Jump to content

How Class Balance Happens

EricMusco

By EricMusco
in General Discussion

 Share

Recommended Posts

  • Dev Post
EricMusco

EricMusco

Posted
  • Dev Post
Posted

Hey folks,

 

Over the next few weeks, as you know, it is our plan to give you specifics on upcoming Class/Discipline changes. One things that we have never done before is really get down in the trenches and explain how we balance Classes. With our continued move towards more transparency, we really wanted to lift the hood and provide as much context as possible for those changes. Below you will find an elaborate breakdown from our combat team on not only how they balance Classes, but why. I highly recommend you read the entire post (I know it’s long!). For the non-reader among you…

 

TLDR – We do not balance Disciplines against each other directly. Balance is based on target values which are determined by what type of damage dealer you are; ranged, melee, burst, sustained, etc.

 

Here is the detailed breakdown from the combat team:

Opening note, for anyone who might not be familiar with the abbreviations and terminology used in this post:

  • DPS = Damage per second
  • HPS = Healing per second
  • DTPS = Damage taken per second
  • “Nerf” = Reduction to effectiveness
  • “Buff” = Increase to effectiveness
  • PvP = Player(s) versus player(s) combat
  • PvE = Player(s) versus non-player(s) combat (non-players are controlled by artificial intelligence programming)

 

Before we dig into the details for any Classes or Disciplines that will be seeing damage or healing “nerfs” in 5.3, we wanted to take a moment to share a bit of our Class design philosophy with you. Whenever we “nerf” the damage or healing a Discipline can produce, it is not because we compared them to other Disciplines and thought that they were too good. Likewise, we don’t “buff” a Discipline because we think they are not as good as the other Disciplines and need a boost. We do not play the Class comparison game for damage, healing, or tanking changes.

 

The deeper reason for a “buff” or a “nerf” lies in a Discipline’s ability to perform at their target DPS (for damage dealers), HPS (for healers), or DTPS (for tanks). The target DPS, HPS, and DTPS values determine which content (Veteran Operations, Master Uprisings, etc.) players can successfully complete in a given set of gear. And in PvP situations, they determine the speed of player kill times and the ability for healers and tanks to extend the life of an ally under attack. So to clarify, there is a comparison that takes place for damage, tanking, and healing balance, but that comparison is not among Disciplines or Classes – it is between the target DPS, HPS, or DTPS value and the Discipline in question.

 

Due to the variances with the item rating of the gear your character has equipped and the augments, Class buffs, stims, and adrenals your character may or may not be using, we are not providing you with a set number for any target values. Also, these values are average values. Every damage, tanking, or healing parse tends to be unique, and sometimes those parses can stretch quite high or low for several different reasons.

 

That said, we do have something a little special which we would like to share with you today: damage grouping categories. With no further ado…

 

Damage Groupings for Damage Dealing Disciplines

Melee Sustained Damage Dealers (up to +5% of target DPS)

  • Annihilation Marauder / Watchman Sentinel
  • Hatred Assassin / Serenity Shadow
  • Lethality Operative / Ruffian Scoundrel
  • Pyrotech Powertech / Plasmatech Vanguard
  • Vengeance Juggernaut / Vigilance Guardian

Melee Quasi-Burst Damage Dealer (up to +2.5% of target DPS)

  • Fury Marauder / Concentration Sentinel

Melee Burst/Ranged Sustained Damage Dealers (at the target DPS)

  • Advanced Prototype Powertech / Tactics Vanguard
  • Carnage Marauder / Combat Sentinel
  • Concealment Operative / Scrapper Scoundrel
  • Deception Assassin / Infiltration Shadow
  • Engineering Sniper / Saboteur Gunslinger
  • Innovative Ordnance Mercenary / Assault Specialist Commando
  • Madness Sorcerer / Balance Sage
  • Rage Juggernaut / Focus Guardian

Ranged Quasi-Sustained Damage Dealer (down to -2.5% of target DPS)

  • Virulence Sniper / Dirty Fighting Gunslinger

Ranged Burst Damage Dealers (down to -5% of target DPS)

  • Arsenal Mercenary / Gunnery Command
  • Lightning Sorcerer / Telekinetics Sage
  • Marksmanship Sniper / Sharpshooter Gunslinger

 

PLEASE NOTE: The Disciplines are in alphabetical order within each grouping, not by damage dealt (for example, Deception Assassins are not intended to deal higher DPS than Saboteur Gunslingers). There are also groupings for healers and tanks, but we are not listing them since all healers are in the same groupings and all tanks are in the same groupings.

