Guild Wars 2 Balance Philosophy

[Rubi Bayer.8493]

Hi all, I have a message from Skills and Balance Lead Cal Cohen:

Hi, everyone,

Today we'll be going over our current balance philosophy for Guild Wars 2. Our goal is to give you all insight into some of the things we consider while working on balance.

 his is a living document; our balance philosophy has changed over the years, and it will continue to adapt as needed for the health of the game. When the philosophy needs to be adjusted, we'll communicate what is changing and why.

This document will cover some ideas that are not perfectly represented in the current state of the game. We'll be working to resolve any balance issues that don't align with the philosophy, but this is something that will happen incrementally over time.

Goals

Fundamentally, our goal is to ensure that the moment-to-moment gameplay in Guild Wars 2 is enjoyable for as many players as possible. To that end, we want to capitalize on the depth of the combat system to build a fluid and fast-paced combat experience that allows players to express their mastery of mechanics. We also want to create a substantial number of viable build options and allow for a broad set of combat strategies in order to enable a wide range of playstyles.

Combat Depth and Build Complexity

Guild Wars 2 has a deep combat system, and players have a very wide range of mastery of its mechanics. To put things simply, we want to build a game that is both rewarding and accessible for all types of players.

We want to design builds that allow players with a high level of mastery to demonstrate their prowess and be appropriately rewarded in terms of effectiveness. At the same time, we want to ensure that there are builds for every profession that require less mastery to be effective. These builds should allow players to succeed in parties and clear content, while still having room for them to improve their mastery over the combat system and increase their effectiveness.

This is also an important consideration for balance in competitive game modes, as the builds that are effective can vary significantly between different levels of mastery. Our goal is to create a fun and diverse metagame for as many players as possible, and that involves addressing builds that are problematic at any level, even if they aren't problematic at every level. When bringing down a build that only overperforms at a particular level, we'll try to target changes to minimize the impact on other levels or attempt to otherwise compensate in a way that is less problematic at the targeted level.

 

Gameplay Roles

Roles are the general playstyle a player wants to achieve, and they determine which responsibilities a player fulfills in a group scenario. For the purpose of balance, we consider a few main roles for each game mode. While not every build needs to perfectly fit into a role, these are the most common archetypes that we look at when balancing.

Within a role, we want different builds to have distinct strengths and weaknesses, and therefore different considerations when building a composition. There will always be some overlap (damage dealers are all good at dealing damage), but the secondary elements should be different enough that each build feels unique.

Player vs. Environment (PvE)

PvE group compositions are typically built to maximize damage output through high might, fury, quickness, and alacrity uptime, with just enough support and defense to keep everyone alive.

Damage Dealer

The primary source of damage. Built to maximize damage output, they bring minimal group utility, though they may share some offensive boons in limited amounts.

Boon Support

A hybrid role focused on providing high uptime of key offensive boons, though a single build should not provide both quickness and alacrity. They also contribute to damage or healing in lesser amounts than dedicated builds for those roles.

Healer

A support role that focuses on keeping allies alive through defensive boons and raw healing. They may also provide some offensive boons to the party.

World vs. World (WvW)

WvW group compositions have a similar makeup to PvE group compositions, with a focus on damage dealers to deal damage and support characters to defend them. Stability is always in high demand and is essentially a requirement for every group.

Support

Support is a broad term, and there are a variety of distinct tools that can be the focus of a support build in WvW, including healing, condition removal, boons, crowd control, and other general utility tools. Most support builds bring more than one tool to the table, but it's important that a single build can't excel at too many things.

Damage Dealer

These are builds that primarily exist to deal damage, but they also commonly bring additional pressure tools such as boon removal or crowd control. As fights get larger, area-of-effect damage becomes more important and single-target pressure loses some of its value.

Player vs Player (PvP)

Over the years, we've seen metagames dominated by both team fight compositions built around the support role, and more split compositions built around bruisers or mobility. Ideally, we'd like to get to a state where multiple styles of team compositions are viable.

