they dont want to mess with). In other words, the best clan mechs get rubbish base mobility despite having large engine caps, as a balancing feature.


Kodak was hurt a bit...Night Gyr hurt...not so much. You are right decoupling didn't affect all mechs equally, it affected (at least in the PTS) all classes equally. To survive in many IS mechs you need to be able to twist and twist fast to spread damage. In the PTS some of the WORST mechs in the game are only viable because of that feature. Did ya test a Dragon? Nerfed by decoupling, How about a QuickDraw? A Victor? A Zeus? They certainly weren't all slowed tremendously and with skills you could get them back up close to live server performance, but that is the point...nerf the worst mechs in a system where all mechs have the same options to improve their "skills" and a mech light the Night Gyr is even MORE dominant, because it can spend freely on maxing out weapons and armor and jump jets, where the slowed mechs (yes even the poor Kodak) must now spend on 30 nodes just to get back what they lost in mobility. They are at a deficit now. They are at a worse deficit in the new system.

View PostWidowmaker1981, on 16 March 2017 - 04:55 AM, said:

Now, not all agility stats were correct, for example IS 100 tonners need to be more agile than Dires or Kodiaks. Also, IS mechs losing weapon quirks they needed to be balanced with their competition should be compensated in some way (either with more base agility, or with armour/structure).. but those are details not related to the concept and can be changed.


This too goes to my point. They aren't being compensated. They are being nerfed. In some cases doubly so (see above). If I believed PGI is actually going to be diligent and address the quirks or make other changes promptly to the most harmed mechs, then I would not be concerned about any of this. But I don't trust them and their history backs me up here. Once theses nerfs are made they will likely never correct or undue them. It isn't "just details" when you are taking the worst mechs in the game and making them objectively worse, it is totally trashing balance and diversity which some folks...namely PGI...assert is the opposite of the goal of the skills tree.

#191 Athom83

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Death Wish
  • The Death Wish
  • 2,529 posts
  • LocationTFS Aurora, 1000km up.

Posted 16 March 2017 - 09:37 AM

View PostBud Crue, on 16 March 2017 - 06:26 AM, said:

Kodak was hurt a bit...Night Gyr hurt...not so much. You are right decoupling didn't affect all mechs equally, it affected (at least in the PTS) all classes equally. To survive in many IS mechs you need to be able to twist and twist fast to spread damage. In the PTS some of the WORST mechs in the game are only viable because of that feature. Did ya test a Dragon? Nerfed by decoupling, How about a QuickDraw? A Victor? A Zeus? They certainly weren't all slowed tremendously and with skills you could get them back up close to live server performance, but that is the point...nerf the worst mechs in a system where all mechs have the same options to improve their "skills" and a mech light the Night Gyr is even MORE dominant, because it can spend freely on maxing out weapons and armor and jump jets, where the slowed mechs (yes even the poor Kodak) must now spend on 30 nodes just to get back what they lost in mobility. They are at a deficit now. They are at a worse deficit in the new system.

In what kind of world would the same motor be able to twist the 80 something tons of an Assault torso the same speed as the 10 tons of a Light torso? This is why coupling twist speed with engine rating doesn't make sense. The point was to reduce Assaults into the big, slow, and cumbersome walking fortresses they are. Because of how little mobility skills actually affect lower numbers, I would never spend those 30 nodes in mobility. I would spend those in armor and firepower to become an unkillable beast that can lay out stupidly tremendous amounts of firepower. This makes mechs (like the Night Gyr) pick some mobility nodes over firepower so they could spread out the damage as they cant tank like the assaults.

And in all honesty, how can you compare the 80 Victor to a 100 ton mech with a different tech base like the Kodiak for how much was removed? It only really lost the same amount as the Night Gyr and other 75 ton mechs (as its only a 5 ton difference), and can be brought up to the same levels. Same with compairing the Dragon and QuickDraw to the NightGyr, they're the bottom of the weight class and are more like the heavier medium mechs (Griffon, Shadow Hawk, etc). However, the current "meta" does favor Clan mechs, but it will shift with the upcoming Tech. You will probably being seeing a lot more of these mechs, especially if they get some 3060 variants coming. Like the strange but epic 8X and 8P variants of the QuickDraw which gives it ECM (and stealth armor for the 8X), moves the cockpit to the CT and frees the head for 2 energy mounts in a rotating turret. Or the Victor 10L with ECM + Stealth Armor, Gauss, and Streak 6s, or the 10D with a UAC20 and ERMLs, or possibly the 11D with a light engine, T-comp (C3 slave), ERLLs, and a RAC/5. Keep in mind those keep the jump jets.