 

Now let us explain the theory behind the planned differences for the damage groupings above. There are really only four sub-categories we use in order to classify a damage dealer: Melee (meaning short-range), Ranged (meaning long-range), Burst (meaning immediate or quick damage), and Sustained (meaning damage over time or slow damage). We feel both Melee and Sustained damage types are inherently disadvantaged, and that is why Melee Sustained damage dealers (with their two inherent disadvantages) make up the top grouping. Melee damage is disadvantaged because it must stay close to its target (often within four meters) in order to maximize damage output, and Sustained damage is disadvantaged because it takes several seconds to ramp up to its full damage potential each time it acquires a new target (and in situations where rapid target switching is required, it will often never reach its full damage potential).

 

Due to these inherent disadvantages, those classified as Sustained damage dealers are given a damage output advantage over those classified as Burst damage dealers. Similarly, those classified as Melee damage dealers are given a damage output advantage over those classified as Ranged damage dealers. Melee Burst and Ranged Sustained damage dealers fall into the same grouping because they each have one inherent disadvantage, even though those disadvantages are different (one being Melee and the other being Sustained). And in the last grouping, we have Ranged Burst damage dealers, which have no inherent disadvantages. They can quickly change targets without needing to close a gap, and they have little to no ramp up time before they start dealing high DPS on a newly acquired target.

 

It can be noticed that two of the groupings are missing in the above explanation. Those groupings belong to the Advanced Classes that have three damage dealing Disciplines instead of two damage dealing Disciplines and one tanking or healing Discipline. We could have technically tried to squeeze these Disciplines into one of the other three damage groupings, but we separated them out to differentiate them within their Class while still keeping them in line with what other Classes have to offer.

 

Another thing about the damage groupings above that might be a little confusing is the reference to “target DPS,” so let us give you a fictitious example to help clarify our meaning. Remember, target DPS varies based on your equipped item rating and any augments, Class buffs, stims, or adrenals that you may or may not be using. For ease of calculation’s sake, we will say the fake target DPS for our example is 10,000. With 10,000 DPS being the target value, that would mean a Rage Juggernaut / Focus Guardian should average 10,000 DPS in a fight against a single target that lasts about five minutes (because they fall in the “at target” damage grouping). In the same fight, a Pyrotech Powertech / Plasmatech Vanguard should average above that target DPS, up to 10,500 DPS (because they fall in the “up to +5% of target” damage grouping). A Virulence Sniper / Dirty Fighting Gunslinger performing in this same scenario should average below that target DPS, down to 9750 DPS (because they fall in the “down to -2.5% of target” damage grouping).

 

Before we move on from this topic, we want to note the point about the fight lasting around five minutes is rather important. In a much shorter fight, the bursty Rage Juggernaut / Focus Guardian might have higher DPS than the Pyrotech Powertech / Plasmatech Vanguard, because it takes a while for the latter’s sustained damage to ramp up to its full potential.

 

Okay, so now you might be looking at the groupings above and scratching your head thinking, “That’s not AT ALL where my Discipline is!” If so, you might be correct (depending on which Discipline you play). In fact, that is precisely why we do Class balance changes: to fix things that often fall out of balance when we do something like a level increase or push out new abilities and skills. Achieving the damage groupings above is the goal of our Class balance changes beginning with 5.3 and continuing in the patches beyond it as necessary, until the damage grouping balance above is a reality in game and not just on paper.

 

We hope sharing these damage groupings with you is informative. When you see all the changes coming with the 5.3 update, we believe you will be able to decipher whether your Discipline is going to be seeing a “buff” or a “nerf” in the future. If you see another Discipline in your grouping got “nerfed” and your Discipline’s damage output is now quite a bit higher than theirs, then it is a good bet when we get around to balancing your Discipline it will also be seeing a reduction in damage output.

 

At this point, you might be wondering how we decided which Disciplines to balance first. We used several factors including player feedback, player parsing data, and our own internal PvE and PvP statistical data. For 5.3, we focused on the Disciplines we felt were most out of line with the rest of their damage groupings. That gave us the Virulence Sniper / Dirty Fighting Gunslinger, Engineering Sniper / Saboteur Gunslinger, and Innovative Ordnance Mercenary / Assault Specialist Commando on the overperforming end and the Pyrotech Powertech / Plasmatech Vanguard, Hatred Assassin / Serenity Shadow, and Madness Sorcerer / Balance Sage on the underperforming end. If you are wondering whether Arsenal Mercenary / Gunnery Commando and Corruption Sorcerer / Seer Sage got “nerfs” or “buffs,” well… just have another look at those damage groupings and take this hint to heart: it is the same story for Corruption / Seer healing as it is for Arsenal / Gunnery damage.