Support

This role empowers and defends allies. They are most valuable in larger fights where their lack of damage is made up for by other teammates. Most supports are good healers and have access to group condition cleansing, but beyond that, they may specialize more in defense (defensive boons, healing, and cleansing), offense (offensive boons, crowd control, and personal damage), or a mix of both.

Team Fight Damage Dealer

These are damage builds that typically sacrifice some defensive options for more offense, with the assumption that they will usually be fighting alongside a support role. Their defensive tools are more often focused on hard mitigation, with minimal self-sustain.

Bruiser

These are bulkier damage dealers who trade some damage for more self-survivability and crowd control. They are generally good team fighters, but they can also flex on to the sidenode in some matchups.

Roamer

This is commonly a bursty damage dealer with high mobility. They look to capitalize on number-based advantages and end fights quickly. They have some defensive capabilities but are usually unable to stay in drawn-out fights.

Sidenode

The duelist. They're most effective in 1v1s and smaller fights but can sometimes flex into team fights.

Skill and Trait Design Guidelines

The following are a few key ideas that we consider when balancing skills and traits. This list isn't intended to be absolute in all cases, but there should be a strong reason when a skill breaks one of these rules.

Purity of Purpose

Purity of purpose is the idea that a skill (or trait, or weapon, etc.) should have a well-defined identity. In other words, skills should not do too many different things at once. Some common skill identities include damage, defense, support, control, and mobility.

Holes in Roles

This is an idea similar to purity of purpose, but applied to builds or professions. As we touched on when discussing identity, we want every profession to have distinct strengths and weaknesses. Professions should have things that they excel at, things that they are less effective at than other professions, and some things that they simply cannot do. If one profession does everything and has no holes, there's no reason for players to play anything else.

Power Budget

For a given skill or trait, there is a "power budget" that can be spent on individual elements. A skill that only deals damage can deal X damage, but if that skill also applies conditions or provides other value, then it needs to deal less damage to stay within budget. In the context of a weapon, the budget is considered across the entire kit, so some skills may be weaker than average in order to allocate more power to a particular skill. Budget can also vary depending on a skill's cooldown; skills with longer cooldowns are generally more powerful.

Trait budgets are to be considered on a per-tier basis; adept traits should have less power than grandmaster traits.

Play and Counterplay

Counterplay is a fundamental piece of competitive gameplay in Guild Wars 2, and it's important to build skills that can be interacted with effectively. This means we generally don't want instant-cast skills that heavily impact an enemy because we want players to be able to see and react to what their opponents are doing. Instant skills are usually best as defensive skills, though we also want to avoid instant healing in significant amounts.

Minimizing Bad Choices

This is just another way of saying that we want as many build components as possible—weapons, slot skills, traits, etc.—to have situations that they are viable in. Some skills may be restricted to more niche applications, but we want to avoid cases where a skill simply has no relevant use case. This can sometimes be difficult when considering the needs of multiple game modes, but that leads to our next topic: skill splits.

Skill Splits

Guild Wars 2 has three primary game modes that are considered for balance: PvE, PvP, and WvW. Each of these modes require different balance considerations, and it's not always possible to design a skill or trait that fills the needs of every game mode without any adjustments.

Skill splits began as a system that allowed adjusting a skill or trait's effectiveness between game modes while maintaining consistent functionality in every game mode. If a skill applied quickness, it was required to apply quickness in every game mode with different durations. Over the years we've seen the limitations of this approach, and we believe that the needs of each mode are different enough that skill splits also need to include some mechanical or functional changes. We've started to make broader splits over the last few months. When we decide a functional split is necessary, we still want the general purpose of a given skill or trait to be consistent across all game modes. Defensive skills should be defensive skills, but the way a skill provides that defense may be different between modes.

