Though I am loathe to do it, let's talk for a moment about statistics. The disclaimer here, as always, is that when it comes to math I am functionally handicapped. Numbers confuse me, but I'm plucky and optimistic and I don't let my disabilities keep me down. I've seen enough Hollywood movies to know that by the end of this column, I will triumph over adversity and end up winning the heart of a chick that's way too hot for me.
There are sites like World of Logs and RaidBots that compile raid DPS data into tables that show which specs do the most average damage in the game's current end raid. They take a cross-section of the raiders in the game's top guilds and log their performance over a number of fights, presenting us with a gauge of who does what kind of damage to whom, and whic specs are best on which fights. This is helpful data. I'm glad we have it.
The problem it creates, though, is that we tend to accept it as infallible. Sites like these do not represent every raider out there. They only represent the guilds who choose to utilize their specific stat-gathering client. That's already a small fraction of the actual raiders out there. This isn't a massive sample size.
Let's take a look at the Baleroc fight as an example. It's a low-movement fight, one where mages don't have to worry too much about interrupting their most potent rotations with many odd mechanics or lengthy movement phases. It should provide us with a good baseline to work from.
I'm looking at the latest numbers from the normal 25-man version of the fight, which you can see here. The top mage spec is arcane, with a median DPS of 25,561. The next mage spec on the list is frost with 25,349, and fire brings up the rear with 22,718. So from that, it looks like arcane and frost have very similar DPS potential, but fire lags well behind. Blizzard seems to recognize that fire is lacking and is addressing the disparity as we speak with the mage changes on the patch 4.3 PTR. So let's focus on frost and arcane.
Here's the part I find deeply disturbing: The number of samples for arcane (or the number of arcane mages they got data from) is 6,142. The number of samples for frost is 115. Let me emphasize that:
Arcane samples: 6,142
Frost samples: 115
Can anyone explain that disparity to me? In what world does a 200 DPS difference justify such a massively disproportionate sample size? And here's the other thing: No other class has the same representation issues. Arcane has the most samples of any class/spec, by a crazy big margin. The next most represented spec is marksmanship hunters at 4,925. By contrast, frost is easily the least represented spec, behind beast mastery hunters at 174.
What that tells me is that a much larger percentage of mages are playing our top spec and a much smaller percentage of mages are playing our perceived bottom spec than any other class in the game. Some of you will say that just means we're better as a class at identifying our best spec and then gravitating towards it. I say it means we're better not just at turning people into sheep, but also at being sheep. We conform. We self-fulfill our own prophesies. As a mage community, I firmly believe this is our biggest flaw.
So, the question becomes: Does reality dictate our perception, or do we let our perception dictate reality. The argument generally goes like this:
If arcane is better, we should all be playing arcane.
But that argument only floats if you can prove, demonstrably, that arcane really is best by a sizable margin, at all times, in the hands of every player, in every situation. And of course you can't do that. Nobody can. Blizzard generally keeps the specs close enough, potential-wise, that the margins simply aren't wide enough to assume that one spec is always best, every time. Is arcane going to do more DPS most of the time, in the hands of a majority of players, on most fights? Probably. But that assumption is simply never going to be iron-clad. I don't care what your raid-leader is telling you. I'm telling you his logic is faulty. You can tell him I said so. He can meet me after school behind the gym. I will cut him.
Having said that, there are certainly numbers out there to support the prevailing idea that arcane is best. In other fights, where mechanics don't necessarily favor frost's turret-casting needs, arcane's advantage is clearer. On Ragnaros, the median DPS gap between frost and arcane is about a thousand. On Shannox, the gap is almost 400. On Staghelm, frost is actually up by a little over a hundred points. But in every case, the sample size for frost is so pathetically small and the sample size for arcane is so ridiculously inflated, it kills the credibility of all of those numbers.
I'm convinced that the reality is somewhere else entirely. What's happening in the mage community is the same thing that happened in my elementary school in the 80's. Acid-washed jeans were cool. They made you cool when you wore them. If your jeans were of the regular-washed variety, with no evidence of any acid being involved in the washing thereof, you were not as cool. This was an accepted fact. Remember that mulletted pre-teen in the header picture? He was the accepted personification of rad. On the playground, that kid beat you up and took your lunch money, scored the touchdown and made out under the bleachers with the one chick in your class who had gotten her boobs.
So what did everybody else do? They got their parents to take them to Montgomery Ward and buy them acid-washed jeans and neon-colored Jordache shirts and oversized sunglasses and LA Gear sneakers. Because everybody knew that was what made you cool. It's been a while since I was 12, but I'm pretty sure none of those things are true today. No matter how much I wish they were. Teenagers today worry me. I mean, pull up your pants, guys? OK? Please? I don't want to see your boxers, and I'm scared that when the zombie apocalypse comes, you won't be able to outrun the ravenous undead with your pants belted around your mid-thigh.
My point is this: Just because a large percentage of the population accepts that a thing is good doesn't actually make it good. That's trite, and I've said it before, but it's true. As a mage community, we have made arcane better than it actually is through our overwhelming perception that it is better.
The reality, I believe, is a little bit different.
Arcane is generally better. More people have will produce higher numbers in more fights with arcane than with frost. The argument can be made that arcane brings a more frequently useful raid buff to the table. But is it the absolute best spec?
Frost is very close. In some cases, better. Once the 4.3 buffs to fire go live, I believe fire will become competitive again, also. The bottom line?
If you are skilled at playing a frost mage -- and after the patch a fire mage -- and your abilities meet the needs of the fight and the needs of your raid, the numbers are close enough that no one should ever be able to tell you that you should be playing arcane instead. No one.
Any more than anyone should be able to tell you you should be wearing acid-washed jeans.
Every week, Arcane Brilliance teleports you inside the wonderful world of mages and then hurls a Fireball in your face. Start out with our recent beginner's guide to being a mage, then check out our three-part State of the Mage columns on arcane, fire and frost. Don't forget to look at some of the addons your mage should probably be using.