Nerf the Interesting, Buff the Boring

There’s a bit of a thread going on in the official forums about the fire mage nerf.  The gist of the thread is that it’s supposed to provide light on the subject rather than the heat of the many fire mages complaining about the nerf.  I provided my two cents, and a lot of other people did as well.  It’s worth a read.

There was one post there, however, which was extremely disturbing to me, and while I understand the notion that it’s important to provide light rather than heat, I can’t help but get a little heated about it.  On the second page of the thread, Lhivera, a very well respected blogger of things magey, said:

Honestly, I don’t see why this would be a pessimistic point of view. If 95% of Mage players were using Fire, we have to assume that a good portion of them would have preferred to be using Arcane or Frost, but didn’t feel free to do so because Fire was putting out more damage. Enabling people to switch to their preferred spec not only probably was part of the motivation to nerf Fire, it should be part of the motivation to nerf fire, because improving balance between the three specs is the only way to give people that freedom.

What we’re trying to get at in this thread is the results of the change. If they struck a decent balance, we should see fewer people playing Fire and more people playing Arcane and Frost. If the number of Mages playing Fire drops from 95% (spitballing) to 30-40%, that is not evidence of a catastrophic nerf — it’s evidence that they’ve achieved a pretty good balance, so that the number of people using each spec is reasonably similar.

If, however, they overshot the target, we may see 5% of people using Fire and the problem has just been reversed. Blizzard likely already has a pretty good idea of whether they hit the target or not based on their internal data. This thread is more intended for us as players to try to determine that for ourselves.

This represents for me a rather disturbing viewpoint: Lhivera (and probably Blizzard) evaluates balance of the mage specs not on the basis of their performance, but on the basis of the distribution of players among the specs.  He assumes (without evidence) that people would rather play frost or arcane if those specs did more damage, but chose fire because it was putting out more damage.

Here’s the thing, though.  Fire does not put out more damage.  As of 5.0, if I recall correctly, Frost was theoretically better than Fire, and Arcane was insignificantly behind Fire.  All three specs were perfectly viable and all three specs were close enough to each other such that skill at a certain spec would overwhelm any theoretical difference between the specs.

What happened with 5.1 before the hotfix?  Well, arcane was buffed.  The result was that, theoretically, arcane was bumped up to the third best dps spec in the game.  Frost was number two.  And fire was down to number seven, a little over 5% worse than frost.

I’ll say that again. As of the release of 5.1, fire was theoretically significantly worse than frost or arcane.

But in terms of actual results, fire dominated.  There is no question about that; a look at the leaders in dps for Mogu’shan Vaults is littered with fire mages, while arcane mages have but a few spots and frost mages are nowhere to be found.

Why would this be?  Well, Lhivera offers a clue, even if you have to read the post backwards to find it:

Yeah, Frost didn’t actually receive a buff — but we’ve long known (and the devs have said this as well) that the simple fact of being underplayed will seriously depress a spec’s aggregated parse numbers well below what the spec is actually capable of. Simply increasing the number and quality of players using a spec will make a huge difference in how it appears to perform.

I think it’s worth knowing the sort of numbers people are seeing Frost produce now simply because if more people are playing it, the picture of what it can do will be more accurate.

Put another way: Fire was performing better because the top players in the game were playing fire (and indeed most people were playing fire), while arcane and frost mage players were simply not as skilled. (This is, of course, supposition on Lhivera’s part, but it makes absolute sense given fire’s superior performance despite a lack of theoretical advantage.)

The thing about it is that I know for a fact top players consider SimulationCraft and other resources reliable measures of performance. They chose fire regardless of its theoretical inferiority.  Why?

Allow me to make two suggestions, one silly and one serious. People chose fire because they like the color orange. Mages seem more natural with orange things flying from their hands instead of blue and purple. The fireball animation itself is far superior to arcane’s arcane blast non-animation. The gist of it is that fire mages look like mages, while frost and arcane mages look like something else entirely, something not nearly as exciting.

Okay. I doubt that’s true for anyone other than me and a small handful of people. But it gets at a more general point. Why would people choose fire over arcane and frost if the damage is equal?  Because fire is more interesting.

Here’s how I’ve described the fire mage “rotation” versus bosses:

• If you get a Heating Up buff and Inferno Blast is off cooldown, cast Inferno Blast.
• If your bomb of choice is not active on the target, cast it; refresh it as appropriate based on the particular spell’s criteria.
• If you have the Pyroblast! buff, cast Pyroblast!
• Cast fireball.

This is a highly simplified version of the basic rotation; there are times when you will want to delay your Pyroblast cast. What this does not include is the use of the Combustion cooldown. Combustion forces you to monitor the size of your ignite buff on the target and use it strategically. (For me, typically two seconds before my raid leader activates Time Warp.  Grumble grumble.)