In cases where the core mechanic of a skill or trait is problematic in a particular game mode, we'll investigate if there's a way to rework the mechanic that feels good for every mode. In extreme cases, we may decide to significantly adjust how a skill behaves in a single mode, but this would only happen if the skill causes a major balance issue, there isn't a viable rework, and the skill cannot be balanced effectively while respecting the usual considerations of skill splits. Ideally, we want to avoid splits of this nature as they significantly increase the learning curve for players who play multiple game modes, but we will still utilize them when necessary.

Incorporating Player Feedback

The final topic we want to touch on is how we utilize player feedback throughout the design process for a balance update. Every update starts by determining what changes we want to make. We do this by reviewing recent live data for each game mode to identify overperforming and underperforming builds, but also by reading through player feedback to check for common pain points. After we have a plan, we design, prototype, playtest, and iterate until we've taken our initial goals and turned them into a finished set of changes, ready for wider community feedback.

Once we get to the preview stage, we collect player feedback from a wide variety of channels. At this point, feedback is even more important, as it gives us insight into what potential problems exist in the upcoming update, and what things we need to resolve between the preview and the live release. Every change that makes it to the preview is there for a reason, and it's important for us to provide that reasoning to make sure that everyone is aware of our goals. When community sentiment is negative around a particular change it's important for us to understand why that is, so that we can consider those reasons against the initial reason for the change and determine whether there's a viable solution.

One last note about previews and feedback: mechanical changes are the things we focus on most in the feedback phase. We do still look at feedback around numerical changes, but a lot of the time these can be more difficult to evaluate on paper and we usually want to get some actual data on how they play out. Numerical changes are also much easier to tune in later updates, even outside of the regular balance cycle. This isn't to say that numerical changes will never get adjusted because of preview feedback; they're just a lower priority for us compared to any larger mechanical updates.

Player feedback is an extremely valuable tool in the game design process. Thanks to everyone who regularly contributes to the discussion.

Conclusion

As we mentioned back at the start, going forward we'll be working to identify areas where the philosophy can be improved and resolve any outstanding issues where the live game doesn't align with the philosophy.

One final note: we've found the opportunity to gather initial feedback from either preview streams or forum posts to be incredibly valuable over the last few months, and we'd like to try to find more ways to get the community involved early. Finding the proper timing is the hard part, as we want to make sure we have enough time to put together an impactful update, but we'll be thinking about ways that we can improve the current process.

Thanks, everyone, for reading. We're looking forward to following the discussion.

Cal "cmc" Cohen

Skills and Balance Lead

[Kcquipor.6730]

We need this : https://wiki.guildwars.com/wiki/Test_Krewe 

[DanAlcedo.3281]

I just want to say:

Thanks for the Communication.

[Caysadia.2548]

Shattered Aegis

[Bookah pls.9352]

Hello, when do reapers get to spend the rest of their tickets?

[TheGrimoire.4039]

You keep asking for feedback but there's YEARS worth of feedback on the Elementalist forum, and nothing.
Sad to see.

[Zuko.7132]

I would love to see a breakdown of what roles Anet sees each profession or elite spec as capable of or meant to perform in pvp, and balance to back up and make those roles viable. For example, I assume berserker is supposed to be a teamfighter, but it clearly is failing at succeeding in that role and thus needs buffs. An acknowledgment of what roles professions and elite specs are supposed to play along with whether or not Anet thinks that those roles are currently effective, underpowered, or overpowered would be very helpful.

[Antina.5973]

That's a good start!

[jozze.9532]

5 minutes ago, Bookah pls.9352 said:

Hello, when do reapers get to spend the rest of their tickets?

And when does firebrand give reaper their tickets back 😄

Was nice to see the stream. Looking forward to the extra live stream and the patch preview.

[Vampiremax.8350]

Very good stream and document! Keep up the great communication 🙂

[SamuelW.2685]

8 minutes ago, TheGrimoire.4039 said:

You keep asking for feedback but there's YEARS worth of feedback on the Elementalist forum, and nothing.
Sad to see.

