Yesterday’s post about SV’s dps generated a lot of conversation, and let me say up front how pleased I am that so little of it devolved into name-calling. We hunters truly have a fantastic community that we can discuss and disagree with logic and questioning and discussion of data validity rather than insults (though I was called a panda bleeper in an email from what I suspect is actually a DK).
For the purpose of summation, here’s the short version of the post: Dawn told us that the top guilds are actually hunter (and warlock) stacking because their dps is so good. This matched what I was seeing from SV dps in-game and from browsing raid parses. I suggested that a nerf was increasingly likely, and that I’d go with a gentle downward push of 5% or so, as well as a gentle upward push of perhaps a bit more to BM and MM to make them line up.
Many people agreed, and many people disagreed, and the split was fairly even, leaning a bit in the disagreement column.
I love you agreers dearly, but this post is dedicated to the dissenters, who made some good points that deserve to be addressed.
In the disagreement camp we had several different arguments, some of them anecdotal, some of them quite valid. I’m going to address a few of those, but first I’d like to relate the story of my absolute favorite dissenter. This should be a Frostheim story corner, really, but I’ll try to shorten it since this is probably going to be a long post without it:
I logged in to grind through my dailies. I was told by a guildie that someone named Mealy (not his real name) was looking for me. I had no idea who this was, but I was soon to learn. That he hunted down and pestered my guildies was probably the first sign of bad things to come.
So Mealy logs on and sends me a tell, starting with something like “I have a problem with you. Why are you saying SV is OP? It’s not!”
He then proceeded to tell me over the next half hour various stories of how he was outdpsed by this class and that class and this other class, and that even far worse-geared players of various classes were doing nearly the same dps as he was. I think he realized how this sounded, and then pointed out that these were in heroic runs where he didn’t make a single mistake in rotation, movement, anything. He was playing as good as you can in his gear, and, incidentally, so were the other classes he grouped with.
I finished off my cooking daily (forgot about that one and only just started them) and headed out to Deepholm, continuing to answer his tells between killing and questing.
I pointed out that anecdotal evidence isn’t conclusive in these things, and that you’d need vast quantities of that kind of info before you could come to a conclusion, and that heroics are a pretty poor metric anyway. He quickly retorted that he would send me three screenshots of damage meters to prove it.
I explained again how this was anecdotal — not that I didn’t believe he was out-dpsed — and he informed me that he was a biology student and he knew all about data. He then rounded things out by complaining about my bad-talking Camouflage and then telling me how awesome of an ability camo was.
Later that night I got an email from him with his screenshots. I replied, again pointing out that this was anecdotal evidence and failed to prove anything. He told me it was scientific data, that I spelled a word wrong and that I pronounce melee wrong on the podcast (I often do, it’s true) and then made the aforementioned reference to my carnal knowledge of the majestic and sexy panda.
In another world — or perhaps in another mood — this guy would have been annoying or even creepy, but I found it to be one of the most entertaining complaints I’ve ever had, and it lightened my heart. Euripides scolded me for not getting screenshots, which would have been gold, truly.
At any rate, back to the business at hand.
SV does not need a nerf
There were a lot of different arguments for why SV is not overpowered and certainly doesn’t need any kind of nerf, and I’m going to address a bunch of them. If you made a particularly salient point in the comments or an email that I’m not bringing up, it’s not because I’m deliberately dismissing you, but just because it’s not standing out in my memory at 3am.
You can’t base calls for nerf on a healer’s perspective:
Firstly, be nice. We can ignore the fact that Dawn’s a healer here — what she is for this discussion is a progression raider who keeps track of what the very top raid guilds are doing. Much, much closer track than I do. Further she has friends in some top guilds. So much like many of you believe me when I report hunter news, likewise I believe Dawn when she reports news from the very top tier of raiding.
While what the top raid guilds do isn’t conclusive in itself — after all they could be doing the wrong thing — we often see the next rung of raid guilds (the server-first rather than world firsts) emulating them in the following weeks. This snowballs on down through the competitive raid guilds until it becomes “common knowledge” on the forums and the QQ machine shifts into overdrive and flips the nitro switch. This is why I say top guild class stacking is one of the first signs of imminent nerfs.
And for reference, there is no certain number of a class that fits the definition of class stacking. It’s stacking when the guild pushes to get as many of that class as possible, having anyone with a raid-viable alt in that class play it.
SV is not OP, it’s just a scaling issue:
There were a variety of scaling comments and emails. There are two theories: the first is that hunters are just high dps at low gear levels, and once we get raid gear the other classes will catch up. The other is that MM scales better with gear than other specs and will be one top just from gear upgrades in a tier or two.
I can’t say for certain if this is true or not — after all, I don’t know how other classes scale with their gear, so I can’t compare.
However, I’m very skeptical of this argument because I suspect it’s a mis-application of history. Back in Wrath when the giant BM nerf hit many of us immediately felt it was a massive over-nerf (and it was) and that part of the picture was skewed because of hunter scaling. The key part was that hunter pets did not scale linearly with our gear — attack power was the only dps stat they inherited from us. So as our gear improved, the pet portion of our dps didn’t scale linearly with the gear.
This is no longer the case. Our pets now scale with all of our dps stats. In theory with insane crit levels our pet scaling could drop off when they get basic attack capped, but I don’t think we’ll ever have crit levels that high.
On the MM scaling end, it’s also worth mentioning that the crazy non-linear MM scaling was mostly because of armor penetration, which is gone, and somewhat from base weapon scaling, which has been greatly reduced with base weapon damage changes. Sure, MM might get more dps from mastery and SV might get more dps from agility, but the amount more scales linearly, not exponentially, so the spec scaling should be mostly linear.