Let’s look at what Icy Veins has to say about Arcane:

• Use your bomb of choice.
• Cast Arcane Missiles when it procs.
• Cast Arcane Barrage when you have between four to six stacks of Arcane Charge.
• Cast Arcane Blast.

Superficially, there are the same number of spells in the rotation. However, Arcane is much simpler. Arcane Missiles, when they proc, do not disappear if not used (unlike the Heating Up debuff). In fact, you can get two stacks of Arcane Missiles procs, and the recommended strategy is to wait until you have a lot of Arcane Charge stacks or you have two stacks of Arcane Missiles to use Arcane Missiles.  Arcane Charge, unlike Pyroblast!, builds very predictably; each Arcane Blast adds one charge.

Arcane’s major cooldown, Arcane Power, is also a lot simpler. Icy Veins recommends that it be used whenever it is available (although my suspicion is that if you can line it up with heroism or other fight mechanics you may be better off).

Now let’s look at Frost:

• Use your bomb of choice.
• Cast Frozen Orb on cooldown.
• Cast Frostbolt until you have three stacks of the debuff it applies.
• Cast Freeze from your Water Elemental when it is available.
• If Brain Freeze procs, cast Frostfire Bolt.
• If you have Fingers of Frost, cast Ice Lance.
• Cast Frostbolt.

Superficially, this looks more complicated.  There are six spells here. Frozen Orb, however, has a 60 second cooldown and I’m not sure why they put it in the rotation rather than in the cooldown section. There are two procs to watch for, but like Arcane and unlike Fire, there is no significant penalty for missing the proc.  (There is a slight chance that the proc will happen again, but Fingers of Frost will stack twice, practically eliminating that risk.)

Frost’s big cooldown is Icy Veins, and like Arcane, it should be used every time it is available as soon as it is available.

What’s the gist of all this?  To me, it’s that fire has a more interesting play style.  It’s far more reactive than arcane, and more reactive than Frost.  The cooldown is more interesting to manage. It is a spec that offers greater reward for skill as compared to the other mage specs. Even if I didn’t like the color orange, I believe I would choose fire, all other things being equal, because it is more fun.

I think most people agree.  Fire was overwhelmingly more popular, despite it being roughly equal in theoretical DPS to the other specs.

I believe the Blizzard developers did an admirable job balancing the specs in 5.0.  Fire, Frost and Arcane were within a stone’s throw of each other. In 5.1 the balance was thrown off a little bit, but it was thrown off to fire’s detriment. More people should have chosen frost because it was the highest dps spec. But all three specs were viable, and the work balancing things was pretty darned good considering the difficulty of doing it.

And then came the hotfix.

With no testing, no feedback, and no warning, fire was nerfed to the ground.  Despite it being the worst spec pre-hotfix, Blizzard saw fit to cut Fire’s dps by over 10000, rendering it approximately 15k dps behind both Arcane and Frost (which are now the highest dps specs in the game).

The result of this nerf will be predictable: most mages will switch from fire to frost and arcane.

At this point I’d like to go back to a sentence from Lhivera’s post I quoted above:

If 95% of Mage players were using Fire, we have to assume that a good portion of them would have preferred to be using Arcane or Frost, but didn’t feel free to do so because Fire was putting out more damage.

No, no, no, no, NO. Absolutely not.  I respect Lhivera and his contributions to mage theorycraft but this is one of the most boneheaded things I could possibly imagine.  We know the theory. We know that in 3.3, 95% of mages chose arcane because it was the only viable raid spec. We know that in 4.2, practically all mages were arcane.  We know that Blizzard did a damned fine job going into 5.0 of balancing the specs and we know that they did roughly equivalent damage.  We know than as of 5.1, pre-hotfix, Frost and Arcane were significantly better than Fire.

When offered a choice between three specs that do roughly equivalent damage and finding that most people choose one spec, the reason is not because of that spec’s increased damage, not unless you can find a flaw with the theoretical measure of damage. (To my knowledge, other than pointing out that the top mages were all fire, no one offered a theoretical reason why the simulations were wrong.)  It must be because they preferred the spec.  To me the reason is obvious: fire is interesting and fun.  Arcane and Frost are boring.

What Blizzard has now done, instead of making boring specs more interesting, is to force people to play boring specs they’d rather not play in the name of helping their guild progress. They have done this for some arbitrary reason of wanting there to be an equal division of the specs.

I don’t know about you, but encouraging people to play specs they don’t like out of some arbitrary desire for balance in spec choice does not strike me as a good business move.

Blizzard had it right in 5.0, and in pre-hotfix 5.1, when the specs were relatively balanced, and people could choose a spec without fear that it would do significantly worse dps than a non-preferred spec.  Two days later, in a state of panic for which I have no explanation, they decided that wasn’t good enough.  They nerfed the theoretically worst spec and the strategically most interesting one to the point where it is no longer viable.

That is indefensible.