I get it, I main Thief and I got burned pretty bad from the October changes to Specter and I know Ele has been shafted for like, ever, but we just got a new balance lead, balancing things take time, especially with how AWFUL it's been prior to Cal coming in.

Will he improve things overall? Who knows! There were good and bad things about the October patch, but we need to wait and see for more patches to form a good opinion. The communication in the form of this stream is a very good sign and them saying the read the forums is very good too.

Edited October 28, 2022 by SamuelW.2685

[Arilozen.2756]

Awesome stream! I have some feedback to share for the hydro homies (Roy and Cal).

Long time warrior main. I've always wanted to use rifle, but compared to every other weapon it's fallen short in PvE.

I've done hours and hours of testing, trying my best to squeeze the maximum DPS out of the rifle that I could. (The Predator was my first ever legendary near the start of the game  so I'm pretty motivated.)

I understand that it's largely balanced around its burst skill. That being said, even with an optimal rotation in perfect conditions it struggles to compare to even the most basic rifle Mechanist using minimal effort. Is it plausible to shift some of the power budget away from the rifle burst skill to have an improved auto attack? What are your thoughts?

Much love.
 

[Gibson.4036]

2 minutes ago, Zuko.7132 said:

I would love to see a breakdown of what roles Anet sees each profession or elite spec as capable of or meant to perform in pvp, and balance to back up and make those roles viable. For example, I assume berserker is supposed to be a teamfighter, but it clearly is failing at succeeding in that role and thus needs buffs. An acknowledgment of what roles professions and elite specs are supposed to play along with whether or not Anet thinks that those roles are currently effective, underpowered, or overpowered would be very helpful.

This kind of specificity is what we actually need.

The stream spelled out general philosophy. Okay, so where does ANet think they are meeting their goals? Where do they see themselves falling most short? On which professions? In which game modes? Which specs are targeted to fill which roles in which modes?

Without communicating these things specifically, player feedback is a shot in the dark.

Hopefully future patch notes will also call back to the design philosophies, instead of largely being "overperforming so we nerfed" and "underperforming so we buffed".

[Forgotten Legend.9281]

i think that the one thing many people wanted to know: was how do you balance power vs condition specs? because there are some condition specs that will melt someone 100%-0% faster than power spike builds....

 

and as an elementalist main... where is our ranged DPS weapon? (reminder, that Anet classifies staff as a support wepon.) especially since we're listed as a class that prefers ranged, but all the elite specs are designed around "close range" a lot of us (not the PvP community obviously) were hoping for a longbow. there's a lot of history in GW1 with elemental arrows, and it makes sense to give elementalist longbow some of the block-hate from GW1 skills like glass arrows. that block-hate is something unique that many Open World Ele's would love. it could work in competitive modes as well against hard to kill melee classes (which are counterplay to non cc-spamming ranged classes)

[Barraind.7324]

So... that wasnt design philosophy.  Maybe you could call it an abbreviated high-level overview, but the philosophy was completely missing.

 

"We see the archtypes in PvE being DPS, Boon DPS, Healers".  Right, ok, so anyone with a pulse knows that. 

But what we dont know is:

 

How you feel boon application should be handled.  Should alac/quick be something like renegade OFA? Should it be like herald facet spamspamspamspamspam all day?  Firebrand mashing 3 skills on cooldown?  Like mirage being tied to mirage cloak staff 1 with clones? 

 

What should the difference in output be between power and condition builds? in a lower intensity power build vs a condition build with nonsensical apm and positioning requirements?  Does forced movement or forced immobility get taken into account?    Where does the line for relative fun/complexity sit with regards to damage?  Where is the line for changes being made to specs being popular but performing well below other specs that are less popular?     What is the ideal perfect point of complexity vs fun on a damage output chart?

 

Theres a lot of "design/balance philosophy" topics, and about 0 of them were even mentioned.

Edited October 28, 2022 by Barraind.7324

[soul.9651]

Firebrands and mechs stole all tickets 👮

[Awesumness.1823]

Good stream.

