The many roles of DPS... Ish.

Greg “Ghostcrawler” Street released a blog today, speaking specifically about how DPS roles can be constructed and designed within the confines of a game. He’s the Lead Systems Designer for World of Warcraft, but as TOR follows suit in its basic combat model (the GCD along with tanks, healers and damage dealers) the commentary is just as important for those of us running around the Vaiken Spacedock.

First of all, let’s get this straight; I hold Mr. Street personally responsible for the Azerothian malaise that has essentially killed off the game. That said, a lot of his blogs cover topics that are emotive, difficult to handle and hard to discuss. When it comes to DPS and how your specializations should be designed, he’s messing about with the vast majority of his player base.

He gets kudos just for taking it on.

Secondly, though, credit has to go where its due. PvE performance, in this author’s opinion, has never been more balanced than it is now and it’s Mr. Street who takes the plaudits for getting it there. Unfortunately, in lieu of any real analysis being possible in Star Wars as of yet, it’s hard to tell just how “balanced” the game is in its current state.

That said, the whine is strong in many.

Let’s take a look at his post, then.

The Concept

Essentially, the problem starts because there is a huge disparity when classes have three choices of talent specialization, but they don’t all have the same choice in roles. In the case of the paladin, shaman or druid, the problem doesn’t effectively exist because each spec does something vastly different. The paladin, for example, can either tank, heal or melee DPS. The shaman can heal, melee DPS or ranged DPS. The druid… You get the idea. Each specialization has a distinct “role” that utterly changes what is expected of the class when it’s in that spec.

Unfortunately, the other end of the scale has pure DPS classes that can only perform one role (DPS); and none of them can switch between ranged and melee, similar to the shaman or druid. Rogues have three choices between melee DPS while hunters, warlocks and mages have three choices between ranged DPS. Their role, effectively, never changes in a raid group. Obviously priests, DK’s and warriors fall into this category but not necessarily to the same extent.

What, then, is the point in the three specs?

When a player can only fill one role in a raid, what makes for a compelling choice?

Here are Mr. Street’s options:

1) DPS and utility is equal, you play what you like most.
2) Specs are specific for single-target, AoE or utility.
3) Each spec is roughly the same, but changes utility.
4) There are designated specs for PvE and for PvP.
5) Multiple DPS specs bite the dust in favour of hybrids.

I’m only really going to comment on things that the blog author hasn’t.

Model 1:

Despite model one looking good on paper, it’s practically impossible when you move away from a target dummy. Additionally, we live in a world where the community has decided that the only thing a DPS player should be concerned with, is DPS. That’s it. If they’re not competitive on damage, they’re not a competitive class. It completely ignores that key word, “utility”.

Sure, there is almost no encounter problem that can’t be solved with higher damage – but that doesn’t mean utility should suffer the vilification it does at present. If a class utility doesn’t buff the damage of said class, players write it off as “pointless utility that shows clueless design”.

While this could be considered a problem with raid design not offering enough scope for utility, it’s NOT a problem with class design. The community has made it thus. And we’re also talking about a community that has proven, by and large, that you should change from your favourite spec to the one that does the theoretical 1% of extra DPS.

The developers cannot win here.

Model 2:

It’s amazing that the same community showing such contentment at changing spec for a miniscule damage increase (even if only theoretically) would be so poisonously hateful if the developers chose to go down the route of model two. Imagine it.

“We’ve decided that Fury is for AoE and Arms is for single target”.

This could then be the most balanced the game has ever been, but I think we can all agree that the outcry would reach critical decibel levels. Switching from Arcane to Fire is fine for a 1% increase in damage. Switching from Arcane to Fire is not fine if it’s because Blizzard designed the encounters with that in mind.

Of course, there is another problem with this model; what about Retribution or Shadow? Classes that only have a single DPS spec would either have to be designed for both AoE and single target (thus, overpowered), or designed for only one (thus, underpowered). Buff utility? Pah. I’m a DPS player, man. I do DPS.

The developers cannot win here.

Model 3:

Personally, I think there is a lot to be said for model three but I don’t think the community would widely agree. As previously hinted, players don’t like utility and I think the root cause of that is because it’s hard to judge whether they’re being effective or not. Sure, World of Logs analysis can help to figure this out, but the game really needs to find a way of showing how effective utility is when a player doesn’t compete in raw DPS.

