DPS Rankings after Day 1 of 7.1 Changes

[neocount]

Can someone please explain the whole percentile thing? Math is a mysterious thing to me, lol.

[Dch48]

seems like it would be hard to trust these results 100%

how many world class players are taking up a mythic raid slot with survival?

i'd worry about this being a comparison between pro and amateur more than spec vs spec

This is exactly right. All the top tier Hunters play MM because they have been convinced that it's the best. If some of them played BM, the difference between the specs would be much less. All of these kinds of "rankings " are invalid and Blizzard has said that's why they don't pay much attention to them. They shouldn't.

I play BM and in my guild raids, I can keep up with a MM Hunter easily. BM AOE is far better than ever before with the artifact traits and on a single target, I can beat our MM (he's a better overall player than me too). In Heroic and Mythic dungeons, I am the top dps about 50% of the time. This was without my Legendary ring which I just got from a Wardens Emissary chest and haven't tried out yet.

[Daan]

6. Hunters have to roll MM to be competitive.

Welcome to Blizzard's well-wrought system. Giving Hunters a lot of new pets, but the only reasonable specc is MM without pets.

[ejpaints]

Welcome to Blizzard's well-wrought system. Giving Hunters a lot of new pets, but the only reasonable specc is MM without pets.

I swear if I ever find a game that lets me tame what I want the way WoW does, that might be what finally steals me from WoW. It's the biggest draw for me on the hunter class. Every MMO I've played has some sort of archer but WoW is the only one I've tried that lets me customize pet choice. And it's a bottom of the barrel subpar spec again.

[Deleted]

>75%
into the garbage dumpster you go, OP

[Larynx]

Can someone please explain the whole percentile thing? Math is a mysterious thing to me, lol.

75th percentile = top 25% of entries

99th percentile = top 1% of entries

Player skill is more or less a bell curve, so the difference between 75 and 99th is much larger than the difference between 74 and 50, for example. A 74 and 50 might have a difference of 30k dps between highest and lowest, but 99 and 75 might have a difference of 60k between the two.

[BiggestNoob]

Can someone please explain the whole percentile thing? Math is a mysterious thing to me, lol.

https://en.wikipedia.org/wiki/Percentile

Basically a 75th percentile includes data points for all parses above the 75th percentile. That is, the best 25% of all logs for each given spec. A 90th percentile only includes the top 10%.

I'm not entirely sure what WCL is showing because it's not communicated well. It shows overall numbers in each category, doesn't indicate what set of parses are in the graph shown, or how the score is obtained.

I'd much rather an option to export the dataset so we can use statistical analysis tools to better figure out how specs are performing. Discard outliers due to deaths, group on legendaries, etc.

[JyuNyar]

I'd much rather an option to export the dataset so we can use statistical analysis tools to better figure out how specs are performing. Discard outliers due to deaths, group on legendaries, etc.

This is something I've wanted to do as well.

I resorted to pulling data straight from warcraftlogs using their API. I put some early results up on the wow subreddit.

Its definitely not perfect, but it gives a better idea of what kind of variance your dealing with from spec to spec.

[Deleted]

Never get into any kind of game development.

Ever.

Just ignore him. He keeps spamming the mage forums with how arcane should be buffed by 50+% in flat damage so it can "keep up".

- - - Updated - - -

That isn't true, because they include all players, with legendaries or not. Unless you're saying Fire has a higher legendary droprate than other specs and classes, which would be a pretty big surprise to everybody.

OR he might be saying that fire has a much wider spread of parses than most other classes, with the people who have 2 good legendaries making up the vast majority of the 75th percentile where as from the 50th percentile down you get all the people who are miles behind due to not having said legendaries (specifically the bracer). Which makes fire good at the high-end with legendary, but godawful at the low end without it.

[cFortyfive]

All of these kinds of "rankings " are invalid and Blizzard has said that's why they don't pay much attention to them. They shouldn't..

Sure because they have proven to be doing such a great job .

[Deleted]

When will they buff warlocks to not be underperforming with all 3 specs?

I know we're not mages, but I'd like to at least be up there with specs like Ret, Havoc, Frost dk etc.

[Notdev]

I am late to this thread. But this "evidence" is grossly misleading.

The literal only reason Fire is so "high" is because of the last 30% of Xavius.

They are middle of the pack on Nythendra and Ursoc, lower middle on Il'gynoth and Cenarius, and decent on Elerethe. They are strong on Xavius only, and literally only because of the complete ignite pad on the tentacles at the end.

Fire is a middling spec, this is very misleading. Needs a single target buff, and maybe a small ignite cleave nerf to compensate. If, somehow, you were able to take rankings from 6/7, they would drop 5-10 spots on that list in OP.

[Deleted]

the only class they knee jerk nerfed was outlaw rogues... which in there words was stupid.

difference is for shadow, one talent is breaking the whole spec everything else outside of that talent is absolutely fine, do you just propose they give a 20% dmg nerf for 1 talent and absolutely hammer the class into nothingness outside of raiding? it's better they just leave it till they can fix it properly.

