Graphs proving that build diversity and claim buff are messed up

[Riba.3271]

Build diversity across different metas

As you see, Pre-HoT and Post-HoT 2-guardian metas both follow somewhat healthy distribution but post-PoF things have gotten very very messy. Builds between Cele/soldier and berserker are marauder or mix of valkyrie + berserker

Claim buff affecting meta

Well as you see, your advantage against enemy was somewhat linear so it was easy to find a spot to have good spots to fight against enemy, win a fight at outer gate? Maybe you will lose at inner gate? If not, then maybe at a lord. Now there is arbitrary point between keeps where advantage has a superjump between objectives. The effects of claim buff are quite obvious from this and it can't be denied. Only spots equally strong groups can fight each other are at 3rd servers objectives, who will then have advantage instead.

Didn't take into account presence of the keep which would make this even worse regarding keeps, this graph applies to all objectives nevertheless

I know I know, my graphical talent with paint is immense. Unfortunately my excel license ran out.

[Infusion.7149]

You're not factoring the effect that marauder and other 4 stats (minstrel , trailblazer) had.

Also remember karka potions which are still not removed from WvW.

[Riba.3271]

@Infusion.7149 said:You're not factoring the effect that marauder and other 4 stats (minstrel) had.

Also remember karka potions which are still not removed from WvW.

Minstrel and marauder existed during HoT meta also. And marauder and valkyrie+bersk builds are between berserker and soldier

[Strider Pj.2193]

From a simplistic view we went from 6 builds in the meta to 10 builds in the meta. (Assuming what you posted is accurate)

Yes, the middle of the road builds have phased out, but many of those are in the roaming scene.

Overall diversity is up.

The game has moved towards min maxing. Other builds in the middle are viable, just not considered meta. Not just in WvW, but in raids, fractals etc.

Until you eliminate the viability of full support builds (which wouldn’t be healthy in a cooperative mode) you are going to see this continue to evolve.

Now, not coming from a ranger player, it’s concerning that In none of these threeMeta snapshots is Ranger included.

[Infusion.7149]

@Threather.9354 said:

@Infusion.7149 said:You're not factoring the effect that marauder and other 4 stats (minstrel) had.

Also remember karka potions which are still not removed from WvW.

Minstrel and marauder existed during HoT meta also. And marauder and valkyrie+bersk builds are between berserker and soldier

Economically minstrel was rarer. Raids were introduced with HoT (so legendary armor stat swap was not that simple) , winterberries only give you one trinket + back/ring/accessory (November 21, 2016), and I remember each pearl was 7g each.

[Stand The Wall.6987]

it would be nice if balanced builds were viable again. for cele, i wouldn't add in condi/ boon duration but a bit more power. the game really needs much better 4 stat combos that support balanced builds, such as +power (or condi) +heal - precise -vital (or tough). this would only work for classes that have + precise and + feroc traits. it would be awesome if we could just have gear that added points into a shared pool that we could then assign to each stat as we saw fit (with limits ofc).

[Riba.3271]

@Stand The Wall.6987 said:it would be nice if balanced builds were viable again. for cele, i wouldn't add in condi/ boon duration buta bit more power. the game really needs much better 4 stat combos that support balanced builds, such as +power (or condi) +heal - precise -vital (or tough). it would be awesome if we could just have gear that added points into a shared pool that we could then assign to each stat as we saw fit (with limits ofc).

Issue is that they added legendary gear into WvW. I would have liked them to implement double, triple or quadruple amulet system where only half your stats can be minstrel/trailblazer/dire etc and another half would have to be cele or 3 offensive stats + 1 defensive.

Or not add minstrel in the system at all

[ArchonWing.9480]

Umm, no.

Cutting edge meta builds are based off of open field combat, skrims, and gvgs, not fighting in structures.

Yes the claim buff is dumb, but this is a lot of conjecture. The line is much more reasonably drawn with the introduction of Firebrand because that is when guardians went full minstrel. And to drive home this point, all dps guardian builds today are core guardian or dragonhunter.

Warrior going minstrel has much to do with the damage potential of hammer being sacked, destruction of defense, and rework of tactics.

Engi going minstrel has much more to do with the addition of purity of purpose and med kit being viable. There was almost no role for Engi before that besides stealth.

The idea that any of these build changes relied on keep buff being a major factor, is well, to put it nicely, misguided.

[Riba.3271]

@ArchonWing.9480 said:Umm, no.

