Best and Worst DPS Specs in the game (by stats)

TL:DR: Below are the specs you want to main if you want Blizzard to give a damn about your existence.




Hey all,

I was curious to find out if Blizzard has any preferential treatment when it comes to balance for specs, so I decided to take a look at logs dating back to Highmaul (that furthest ones I could find).

The provide some interesting findings, which I’ll describe and summarize below. I also included the Excel file image below (with sources) for anyone curious as to where I got the data.

Note that for all of these, I looked only at Raid and at All Bosses ranking (which is flawed, I know), 90th Percentile and then ranked the DPS specs (i.e. in the current tier, Destro would be rank 1, then Demo at rank 2, etc…). Knowing that there are 24 DPS specs, that means that if your spec performs right in the middle on average accross all raid tiers, it should be ranked 12th on average. Specs ranked less than 12th are therefore, on average, given preferential treatment (historically), and specs ranked more than 12th are, on average, not as cared for by Blizzard’s balance team.

I’ve also included the number of times an individual spec was top 5 DPS for that specific raid tier, as one spec could technically be doing well (i.e. Havoc DH, as we’ll see) but never actually be “one of the best DPS” for any given raid tier.




So here’s the data from Highmaul (WoD) to Sepulcher of the First Ones (SL):

On Average, these specs are ranked the following out of 24 across all raid tiers between now and Highmaul:

1) Shadow Priest: 7.4
2) Fire Mage: 7.8
3) Affliction Warlock: 9.1
4) Demonology Warlock: 9.1
5) Balance Druid: 10.1
6) Havoc Demon Hunter: 10.2
7) Destruction Warlock: 10.5
8) Fury Warrior: 10.5
9) Arms Warrior: 11.3
10) Elemental Shaman: 11.5
11) Frost Death Knight: 11.6
12) Outlaw / Combat Rogue: 12.4 — NOTE: This is where the Average DPS should be —
13) Assassination Rogue: 12.5
14) Subtlety Rogue: 13.2
15) Enhancement: 13.4
16) Feral Druid: 13.9
17) Marksmanship Hunter: 14.0
18) Retribution Paladin: 14.5
19) Windwalker Monk: 14.7
20) Survival Hunter: 15.4
21) Arcane Mage: 15.5
22) Unholy Death Knight: 15.6
23) Frost Mage: 16.7
24) Beast Mastery Hunter: 16.7

Here are the specs ranked by how many raid tiers they have been top 5 DPS specs:
1) Shadow Priest: 10 raid tiers
2) Fire Mage: 8 raid tiers
3) Affliction Warlock: 7 raid tiers
4) Demonology Warlock, Destruction Warlock, Subtlety Rogue: 5 raid tiers
5) Assassination Rogue, Outlaw/Combat Rogue, Elemental Shaman, Arms Warrior, Fury Warrior, Balance Druid, Frost Death Knight: 4 raid tiers
6) Survival Hunter, Windwalker Monk, Retribution Paladin: 2 raid tiers
7) Beast Mastery Hunter, Marksmanship Hunter, Frost Mage, Arcane Mage, Enhancement Shaman, Unholy Death Knight: 1 raid tier
8) Feral Druid, Havoc Demon Hunter: 0 raid tiers




Here’s the data from Emerald Nightmare (Legion) to Sepulcher of the First Ones (SL):

On Average, these specs are ranked the following out of 24 across all raid tiers between now and Emerald Nightmare:
1) Shadow Priest: 6.7
2) Affliction Warlock: 8.1
3) Fire Mage: 9.4
4) Frost Death Knight: 9.5
5) Demonology Warlock: 9.8
6) Fury Warrior: 10.2
7) Havoc Demon Hunter: 10.2
8) Assassination Rogue: 10.3
8) Balance Druid: 10.3
9) Destruction Warlock: 11.2
10) Outlaw Rogue: 11.5
11) Elemental Shaman: 12.0 — NOTE: This is where the Average DPS should be —
12) Arms Warrior: 12.1
13) Subtlety Rogue: 13.1
14) Feral Druid: 13.2
15) Survival Hunter: 14.5
16) Enhancement Shaman: 14.7
17) Marksmanship Hunter: 15.2
18) Tie: Frost Mage, Windwalker Monk and Retribution Paladin: 15.4
19) — (above) —
20) — (above) —
21) Arcane Mage: 16.5
22) Unholy Death Knight: 17.4
23) Beast Mastery Hunter: 18.1

Here are the specs ranked by how many raid tiers they have been top 5 DPS specs:
1) Shadow Priest: 9 raid tiers
2) Affliction Warlock: 6 raid tiers
3) Fire Mage, Destruction Warlocks: 5 raid tiers
4) Assassination Rogue, Outlaw Rogue, Subtlety Rogue, Fury Warrior, Balance Druid, Frost Death Knight: 4 raid tiers
5) Elemental Shaman, Demonology Warlock, Arms Warrior: 3 raid tiers
6) Survival Hunter, Windwalker Monk: 2 raid tiers
7) Beast Mastery Hunter, Frost Mage, Retribution Paladin: 1 raid tier
8) Unholy Death Knight, Havoc Demon Hunter, Marksmanship Hunter, Arcane Mage, Enhancement Shaman: 0 raid tiers.