View PostBud Crue, on 16 March 2017 - 06:26 AM, said:

This too goes to my point. They aren't being compensated. They are being nerfed. In some cases doubly so (see above). If I believed PGI is actually going to be diligent and address the quirks or make other changes promptly to the most harmed mechs, then I would not be concerned about any of this. But I don't trust them and their history backs me up here. Once theses nerfs are made they will likely never correct or undue them. It isn't "just details" when you are taking the worst mechs in the game and making them objectively worse, it is totally trashing balance and diversity which some folks...namely PGI...assert is the opposite of the goal of the skills tree.
Question, do you know what the "balance" or "meta" will be in the skill tree with the addition of the new tech? Also, they are being compensated as the HP of equipment is being changed based on size. This makes it so the average internal HP of Clan Assaults drops a little while quite massively increasing the internal HP of IS Assaults (and other IS mechs).

#192 Bud Crue

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 9,946 posts
  • LocationOn the farm in central Minnesota

Posted 16 March 2017 - 10:33 AM

View PostAthom83, on 16 March 2017 - 09:37 AM, said:

And in all honesty, how can you compare the 80 Victor to a 100 ton mech with a different tech base like the Kodiak for how much was removed?


How? Because that is what PGI did. They gave them the same characteristics. Yes they may change them and tweak them later on, but they didn't, nor do I trust that they will.

This is all irrelevant though to my points in this thread (excluding the post above). In all honestly I don't care about decoupling. I care about the nerfs presented in those 56 pages of nerfs. Those are absolute value reductions to the worst mechs in the game. That is the issue for me. Not decoupling, not retaining progress, not refunds, not even the gated skills of which my crap mechs in a non nerfed state will require more nodes than any meta mech in the game to be competitive. I care that they are nerfing...and seemingly randomly so...what WE KNOW, objectively know...to be some of the worst mechs, and making them worse. Those mechs are not going to magically be made good in the new system.

You asked about what I think is going to be meta in the new system? Same as the current system. New weapons might change that, but the skills tree changes nothing. Hunch IIc, Night Gyr, and yes even the poor, sad Kodak 3 are still going to be great (though as an exemplar I'm almost tempted to say Mad IIc). The skills tree is not going to make a newly nerfed St. Ives Blues the new medium of choice. Nor is the nerfed Orion going to suddenly become dominate over the Night Gyr. Etc.

Edited by Bud Crue, 16 March 2017 - 10:34 AM.


#193 soapyfrog

    Member

  • PipPipPipPipPipPip
  • Knight Errant
  • 409 posts

Posted 16 March 2017 - 10:46 AM

View PostAthom83, on 16 March 2017 - 09:37 AM, said:

Because of how little mobility skills actually affect lower numbers, I would never spend those 30 nodes in mobility. I would spend those in armor and firepower to become an unkillable beast that can lay out stupidly tremendous amounts of firepower.

You can easily get mobility, survival and firepower in the new tree.

However "stupidly tremendous" isn't really an option as the firepower buffs are really not very significant.

Quote

This makes mechs (like the Night Gyr) pick some mobility nodes over firepower so they could spread out the damage as they cant tank like the assaults.

No need for sacrifice. Again, the Night Gyr can get mobility, survival and firepower with no need for real sacrifice. Well it has jumpjets so if it wants better jumpjets it might have to give something up for that maybe? Probably not though.

Quote

Question, do you know what the "balance" or "meta" will be in the skill tree with the addition of the new tech? Also, they are being compensated as the HP of equipment is being changed based on size. This makes it so the average internal HP of Clan Assaults drops a little while quite massively increasing the internal HP of IS Assaults (and other IS mechs).

I don't think you understand how internal damage works. Having higher component HP does not impact how much damage you internal structure can take.

#194 Athom83

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Death Wish
  • The Death Wish
  • 2,529 posts
  • LocationTFS Aurora, 1000km up.

Posted 20 March 2017 - 06:20 AM

View PostBud Crue, on 16 March 2017 - 10:33 AM, said:

I care that they are nerfing...and seemingly randomly so...what WE KNOW, objectively know...to be some of the worst mechs, and making them worse. Those mechs are not going to magically be made good in the new system.
No, it takes time and effort put into something different to make them good. The current system of making every mech have the same skills with the same grind turns people away from those harder to play mechs as they have to put more effort into trying to level. In a system where you can add reasonable capability without needing to grind out 2 other mechs that are also difficult to play makes it more reasonable to play that said mech. When you play that said mech longer, you start seeing better ways of how to play that mech, to do better. You also have to remember, those objectively worst mechs are centuries old in the lore, fighting mechs only decades old. The Victor was designed and built from the 2500s, and the Kodiak from 3001.