Again, however, raid design doesn’t appear to support this model. Most problems are solved by higher output or by getting a specific mechanic right, not by clever use of class utility. In short, encounters would have to change dramatically in order to make this compelling, something that would also run the risk of being too formulaic. If someone really likes using their utility, you can bet your bottom dollar they’ll complain about encounters that don’t encourage its use.

That said, the potential in this kind of idea is staggering. Imagine a Subtlety rogue that was designed around providing threat management. He could improve tank threat (assuming it was still an issue… It currently isn’t), help to reduce DPS threat, or misdirect it to other NPC’s. Imagine an encounter that brought in an add that had to be off-tanked, kited or killed. Your Subtlety rogue, being clever, decides to misdirect the attention of the boss to this NPC!

The developers can win here.

Model 4:

I think, to a certain extent, this would solve a magnitude of balance problems, but create a pile of other problems. The blog already covers them accurately, and I have nothing to add really.

Is ultimate balance really what players want? I’m not sure it is.

The developers can win here, but I don’t like the odds.

Model 5:

While I generally agree that sweeping changes to a mature game with millions of expectations attached to it would be too risky, I know many people who believe that a complete overhaul is exactly what the game needs at this stage in its life.

The biggest problem, however, isn’t the idea – it’s the way it’s presented. Prior to this paragraph, Mr. Street is largely sensible in his approach but he undoes all his good work here by forgetting the history of his own game, while simultaneously rooting himself in it.

Few would support the idea of putting three roles into each class. Tanking mages, healing warriors and ranged rogues don’t make sense. But this is misleading. What makes three the magic number for talent specs? Why not two? Why not one spec that concerns itself primarily with damage, while another spec concerns itself primarily with utility? Star Wars actually got this particular decision right on the money, by realising that pure classes should get the MOST utility because hybrids already have it purely by virtue of being hybrids.

The Sith Juggernaut can tank or DPS, which means his DPS spec does exactly that; damage. Compare that to the utility offered by a Sith Marauder, and you start to see why BioWare went down this road. Both classes do the same damage, roughly, but it’s the Marauder that has the most utility because it’s the Marauder that can never tank.

So, why three specs?

Most will argue that hybrids suffer under this routine because they’d have to lose a role entirely; but that’s not true. Take a look at the paladin who can spec Holy (spellcasting), Protection (melee) or Retribution (melee). We have a shaman who can spec Restoration (spellcasting), Enhancement (melee) or Elemental (spellcasting). Lastly, what about our druid friends? They’re about to play Guardian (melee), Feral Combat (melee), Balance (spellcasting) or Restoration (spellcasting).

In other words, they’re either in melee combat or they’re spellcasting; two options.

Talent trees (or systems… The Pandarian ship has sailed) could be designed around what that character is actually doing, rather than specific buffs for baseline abilities. And don’t forget, Feral Combat used to house the talents for cats AND bears – something that, though sometimes clunky, invariably worked just fine. Therefore, your Retribution and Protection paladin could be based in one tree, while the Holy paladin could be based in another. This applies across the board, and even supports talent choices from other trees if they’re well designed.

This is the problem, as I see it, with Greg Street and the current team. They’re not willing to take risks, or really think outside the box. Utility specs were fun for a lot of people (my sister, a Shadow priest, loved playing it during The Burning Crusade), yet they’re now gone because of shallow raid design that calls for three roles and three only; even hybridization has taken a hit. Yet, there are options for real innovation and they’re routinely sidelined because they’re either too difficult or go against what was seemingly set in stone aeons ago by goodness-knows-who.

So, which model is the best as far as I’m concerned?

It’s got to be a mix of 3 and 5. Rather than designing raid content that simply favours more output, deeper encounters would promote the value of utility or support specs that can grease the tactical wheels beyond raw damage. However, I’m also of the opinion that three specs per class is becoming a burden because they’re impossible to get right in a community that can’t agree with itself.

The community is largely responsible for the other models being unworkable, because they’re vociferously resistant to change that doesn’t address the more fundamental issues. The developers are responsible for not addressing said issues by virtue of meaningful change that might freshen up an overused endgame model. A model that, might I remind people, is bleeding players.