Cutting edge meta builds are based off of open field combat, skrims, and gvgs, not fighting in structures.

Yes the claim buff is dumb, but this is a lot of conjecture. The line is much more reasonably drawn with the introduction of Firebrand because that is when guardians went full minstrel. And to drive home this point, all dps guardian builds today are core guardian or dragonhunter.

Warrior going minstrel has much to do with the damage potential of hammer being sacked, destruction of defense, and rework of tactics.

Engi going minstrel has much more to do with the addition of purity of purpose and med kit being viable. There was almost no role for Engi before that besides stealth.

The idea that any of these build changes relied on keep buff being a major factor, is well, to put it nicely, misguided.

Point of the post wasn't that the 2 things aren't linked. I was just combining 2 graphs, both of which I had idea about today to reach the non believers, into 1 post. So the linking factor is the fact that there is a graph.

[Ovalkvadratcylinder.9365]

Wat?

I mean you said it yourself. This is entirely based on the assumption that there are two equally strong blobs. I have never in seen that happen since launch.And I still don't know what thread is about... Do you wanna talka bout build diversity or the claim buff? Those seem like two subjects in need of two separate threads.And maybe you can make a thread about how shit the pirateship meta was pre-hot aswell? I mean the changes to stab prob had a bigger impact on the game mode than anything you've mentioned so far.

I guess another quesiton is; how do you link the introduciton of the claimbuff to change of meta? It seems like a leap too far. If anything you could prob argue for a change in player behaviour around keeps and towers and evne then you havent provided anything substantial really, but, not once have I and the guilds/theory crafters I talked to taken into acount the claimbuff when figuring out whats most effective in a fight or ameta changeing build for a certain class.

[subversiontwo.7501]

@Ovalkvadratcylinder.9365 said:And I still don't know what thread is about...I'm getting too tired to read through the posts properly, I will check back tomorrow, but at first glimpse I think the thread is about alot of middle-ground stat-distribution falling short at larger scale combat. If that's the case it's also both true and a pretty interesting discussion. There are plenty of things it could branch out into talking about.

[Justine.6351]

Yeah doesn't really make sense to link build diversity to claim buff. Think balance changes are what polarized build stats. Was fairly disappointed to check meta builds and found more minstrels than anything else. Can't get much more boring then that.

[kamikharzeeh.8016]

it feels like half of the people in the format play random stuff anyways, bunch play spvp sets and some maybe cele/solider etc stuff.

at least on our server the top meta players may be like 700% as effective as the kittens, but also way less numerous. we had last MU sth like about 90% of the serverkills done by like 25 people, roughly.

[Riba.3271]

@Ovalkvadratcylinder.9365 said:Wat?

I mean you said it yourself. This is entirely based on the assumption that there are two equally strong blobs. I have never in seen that happen since launch.And I still don't know what thread is about... Do you wanna talka bout build diversity or the claim buff? Those seem like two subjects in need of two separate threads.And maybe you can make a thread about how kitten the pirateship meta was pre-hot aswell? I mean the changes to stab prob had a bigger impact on the game mode than anything you've mentioned so far.

I guess another quesiton is; how do you link the introduciton of the claimbuff to change of meta? It seems like a leap too far. If anything you could prob argue for a change in player behaviour around keeps and towers and evne then you havent provided anything substantial really, but, not once have I and the guilds/theory crafters I talked to taken into acount the claimbuff when figuring out whats most effective in a fight or ameta changeing build for a certain class.

Tbf I do agree that maybe claim buff affecting meta wasn't best way to phrase it as it doesn't have much to do with builds per se but rather how people act and how little competitive fights there are. Win a fight at your objective? Go to enemy objective and get crushed. So people and guilds just tend to defend.

Point was that if your teams are closely matched, in old meta you could go little bit closer to enemy spawn or push their tower and theyd have little bit easier time leading into better fights. But instead there is this massive jump between any objectives where you lose 400 stats and 25% MS and enemy gains those. So if you're equally skilled groups that can have nice fights, guild or blob, claim buff will often be the deciding factor within the fight, and even if you manage to defend an objective often you're stuck defending as you don't stand a chance against enemy at theirs. This is completely different from old meta where group on the offensive swapped with one or two good players joining the map. These days if you're defending side taking more than 1 push to fend off the enemy on the map then you're going to defending side always and there is no hope out from there. And if your teams are equally matched, even going to easiest enemy tower on the map usually leads into crushing defeat so the logical solution for both sides is just to defend.