They could quite literally remove S2M and it would change nothing for spriests outside of raiding/dungeonbosses. It's an overpowered talent in the raid setting where you have an exceptionally long execute phase where a priest can reach 100+ stacks of S2M. Outside of that it's not even worth picking. Doesn't help that void lord and whatever the last talent is are both godawful though.

[Woobels]

I am late to this thread. But this "evidence" is grossly misleading.

The literal only reason Fire is so "high" is because of the last 30% of Xavius.

They are middle of the pack on Nythendra and Ursoc, lower middle on Il'gynoth and Cenarius, and decent on Elerethe. They are strong on Xavius only, and literally only because of the complete ignite pad on the tentacles at the end.

Fire is a middling spec, this is very misleading. Needs a single target buff, and maybe a small ignite cleave nerf to compensate. If, somehow, you were able to take rankings from 6/7, they would drop 5-10 spots on that list in OP.

Fire with correct legendaries is somewhat competitive and I'd put them higher than middle of the pack, still suck on Cenarius though.

The issue is that having the right orange items on that spec makes so much of a difference it's insane, it's comparable to Arms etc imo.

[Notdev]

Fire with correct legendaries is somewhat competitive and I'd put them higher than middle of the pack, still suck on Cenarius though.

The issue is that having the right orange items on that spec makes so much of a difference it's insane, it's comparable to Arms etc imo.

Everything I said holds true when you set it to 95 percentiles.

[Nubulous]

This is exactly right. All the top tier Hunters play MM because they have been convinced that it's the best. If some of them played BM, the difference between the specs would be much less. All of these kinds of "rankings " are invalid and Blizzard has said that's why they don't pay much attention to them. They shouldn't.

I play BM and in my guild raids, I can keep up with a MM Hunter easily. BM AOE is far better than ever before with the artifact traits and on a single target, I can beat our MM (he's a better overall player than me too). In Heroic and Mythic dungeons, I am the top dps about 50% of the time. This was without my Legendary ring which I just got from a Wardens Emissary chest and haven't tried out yet.

I just wanted to respond to this because on our first kills this week for mythic our BM hunter was at

Ursoc-376k
Dragons-392k

I think BM is still undervalued and honestly if you have a good legendary and 26 points its still pretty good

- - - Updated - - -

Welcome to Blizzard's well-wrought system. Giving Hunters a lot of new pets, but the only reasonable specc is MM without pets.

This is a total lie. You can be BM and be just as competitive. For the average player BM is probably easy to do dps on than as marks anyway, and you can actually have more mobility and not have to stop a bunch to cast aimed shots. We are 4/7 mythic and have 2 BM hunters in our raid group

[Arterus]

I am late to this thread. But this "evidence" is grossly misleading.

The literal only reason Fire is so "high" is because of the last 30% of Xavius.

They are middle of the pack on Nythendra and Ursoc, lower middle on Il'gynoth and Cenarius, and decent on Elerethe. They are strong on Xavius only, and literally only because of the complete ignite pad on the tentacles at the end.

Fire is a middling spec, this is very misleading. Needs a single target buff, and maybe a small ignite cleave nerf to compensate. If, somehow, you were able to take rankings from 6/7, they would drop 5-10 spots on that list in OP.

There is at least one mage, or more in the top 10 WoL ranking for every boss on mythic. Just because you don't have good legendaries while other people in your guild do, does not equate to your spec being middle of the pack. Even on Neth, the only pure single target boss in mythic a fire mage is #2.

Fire is more than competitive, theres not much it doesn't have going for it. The spec has like 4 main buttons to press not including cds, so its simple. Its ranged, with excellent survivability cds. And it has good cleave.

[Kihra]

These stats are worthless because you can't discard outliers.

Unless you pick max, outliers should not be much of an issue. I think there is a misunderstanding here of what I show on the statistics page. When you select a percentile for a boss, e.g., 75% on Nythendra, the value you are seeing is how much DPS was done by the spec at the 75% rank position when considering all of the parses for that spec over the sampled period (2 weeks is the default sample period).

In other words, outliers at the top don't cause the 75% number to rise. It sounds like people think 75th means i'm looking only at the top 25% of parses and doing some math on them. That *would* be skewed by outliers, but that's not what I'm doing.

[scwasp]

As an Unholy DK, 7.1 is pretty rough I have to say. Changes to frost combined with lack of changes to uh was pretty much equivelant to blizzard taking us out the back and shooting us.

[Deleted]

There is at least one mage, or more in the top 10 WoL ranking for every boss on mythic. Just because you don't have good legendaries while other people in your guild do, does not equate to your spec being middle of the pack. Even on Neth, the only pure single target boss in mythic a fire mage is #2.

I dont know if WoL is relevant and how many people use it but on warcraftlogs it's only 4/7 in top10 and two of them have variety of classes so it's just rng in that specific fight with procs and boss abilities. Even on WoL your "every boss" is not every boss, it's 5/6 (i can't see any cenarius logs but cenarius is the worst fight for fire mage so i'll bet it is 5/7) and except Ursoc, there is one mage on each boss (and there is your WoL relevancy, because I don't believe there is not more than one mage on fights like dragons or elerethe) . For nythendra mythic, in overall, fire mage is #8 and there is one in top10 dps. So if you want use numbers at least you can use right numbers and don't lead on.