Confirmed a lot of what the hyper engaged community has been saying, so it's nice to get this information disseminated directly from the devs to the broader community.

Especially:

1. Importance of Alacrity and Quickness in PvE. (🔁💨)
2. Skill design tradeoffs and opportunity costs. (🎟🎟🎟)
3. Bad/wrong choices. (🪓+🩸)

[Skyroar.2974]

One thing I would like to know more is also very specific, because, as far as I know, it has only been done... once?

Skill splits within the same profession that are done regardless of game mode.

Mesmer's sword #3, which is different when on a virtuoso. I can understand why you did it, but my question is, if the new virtuoso-specific functionality is so necessary for virtuoso's sword to function, couldn't that functionality be made universal, for the sake of avoiding unnecessary complexity/accessibility? Is the old functionality so important in the older specs that it justifies having two different functionalities at the moment, even within the same game mode? I find that odd, because sword #3, in pve at least, always felt like an unsatisfying/poor/filler skill, so it is weird to me that the most exciting version is arbitrarily gated behind a single spec.

[Alakazarm.1847]

This is pretty disappointingly sparse, but maybe I was expecting something unreasonably thorough. 

I think the thing most obviously missing is some sort of commentary on the general power level of builds relative to each other and relative to past or future states of the game (not with respect to the spec itself as you alluded to with the power budget section, but specifically to other builds). As someone who thinks the game has been powercrept a little too hard, I'd like to see overall damage output reduced by some small percentage (15%ish, maybe?), but regardless of my limited perspective I'd be curious to see what you guys have to say about it, since I imagine you're very much the people through whom power levels of various specs get adjusted.

I also think that boons probably warrant a more thorough discussion. It's been discussed to death, but the smothering power of quickness is absurd, especially relative to alacrity. I think the strength of the actual effects these boons provide is worth considering, despite the fact that players consider them a given in any given team composition. Perhaps it's the case that you just don't consider solo or open-world content when talking about things like tradeoffs or "no bad decisions" or whatever, but it is legitimately impossible to make a good decision if that decision involves forgoing quickness and/or might (and fury for power builds). The other boons are good, but nowhere near as important as those two in terms of dps output, which is where you see most of these tradeoffs (% modifiers most frequently, obviously). In any area of the game where you're not playing with other people you're communicating with, every decision to forgo boons is a bad one--which is ironic considering it's how many people play. 
I dunno, I think it's worth talking about. 

Finally, and this is another point that's been discussed to death on streams and the like, the overwhelming advantage certain builds like firebrand and mechanist have in that they can sacrifice parts of their build for the sake of utility at much less overall cost to their build's output in their designated role. The fact that these professions generate quickness and alacrity off so few skills means they have so much more space to play with when it comes to taking extra stuff. They're not the only specs like this, of course, but they're the most common. 
That's a fun situation to be in--as a firebrand it feels fantastic to know your class well enough that you know how and when to sacrifice parts of your healing or damage output to bring critical CC utility. Conversely, it feels really, really kitten to play quickness herald and know that in order to provide utility you have to kitten your boon output--unless you're willing to swap your gear, but imo gear-swaps are unacceptable in situations like this because of their inaccessibility to such a huge amount of the playerbase relative to the completely consequence-less decision to switch a utility skill or trait. 
Classes like elementalist having so much of their damage output be tied to their utility skills should also be a part of this discussion.

Thank you for your efforts to improve the game, and I look forward to seeing reaper, dh, and daredevil get completely neglected in November. Fix bloody roar.

Edited October 28, 2022 by Alakazarm.1847
grammar, clarity

[Skyroar.2974]

17 minutes ago, TheGrimoire.4039 said:

You keep asking for feedback but there's YEARS worth of feedback on the Elementalist forum, and nothing.
Sad to see.

I think the problem is: as long as the attunement system exists in its current state, the ENTIRE class will always be elite-player-gated. Not everyone can or wants to play piano, working twice as hard for the same result as any other class.