This leads into boring meta where: You're either defender server 90% of the time unless you're willing to get one pushed and completely reliant on enemy coming to you to get content, or you're attacking server that has veterans with better builds and ton more experience than their opposition farming pugs like its a lab farm. There is no such thing as being equally strong to enemy in this meta as it just leads to stalemates, thus bandvagoning to super strong servers became only way out from being reliant on enemy hitting your stuff to get fights and not living in constant stalemate against enemy servers.

TLDR; I chose my words poorly as claim buff didn't affect build meta per se but made the game so that near equally matched enemies end up in a stalemate instead of having ability to choose hitting easier enemy objectives for more thrilling fights. And for one to play offensive they have to bandvagon to a server already full of veterans

[Aeolus.3615]

IMO for some reason they are called specializations??????You are only thinking on the class gimmikc way to play the classes, some classes are quite decent with celestial even after hot or after pof, it isnt just the gimmick to go.

Still the issue is more on the class design itself than balance, maybe some classses need more nerfs in one side to receive more buffs in the other area???

Scourge and DH are still total disaster designed classes, holo still a beast on small gameplay and can be a high damage burster and very sturdy at the same time.

[Ovalkvadratcylinder.9365]

@Threather.9354 said:

@Ovalkvadratcylinder.9365 said:Wat?

I mean you said it yourself. This is entirely based on the assumption that there are two equally strong blobs. I have never in seen that happen since launch.And I still don't know what thread is about... Do you wanna talka bout build diversity or the claim buff? Those seem like two subjects in need of two separate threads.And maybe you can make a thread about how kitten the pirateship meta was pre-hot aswell? I mean the changes to stab prob had a bigger impact on the game mode than anything you've mentioned so far.

I guess another quesiton is; how do you link the introduciton of the claimbuff to change of meta? It seems like a leap too far. If anything you could prob argue for a change in player behaviour around keeps and towers and evne then you havent provided anything substantial really, but, not once have I and the guilds/theory crafters I talked to taken into acount the claimbuff when figuring out whats most effective in a fight or ameta changeing build for a certain class.

So if you're equally skilled groups that can have nice fights, guild or blob, claim buff will often be the deciding factor within the fight,

See this is where I think you're just specualting. Ive had countless fights with zergs fighting within the buff where the win/loss ratio is more or less even. There are too many factors to take into account to what wins a fight or not to just say that a flat 100 stats is THE deciding factor.What about other factors like a commanders willingess to test the boundaries, or squad palyers slacking that one fight. I mean yoi dont need a zerg scenario to figure out that its mroe comlicated.I would probably easier to make the correlation when talking about a duel. But when you're talking about 50 players or more doing something stuff tends to get messy. Not every1 runs the same builds, not every1 has full ascended gear, not every1 has the same ping or fps, not ever1 has the same alot of things.On paper 100 vitality seems like alot but in reality its not that much of a change.

[subversiontwo.7501]

A couple of pointers:

1. I guess there is some merit to listing focus parties for vanilla as there were far more guild groups per pickup back then but it does muddle the overview a bit since I can't remember many focus parties in pickup in vanilla and if you consider guild-groups there's more variety in the HoT and PoF stuff too.
2. Don't forget the destroyer Warriors in vanilla
3. If the PoF chart is to be a full expansion representation keep in mind that the Minstrel SB is something pretty new and it's also questionable how much of a support it is even if it is running Minstrel stats
4. Finally, and perhaps most importantly, keep in mind that this like everything else is very much down to general behaviour patterns and that people less and less play what can work to more and more just do whatever someone tells them to do for convenience reasons - that easily has as much to do with the profileration of minstrel+zerk extremes as anything else. In fact, there are plenty of traits today that would suggest using more middle-ground stats that simply isn't being used as we just coast through leveraging numbers and no one really bothers to look into whether its warranted to throw some half-tank onto something that either get onepushed or sits high on Arc. To cut to the chase with my point: If there were more guild groups (or people playing in groups of friends in general) there would likely be more variety and less just going through the motions and doing what you're being told. There's an interesting and not immidiately appearant overlap between "population balance" and "meta" in that.

[WolfHeart.1256]

@Strider Pj.2193 said:Now, not coming from a ranger player, it’s concerning that In none of these threeMeta snapshots is Ranger included.

Ain't that the dayum truth.