View PostBud Crue, on 16 March 2017 - 10:33 AM, said:

You asked about what I think is going to be meta in the new system? Same as the current system. New weapons might change that, but the skills tree changes nothing. Hunch IIc, Night Gyr, and yes even the poor, sad Kodak 3 are still going to be great (though as an exemplar I'm almost tempted to say Mad IIc). The skills tree is not going to make a newly nerfed St. Ives Blues the new medium of choice. Nor is the nerfed Orion going to suddenly become dominate over the Night Gyr. Etc.
Funny for you pick the Orion as an example of an "underperforming" mech. I nicknamed mine "Clan Hunter" as I just tear through clan Heavies and Assaults in it. With 2 Ultra 5s and some SRM backup weapons, I seek out those clan mechs seen as "Overpowered" (Timberwolf, NIght Gry, Mauauder IIC, Kodiak) and take them head on. Its really fun when you tear off a Marauder IICs ST while they attempt to twist and get an ammo explosion, run circles around a Kodiak and pop his CT from the back as they always seem to front-load all of it, or drill through a Mad Cat that doesn't even glance at you in a few second because you're in a "bad mech". The only 2 mechs I've been called a hacker in was that and the Mauler-2P, and neither of those were "mastered" in the current system.

The whole thought process of "Players are going to naturally skill out and play the one or two variants from each chassis that have inherently better "baseline" characteristics." assumes that they don't already, which they do.

View Postsoapyfrog, on 16 March 2017 - 10:46 AM, said:

Well it has jumpjets so if it wants better jumpjets it might have to give something up for that maybe? Probably not though.

When they finally bring in Death from Above, JJs suddenly become more viable for those heavier mechs.

Edited by Athom83, 20 March 2017 - 06:34 AM.


#195 Bud Crue

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 9,946 posts
  • LocationOn the farm in central Minnesota

Posted 20 March 2017 - 08:51 AM

View PostAthom83, on 20 March 2017 - 06:20 AM, said:

No, it takes time and effort put into something different to make them good. The current system of making every mech have the same skills with the same grind turns people away from those harder to play mechs as they have to put more effort into trying to level. In a system where you can add reasonable capability without needing to grind out 2 other mechs that are also difficult to play makes it more reasonable to play that said mech. When you play that said mech longer, you start seeing better ways of how to play that mech, to do better. You also have to remember, those objectively worst mechs are centuries old in the lore, fighting mechs only decades old. The Victor was designed and built from the 2500s, and the Kodiak from 3001.


You're arguing lore to justify why some mechs in a competitive FPS should be allowed to remain objectively worse than another newer mech.
Okay.

As to your Orion, well, it's a good thing it getting nerfed then. We wouldn't want such a superior clan killer to be allowed to remain in such an obvious OP state.

Edited by Bud Crue, 20 March 2017 - 08:55 AM.


#196 WolvesX

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Machete
  • The Machete
  • 2,072 posts

Posted 23 March 2017 - 06:56 AM

View PostBud Crue, on 20 March 2017 - 08:51 AM, said:

You're arguing lore to justify why some mechs in a competitive FPS should be allowed to remain objectively worse than another newer mech.
Okay.

As to your Orion, well, it's a good thing it getting nerfed then. We wouldn't want such a superior clan killer to be allowed to remain in such an obvious OP state.

Ya, why not balance LORE wise.

The Black Knight is a big Clam Killer!

Why not give it 150% dmg to clam mechs?

---

I can't understand people who are against balance in a game. Its so stupid. Plain moronic.

#197 Widowmaker1981

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Widow Maker
  • The Widow Maker
  • 5,031 posts
  • LocationAt the other end of the pretty lights.

Posted 24 March 2017 - 05:12 AM

View PostAthom83, on 16 March 2017 - 09:37 AM, said:

In what kind of world would the same motor be able to twist the 80 something tons of an Assault torso the same speed as the 10 tons of a Light torso? This is why coupling twist speed with engine rating doesn't make sense.


Just want to point out that this is a HUGE misunderstanding of the current system.

Its not coupled to engine rating, its coupled to Engine rating divided by tonnage. An Atlas with a 300 twists half as fast as a 50 ton mech with a 300 (assuming no agility quirks)





4 user(s) are reading this topic

0 members, 4 guests, 0 anonymous users