But how do you change that? Maybe Anet has had ideas for elite specs where the attunement system is heavily reworked into something simpler, but my guess is that those ideas aren't/haven't been easy to implement, and that's why we keep getting elite specs that only made the complexity problem bigger, instead of smaller.

Edited October 28, 2022 by Skyroar.2974

[KrakensEmbrace.8215]

-Looks at WvW section- I guess a lot of the lackluster changes and nerfs make more sense when it appears the vast majority of havoc and roaming, as well as some group comps, aren't even incorporated in the WvW balance philosophy. 

[ArjukKagrim.6049]

Why does Mechas Shift Signet has 10 times the "power budget" than any other skill in the game? 🤔

[Gibson.4036]

Just now, ArjukKagrim.6049 said:

Why does Mechas Shift Signet has 10 times the "power budget" than any other skill in the game? 🤔

A designer didn't like toolkits, and designed Mech to make sure no one would take them.

[Infusion.7149]

Good overview.

I have concerns with how the existing skill balance is working according to the "philosophy".

* The difference between "low intensity" and literal AFK (see rifle mechanist autocast skills) as far as PVE
* Autoattacks doing more DPS at range than in melee which incentivizes range strategies in PVE or "pirateship" in WVW
---> "counterplay" should also apply to PVE when there isn't a "phase shift" , "exploiter" effect , or reflect bubble
* Elite spec weapons that don't perform at or above core spec weapons for that elite spec (see DH Longbow for the longest time compared to scepter focus , weaver sword right now, untamed hammer, soulbeast dagger against a single target)
* Utility skills that serve dual purpose yet are still top DPS picks (Purging Flames is a big one but so is Mine FIeld on holo engineer)
* Traits that make a skill do 4 or 5 things that makes it far harder for newer players to recognize and counterplay said skill
* Mantra of Solace and firebrand mantras in general in competitive modes which make it more or less unusable outside of minstrel,  or celestial
* Please think about and review how challenge mode ingame content is completed by most players especially if it requires blocks or stability to be effective. If a majority of groups are using "Advance!" on Deimos CM while "Stand Your Ground" or Inspiring Reinforcement on Samarog CM is common then probably you want to think about how future groups will complete it. If this is not an intended way to complete the content then the content should probably be tweaked according to new skill balance.
* Splitting of a core game mechanic on mirage (if you need to nerf jaunt and Infinite Horizon even if the clone ambushes have been toned back it is better than a one dodge mirage in PVP/WVW)
* Completely different functionality in PVE vs PVP/WVW: Draconic Echo and Kinetic Accelerators can probably have quickness at 0.25s instead of might in PVP/WVW and not have a completely different function since they do have effects other than boon generation. That is how Seize the Moment works is it not?
* Passives compared with active skills are not mentioned (i.e. something that happens automatically from a trait versus from a player skill)
* Years ago I recommended corrupts in PVE do CC bar damage on par with fear when there are no boons present. Is that a possibility?
* Does power budget include number of targets hit? If it does then please look at warrior mace in PVE very thoroughly. This is also applicable in WVW since any skill that only hits 1-2 targets is inherently weaker than one that hits 5.
* Can we get pet stow for druid in WVW? It's cannon fodder , borderline animal cruelty, and is "negative value".

If it's possible to do a time to kill metric ingame on the PVP/WVW golem dummy it would be nice too.

 

Edited October 28, 2022 by Infusion.7149
add pet stow for druid request

[Adotiln Urthadar.1823]

I'd be interested in knowing where you stand on previous efforts that have ended up removing tools that could have been used for balance. An example would be making might and fury generation "automatic" in group content by having basically all classes provide some passively (though this hasn't been achieved yet). In my opinion this just gives you less things to balance around, and along with things like minimising attacks that benefit from large hitboxes, or making almost all boon-provision AoEs identical size (PBAoE 360 units), it feels like it leads to a homogenisation and denies classes from having a niche.