[JusticeRetroHunter.7684]

@"Strider Pj.2193" said:Until you eliminate the viability of full support builds (which wouldn’t be healthy in a cooperative mode)

uhh no. Specialization is the hallmark of diverse systems. the fact that support builds appear in a game which at first was nearly nothing but DPS, shows that the emergent behavior of agents has evolved over time to become more complex.

From a simplistic view we went from 6 builds in the meta to 10 builds in the meta. (Assuming what you posted is accurate)...Overall diversity is up.

What he posted isn't exactly scientifically accurate, but it's on the right track. What it doesn't have is a a control group (a constant, non variable) that leverages how much the system has changed from one state to the other state. like you point out we go from 6 builds to 10...but we also went from 9 specs to 27 specs, so even though the amount of builds went up, the ratio between the builds being used and the total of them has gone down.

Imma provide a link to a comment where one can actually calculate diversity by using what's called Simpson's Index, which measures how diverse and how dominant species are in a system...species in this case are buildshttps://en-forum.guildwars2.com/discussion/comment/1303283#Comment_1303283

Even if the OP didn't follow the correct procedure, his graph wouldn't be too far from the reality, that net diversity in WvW is stagnant

[Strider Pj.2193]

@JusticeRetroHunter.7684

My point likely wasn’t well stated. The movement from Soldiers and Celestial builds was made possible by the creation of full support roles pushing players towards full damage, or full support. Damage was already there, as berserker was common even from the start.

If full support roles were removed, you would likely see people take builds with more survivability built in, i.e. more toughness/vitality (although not specifically together)

As to your second point, not only do we have more builds (and yes, a lower percentage of the overall potential specs) but we have more classes. Out of 9, 6 are heavily viable.

Previously, out of 8 classes 4-5 were viable as, for many non GvG groups, focus groups didn’t exist. (If they did they were outside of the margin)

With 27 specs, we aren’t going to see 14 classes in the Zerg meta. Should we? Yes. Will we? Well, given the balance cadence and history, no. We are more likely to see it drop from 10 -> 8 with new E-specs.

But maybe they’ll surprise us.

[JusticeRetroHunter.7684]

@"Strider Pj.2193" said:

With 27 specs, we aren’t going to see 14 classes in the Zerg meta. Should we? Yes. Will we? Well, given the balance cadence and history, no. We are more likely to see it drop from 10 -> 8 with new E-specs.

But maybe they’ll surprise us.

Ideally we should see 27 specs in a group, and again even more ideally we should see different kinds of groups that mix and match due to specialization ei: Groups with 100% of the same spec, vsing other groups that are also 100% of a different spec, or any combination of specs that allow for specialization of a group in it's totality in relation to other groups. This would be closest level we could achieve in terms of "excellent" diversity. Realistically it's never perfect but its definitely not impossible...there are pages of explanation as to why this is the case for gw2... imo, much of it has to do with huge variety of things, including poor design of fundamental systems in the game, and poor application of balance due to lack of knowing how those systems truly work in their fundamentality. But that's just my opinion of course.

[God.2708]

Feel like the graph becomes more interesting if you add a Y-axis with offensive/defensive support on it and change the 'support - damage' line to 'heals - damage'

[Len.1879]

What data are you going off of? Did you really just pour your gut feeling into low res graphs without proper axis labeling and call it "proof" because it makes it look "sciency"?

[Jugglemonkey.8741]

1) These graphs do not prove anything about build diversity or claim buff because that requires data, both qualitative (what is the definition of meta, and over what time period are we comparing?) and quantitative (how many people play these builds?) as opposed to the "class A runs X, class B runs Y" logic you're using. These graphs are illustrations, not proof.

2) The graphs appear to contain errors, at least in the case of thief. For the majority of HoT rev and warrior ran full zerker gear in zergs, but thieves were running marauder/zerk or valk/zerk mixes, meaning thief should be to the left of rev and warrior on that graph instead of where it is now. They also don't show the shifts in metas caused by balance patches over the course of each expansion; the meta can change dramatically with a single patch (see warriors swapping from zerk to minstrel because of the CC damage nerf in Feb), and you will often find players running builds from the previous meta some months after the patch. You would be better off comparing the meta before and after a single change rather than over a time scale of 2-3 years per interval.

3) It is difficult to draw valid comparisons between the graphs when you are changing what professions and elite specs are included. In order to track the change of something over time, you must include the same things in all three graphs otherwise any conclusions will be less useful than you'd like. A more useful thing to track might have been meta build by profession for each period, and to have a series of graphs for each profession.