 

Lastly, you might also wonder why there are not more changes to utilities and survivability for damage dealers in 5.3, and that is a fair question. We are first focusing on the damage and healing output for all Classes before getting too carried away with utility changes. This is because opinions about where a given Discipline stands among the pack can be clouded by high or low damage or healing output without anyone even realizing it. That is not to say we won’t tweak a utility here or there, but before we go making massive changes to a class’s utility toolkit, we want to make sure the damage or healing they can put out is on target and relatively similar to other Disciplines in their grouping. Once we are happy with the damage output, we can start adjusting utilities to give those that need it a boost and others that are too powerful a reduction.

 

We hope this glimpse into our Class design philosophy helps you understand the changes we are making and the rationale behind each decision.

 

-Your friendly neighborhood Combat Team

Link to comment
Share on other sites
  • Replies 489
  • Created
  • Last Reply

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

DenariusJay

DenariusJay

Posted
Posted (edited)

tldr - gunnery commando / arsenal merc will still be OP in pvp..... -_-

 

 

welp I guess its official, time to pack up my melee characters in pvp and play merc full time. so boring, but w/e. atleast we know now.

Edited by DenariusJay
Link to comment
Share on other sites
Dewlmenow

Dewlmenow

Posted
Posted
This is cool, but I still have to question why lightning and to a lesser extent marksman aren't being balanced immediately when there is a ~15% difference between the top melee sustained in lethality and lightning. 15% is a relatively large departure from your stated 10% difference you are trying to work towards.
Link to comment
Share on other sites
yellow_

yellow_

Posted
Posted (edited)

A more serious post

before we go making massive changes to a class’s utility toolkit, we want to make sure the damage or healing they can put out is on target and relatively similar to other Disciplines in their grouping
But utilities can have the exact same impact...especially in PVP

 

Some of the utilities you introduced in 5.0 (*cough* trauma regulators) are so absurdly powerful that the amount of pressure they relieve allows the classes that have them to overperform to extreme levels regardless of how well they parse on a dummy. Marksman is the perfect example of this. It parses poorly but overperforms in PVP thanks to the new utility kit. You can put arsenal down in the garbage bracket with marksman and it will continue to overperform in PVP for the same reasons.

 

Keith...please...

Edited by yellow_
Link to comment
Share on other sites
Lionflash

Lionflash

Posted
Posted
That is not to say we won’t tweak a utility here or there, but before we go making massive changes to a class’s utility toolkit, we want to make sure the damage or healing they can put out is on target and relatively similar to other Disciplines in their grouping. Once we are happy with the damage output, we can start adjusting utilities to give those that need it a boost and others that are too powerful a reduction.

 

Thanks, that was actually an interesting read.

 

I have a question though; I know you said you wouldn't be focusing on utilities yet, but will there be any bug fixes to utilities? If some utilities just don't work or are bugged.

Link to comment
Share on other sites
TUXs

TUXs

Posted
Posted

Awesome post. The communication and transparency from this is remarkable.

 

We'll never all agree on what needs to be nerfed/buffed, but seeing the methodology behind your changes is refreshing.

Link to comment
Share on other sites
Kurj

Kurj

Posted
Posted
That was an interesting read. I'm still absorbing some of it, but it is hugely appreciated to actually be invited into a mature conversation with the developers of my favorite game, and about something that is vital to the core game. I don't know if you guys realize the change in perception you have caused just by doing that, but it makes a big difference to me.
Link to comment
Share on other sites
DenariusJay

DenariusJay

Posted
Posted
A more serious post

Marksman is the perfect example of this. It parses poorly but overperforms in PVP thanks to the new utility kit. You can put arsenal down in the garbage bracket with marksman and it will continue to overperform in PVP for the same reasons.

 

Keith...please...

 

Without getting too far off topic - I'm not a sniper pro at all and I know you are good player in this game, (we play on the same server) how do you reckon marksman is overperforming ? I play my sniper quite avidly, I find a lot of MM's dmg is avoidable (mainly Ambush) and there AOE capabilities is the weakest of the three specs. I would reckon Engi is the overperforming class in PvP atm.