It’s certainly possible that there might be some subtler non-linear scaling things going on that I haven’t noticed or thought of, but I haven’t seen anyone suggest any.
It’s too soon to say:
This may be. It’s entirely possible that for whatever reason hunters are out-dpsing most classes on most fights right now, but in three weeks the other classes will pull up even with SV. I can only say what I’m seeing right now.
You didn’t show any data:
Here is my favorite argument, and the one that I personally think is certainly the most pithy. Gets right to the core of it. Arguments based on data! I love you guys! Some people even linked to a raid parse or two to show that hunters weren’t on top.
It’s absolutely true that I did not show data. In point of fact, I almost never do here on the site, mostly because it would triple the length of every post and make them horribly unreadable. No one want to see sausage being made (and I’ve seen that, by the way. It was nasty.)
But generally if you disagree and show valid opposing data, I’ll dig up my info and we can talk. If you just shout “no that’s wrong” we aren’t going to have much of a conversation.
Like all good science, my thought process on SV started with observation, from which I formed a hypothesis and then went out to collect data to prove or disprove that hypothesis. In this case the observation was that I could out-dps anyone similarly geared (or better geared), and do so even while making a good number of mistakes (unlike our bud Mealy, who didn’t make any mistakes at all in his runs).
After pondering that for a bit and watching it happen more and more, I popped over to World of Logs and poked around. There I saw that hunters were going pretty crazy on the dps of the top parses. Survival hunters, every one, of course. While I love to see my peeps owning the charts, the representation was clearly too high.
Now, this wasn’t a hard numbers thing. It was more like looking at a panda and knowing it weighs more than me. I don’t know how much it actually weighs in pounds or kilograms, but it’s obvious that it’s a lot more than I weigh. It was also clear that warlocks and DKs were pretty skewed as well.
But, I did not gather concrete disprovable data, so I set out tonight to do just that so you can see what I’m talking about.
SV comparative dps data
When it comes to determining how much dps a class or spec is capable of actually doing, the best source of information we have is thousands of raid parses. Where any one or five or fifteen are meaningless, if we can get a large enough sample all those parses suddenly become statistically significant and the single best data source we can possibly have.
There are a lot of ways at looking at this kind of data. This past spring I poured through hundreds of thousands of lines of tens of thousands of raid parses kindly provided by a friendly contact at World of Logs to determine the actual average dps each spec was doing in ICC. It took nearly 10 hours to parse through the raw data, and even if I was willing to do that again that wouldn’t really show what we’re looking for here.
What we really want to know is what the class is capable of when played with great skill, compared to what other classes are capable of when played with great skill. One of the best ways to look at this is just to look at the top dps reports over at World of Logs, and happily the site will show this to you without hours of slogging through raw numbers.
Now every fight is different, and some boss fights favor melee, or ranged, or one class or another. So what I did was look at the top 40 dps results for every boss fight in the entire first tier of raid content. I looked at both the 10-man and 25-man normal-mode raids, and then made an entry in a spreadsheet of how many of those top 40 dps results were hunters.
As a note I did some looking at wider and narrower ranges than the top 40 in a few samples, and the proportional representation stayed about the same on average. 40 was an easy number just because that’s what WoL displays per page.
Graphs are pretty!
Let’s say, hypothetically, that I took all this information and told excel to make a pie chart out of it (I did). Now let’s say hypothetically that this is the pie chart (it’s not – this is mostly made-up):
Imagine for a moment that this is actually what the top dps spread was for each class. It’s not, but seriously imagine it. Rogues and warlocks combine to hold over 50% of all of the top dps spots, as clearly the best dps classes for the first tier of raiding.
Sure it’s possible that more players play those classes, that pure dps classes have greater representation — it’s possible, we don’t know for certain. But what we do know is that is a huge difference in rogue and warlock representation compared to all the classes — clearly a larger difference than class representation alone.
Now let’s be honest with each other — if this chart was real both you and I would be calling for ‘lock and rogue nerfs. They clearly seem to be overpowered based on what all the top players are doing in raids right now.
Of course this chart is not accurate — I did not collect the exact representation of every single class (though I did spot check them, but mostly eyeballed ‘em). In actuality death knights also had a very high representation, so there were three classes eating up most of the chart (and a larger chunk of it) rather than just the two.
But here’s the thing: The yellow slice is precisely accurate, only it’s not rogues, it’s hunters.
Hunters hold about one third of all the top dps benchmarks in currently available raids (32% when I checked), more than any other class, and of course they’re all survival hunters. Hunters and warlocks combine to own over half of all of the top dps results in raids. Add in DKs and that’s over 75% of all the parses — the other seven classes get to split up the remaining 25%.
It’s certainly possible that things will change with gear, but what I’m observing in game, what I’m seeing in top raid parses, and what I’m hearing from top raid guilds all suggests the same thing: right now classes are not balanced, and SV hunters are up in the top end of the imbalance.
I emphatically don’t think we need a big nerf. I don’t think we need anything like the great BM nerf (which we didn’t need then either). But right now it looks like we need our SV output lowered a little bit, and the other specs raised a bit.
And if by tier 2 things change, then we should rebalance for that. The game is constantly changing, and we want our PVE balanced for the current raid content.
If any of you guys have data compiled that contradicts this, please share it. I’d love it if the SV dps was right where it should be and it was just BM and MM that needed buffs all the way up to that level.
I would love that dearly, but I can’t find any way to produce data to support that stance.