4) Your line of best fit for pre HoT and post HoT are clearly inaccurate. You appear to be muddling the shape of the graph by using the Y axis to show both the prevalence of each class in the average WvW group and the number of builds that fulfill the criteria on the X axis, leading your peak to be drawn at the point where the guardians & necros sit in the post HoT graph rather than to the right of the line where it clearly ought to be. The prevalence of an individual meta build is irrelevant when considering build diversity, so if you simply had "number of meta builds" on the Y axis, both the pre and post HoT curves would have the same general shape with a concentration of points on the right of the graph causing a peak there, with a longer, more drawn out tail to the left side in the post HoT graph caused by the two support builds being further from the central line, rather than the flattened middle peak you have going on. This, along with the dramatic change to the post PoF line, suggests that the shift between vanilla and HoT was not as impactful as the shift from HoT to PoF. This would in turn suggest that the claim buff is not the cause of any significant shift in the meta given it was introduced with HoT, but it is the introduction of full supports that has done the damage, as said by others in the thread.

Edit: Thinking about it, you do actually have a point in a way, in that keep buff becomes more significant in a min/max meta than it does a more balanced meta. Still, it is a consequence of the change in game design changing the meta, not a cause of the meta changing in and of itself.

[Cyninja.2954]

@"Strider Pj.2193" said:The game has moved towards min maxing. Other builds in the middle are viable, just not considered meta. Not just in WvW, but in raids, fractals etc.

This right here is what players are experiencing and what mostly is refereed to when people mention a drop in build diversity.

Build diversity decreased at the top end because of specialization. Specialization both in terms of:

• more specialized elite specsExample:Firebrand is a stability machine. Past guardian builds did not decrease in their ability to provide stability, they decreased in comparison to Firebrand.Spellbreaker is superb in stripping boons. While the overall classes with noteworthy boon removal increased from 2 to 3 with HoT and from 3 to 4 with PoF (necromancer, mesmer, revenant, spellbreaker), there is now a ranking of how valuable the classes can provide this mechanic, where as before most simply could not.Purity of Purpose completely changed how one can deal with conditions (and might be even to strong or unhealthy for the mode, any conversion might be because it hard counters condition builds).

• more specialized itemization mirrors in essence a more traditional trinity approach in which distribution of roles in favor of overall stronger performance is usedExample:Minstrel has pretty much all the stats a support would ever need in a competitive environment.Trailblazer is the same for smaller scale, maximum survival with primary damage stats for condition damage.Marauder, while technically inferior as pure damage set to berserker, offsets its 4 stat distribution with over 300 additional stat points. This is actually a move away from specialization with berserker but again is cause for prolonged fights.

Outside of those 2 issues, there are core design limitations in place which cause segregation:

• stability is a necessity. The access to this boon is heavily in favor of Firebrand both in terms of ease of use as well as overall availability. So far that most other classes which can provide reliable stability are relegated to backup stability at best.
• boon removal is a necessity. This is similar to stability in that the highest ease of use and availability is concentrated on 2 classes (spellbreaker and necromancer), while the other 2 classes are either to weak on their own (mesmer) or forced into a specific damage type which is undesirable (revenant for condition)
• boon application is by now covered by support classes which fulfill other roles, which makes certain past interactions less important (field blasting) or changing how this is approached (overall stealth blasting primarily now)

All of this causes for a more focused and limited "meta" setup, but actually causes for far more flexibility in the mid segment (good but not meta builds). This is mostly reflected in more damage builds being viable once the core support and boon removal is covered.

TL;DR:The fact that specialization in multiple areas (class, spec, itemization choice) based on some of the foundations which are needed is more efficient than not specializing is the cause of less "meta" builds. At the same time, once those foundations are covered, a lot more classes are "viable" in the mid field.

There are some ways to approach this:A. offer more flexibility in foundation classes (similar to support tempest versus support engineer, while not identical still interchangeable to some degree) which would still cause a specialization meta versus all other classes (if you can't cleanse, stab, boonrip, etc. you are doomed to be a dps) but make not support more interchangeable (as is now)B. reduce effectiveness of specialization in all areas. This would in theory make more classes "meta" again simply by shaving away the top end, but cause potential shrinkage in the midfield because foundation abilities and effects are not provided by a small subset of the group and now need to be provided by more classes (to consider here, pretty much every class has some type of viable build atm even if not meta, this was not the case in the past)