Link to comment
Share on other sites
Ottoattack

Ottoattack

Posted
Posted (edited)

To start off, it nice we have some sort of open dialogue regarding class balance. Though providing the message is nice, the message itself is not.

 

First just looking through the list from PvE standpoint, there are many things from design that do not make sense. How is fury quasi burst? Fury is as burst as it gets. In fact, there is no reason at all it should deal more damage than carnage, as it offers similar burst with no setup (and it does not). In addition, a 10% dps swing between classes is way too high. Though we are currently at 15% between highest and lowest disciplines, which is even a wider gap, than the already wide gap that we are supposed to have.

 

Of course needless to say based on this list reality is much different. Arsenal out dpses AP. Veng is put in the theoretically in the highest damage category, which it is not, etc...

 

Still in PvE:

Lastly, you might also wonder why there are not more changes to utilities and survivability for damage dealers in 5.3, and that is a fair question. We are first focusing on the damage and healing output for all Classes before getting too carried away with utility changes.

 

Where class effectiveness makes the most difference, NiM ops, mobility and suvivability are as important if not more important than pure dps (as long as you can maintain a dps minimum). Fixing dps fixes very little, if your problem is survivability. AOE also is not mentioned, which makes a huge difference when it is required.

 

As for PvP, there are million other factors. There is no reason to sit here and try to list, but dps output on a non moving dummy (or ops boss) means nothing. Many classes with low dps output perform strongly in PvP and vice versa. In fact, almost no dot spec is viable in competitive PvP (annihilation is a maybe). Clearly that is not a dps output issue.

 

Jug issues in PvP have nothing to do with output.

Ops issues in PvP have nothing to do with output.

Merc issues in PvP have nothing to do with output.

Sorc healers issues in PvP have nothing to do with output.

Sniper issues...

 

Getting redundant isn't it? Yes, fixing dps issues will help dps sorc and dps PT, but for the most part it is not the primary issue for the majority of disciplines and if you wont address anything else, well it is pointless. Survivbaility, mobility and utility (range for some disciplines like PT dps and Hatred sings) are significantly more important as balance currently stand. That is why the Hatred changes you guys suggested are borderline useless. If you will not address the class balance issues now, well, what is the point of class balance then?

Edited by Ottoattack
Link to comment
Share on other sites
yellow_

yellow_

Posted
Posted (edited)
Without getting too far off topic - I'm not a sniper pro at all and I know you are good player in this game, (we play on the same server) how do you reckon marksman is overperforming ? I play my sniper quite avidly, I find a lot of MM's dmg is avoidable (mainly Ambush) and there AOE capabilities is the weakest of the three specs. I would reckon Engi is the overperforming class in PvP atm.
Snipers are overperforming in PVP in general because they're far too hard to kill. Engineering even more so because of the plasma probe changes but that's not the point.

 

The point is that marksman parses where it's supposed to according to this post by BW, and yet it is still overperforming in PVP because it, like all the sniper specs, is too hard to kill. There are multiple MM snipers above 3K in solo ranked, they're extremely common and effective in group ranked as well, and good MM snipers regularly top DPS scoreboards simply because they have to put so little effort into dying compared to many other classes that their poor parses mean almost nothing in PVP. If it were the fact that they're a ranged burst class, you'd expect lightning to be similar. But it isn't -- it's awful.

 

I'm making a point about their decision to apparently forgo utility changes for the most part in favor of damage output. It's not necessary that you believe me about MM -- take arsenal for example. What are the odds that Arsenal won't continue to be painfully dominant in PVP if they just slightly tweak its damage output but leave the trauma regulators utility intact? To me the answer to that is extremely obvious - the odds are about zero.

 

As for PvP, there are million other factors. There is no reason to sit here and try to list, but dps output on a non moving dummy (or ops boss) means nothing. Many classes with low dps output perform strongly in PvP and vice versa. In fact, almost no dot spec is viable in competitive PvP (annihilation is a maybe). Clearly that is not a dps output issue.

 

Jug issues in PvP have nothing to do with output.

Ops issues in PvP have nothing to do with output.

Merc issues in PvP have nothing to do with output.

Sorc healers issues in PvP have nothing to do with output.

Sniper issues...

 

 

what this guy said (I don't necessarily agree about dot specs but its beside the point)

 

lethality is garbage in PVP, marksman is top tier - enough said

Edited by yellow_
Link to comment
Share on other sites
K_osss

K_osss

Posted
Posted
Fascinating read, although being an Arsenal Merc, that was as clear as mud in terms of whether we are getting the hammer or not, lol.

 

Read it again, you're getting the hammer at least on damage and maybe the hammer later on utilities.

Link to comment
Share on other sites
DarthCognusSion

DarthCognusSion

Posted
Posted

Eric,

Can you explain why the target dps spread is 10%? Why do you feel you cannot make the spread narrower. The issue with a 10% spread is that when you have fights that have really tight dps checks, the disciplines at the bottom are not favored even if having burst classes is more beneficial for the fight because they either cannot do enough damage or they can but that means the dps at the top just do that much more and trivialize the fight. A narrower spread can still achieve what you want without putting the bottom dps at a disadvantage. For instance:

 

Highest DPS (+2.5% compared to average): Annihilation/Watchman, Lethality/Ruffian, Hatred/Serenity, Pyrotech/Plasmatech, and Vengeance/Vigilance

Mid DPS (+/-0% aka average): Carnage/Combat, Fury/Concentration, Rage/Focus, Concealment/Scrapper, Advanced Prototype/Tactics, Deception/Infilration, Virulence/Dirty Fighting, Engineering/Saboteur, Madness/Balance, and Innovative Ordinance/Assault Specialist

Lowest DPS (-2.%% compared to average): Lightning/Telekinetics, Arsenal/Gunnery, and Marksmanship/Sharpshooter

 

Is it just not possible for you guys to make a spread that is something like 5%-7.5% from top to bottom, or do you just have something against it for some reason?

Link to comment
Share on other sites
ZionHalcyon

ZionHalcyon

Posted
Posted

Can I just take a moment and give kudos to the development team and the new direction the lead producer is taking the game with regards to player communication?

 

Its about damn time, and this was a hell of a read that will help all of us players finally understand where the development team is coming from in terms of class balance. We finally understand the correct frame of mind and what comparisons they are drawing, which will make our opinions more informed and laser-focused.

 

Bravo, and keep up the good work with posts like this!

Link to comment
Share on other sites
Gyronamics

Gyronamics

Posted
Posted (edited)

Well that's interesting reasoning.

 

So this means the highly unplayed Innovative Ordnance Mercenary / Assault Specialist Commando spec is up for a damage nerf.

 

I could write you a few pages on why it's unpopular as hell and only part of it is down to the Gunnery/Arsenal performance. But would you want to hear it?

 

Theoretical or even demonstrated outputs on a dummy have nothing to do with PVP or casual daily fighting with mobs.

 

There's a large gap between why you're adjusting damage and what players are actually playing.

Edited by Gyronamics
Link to comment
Share on other sites
Evolixe

Evolixe

Posted
Posted (edited)

All I really gather from this post is that basically you are going to change Mercs Damage and Healing and not touch their utilities. I mean that's kind of exactly what you're saying. And that's also the exact opposite of what you should do.

 

Doc said it very well. Utilities and that sort of stuff can have just as big an impact.. if not a much bigger one than damage or healing being out of line. I think it's pretty easy to see if a specs damage is out of line or not gathering from experience in PvE. I will tell you straight up that I don't think Mercs are overperforming at all right now when it comes to that.

 

However.. when you go into PvP, they are suddenly living, breathing hovertanks.

They literally take twice if not thrice the firepower to kill from your average Assassin, Marauder or Juggernaut.

 

Where do you think the problem lies then? I really don't think its damage.

 

 

Anyway, I could ramble on for hours about how I suddenly understand but massively disagree with your way of doing "class balance" It won't have any effect.

 

I just want to say one thing though.. when was the last time you guys were actually.. ACTUALLY happy with the class balance in the game? Because I can't personally think of any point in time it was all well-done.

 

Food for thought.

Edited by Evolixe
Link to comment
Share on other sites
Thaladan

Thaladan

Posted
Posted (edited)

8 months soon of total imbalance like we never had before, to the point where the state of PvP is just terrible and this your solution for 5.3? :confused:

 

do you PvP keith? but srly?

 

put 2 clases equal in dps but one with perma snare on each hit and self heal for 3 lives and the other with nothing..but all the same dps possibility. guess who will win !!! :mad:

 

after 8 months you will just adjust some damage % :( there is a major issue about survivabilty, snares/stuns and range.

 

8 months in 5.3 heyyy ohhh wake up where are your metric in regs and leaderboards?

Edited by Thaladan
Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...