What conclusions can we draw from these results?

We can determine from this that Blizzard has historically given overwhelming preferential treatment in raid to the following spec: Shadow Priest, Fire Mage, Affliction Warlock, Demonology Warlock.

We can also determine determine that, historically, Blizzard doesn’t give a crap about the following specs: Survival Hunter, Arcane Mage (can confirm), Unholy Death Knight, Beast Mastery Hunter.

Note that this says NOTHING about current raid tier, where Shadow Priest, for example, is currently ranked 12th, whereas it was ranked 1 in Castle Nathria, as an example. Also note that this says nothing about Mythic+ or PvP, where Feral Druid and Assassination Rogue, as examples, are abysmally ranked, as opposed to their raid ranking.




Below is the dataand sources that I’ve gathered and tabulated for the results shown above:

https://imgur.com/gallery/bled852

3 Likes

Welp, there it is. Looks like it’s time to reroll, can’t be seen on this midtier poop class :slightly_frowning_face:.

Dragonfight will fix it just remain clam and noboday panic.

Cute thread, I like your formatting. I like that Feral is bad. I think MM and arcane should be top 5 more often.

1 Like

didn’t ion say he didn’t want people playing warlocks or a certain spec and that it was purposely neglected? think that was like 2017 or something.

Been saying for a while how these specs have dominated the game for the last 8 years.
It’s not talked about enough.

4 Likes

Inept apathy.

They no longer are willing to do the hard work that once had 12 millions playing.

2 Likes

I am by no means saying any of this data is wrong.
Numbers do not lie, and since you self-imposed yourself with this task, I assume you did it correctly.

But I’ve played my fair share of this game, and if you telling me that Demo Warlocks are in the top 5 specs to play for raiding, and is a spec Blizzard gives a damn about you, I simply can not bring myself to believe it.

This is the first time, in a long time, that Demo has been worth a damn.
And even then their time in the spotlight was cut short as the Destro 4-piece became commonly acquired.

Also, the implication here is that it’s good to be #5 100% of the time.
I’d rather a spec that is half the time number 1, and half the time number 10, opposed to a spec that is always fifth place. Also, the summary of this data would state that both are treated equally.

1 Like

Remember that they are using 90th percentile here. Demo is one of the specs that has more control over its damage depending on skill, so I think comparing the skilled players won’t say much about the overall performance of the class (bad demo locks will do much worse than bad havoc dh, for example).

At the same time, it’s probably a good move to use a high percentile because these are the only people who must care about the specs’ performance because it’s the only way they can improve. People parsing 10, for example, should not blame their spec, they will probably get better results by learning their spec rather than changing to a new one.

Has not happened since Highmaul (for MM) and Hellfire Citadel (for Arcane). About 8 years ago or so.

3 Likes

Seems like fire mage has a safe pick forever. Maybe its time to brush off the shdow priest :).

I was enjoying my Saturday and then you go and attack me like that?
I remember PoF + Arcano Crystal Arcane mage like it was yesterday.

2 Likes

Yeah, it seems Blizzard was thinking the same thing, and might be why these geriatrics keep denying buffs to a 2nd worse DPS spec.

1 Like

Whilst I’m hear rattling my brain thinking about old METAs before bed.

When is this data taken? I assume it’s throughout the course of the entire tier. Which of course, may be the correct way to evaluate this.

But to me, personally, I think class performance during progression weighs far more heavily than class performance during farm phase, and is what Blizzard likely balances around.

Rogue isn’t in the top 5, but Rogue has been an iconic class for the top Guilds to stack during progression over the years.
And despite the low DPS, BM Hunter excels at doing DPS whilst also doing mechanics. That is something that becomes less important during farm phase.

There’s more to class than Warcraft log parses, which I guess is my overall gripe to the summary of this data.

Demo was great in literally 2 of the 3 main raid tiers in BFA. It was good when SoD released in 9.1. I’m not trying to start something but this is just bias.

And demo was just as good in prog. Your class, not just demo, is a definite favorite of Blizzard. Whether you wanna accept it or not :man_shrugging:t6:

Because of their survivability. And before WoD, rogues were routinely S+ tier dps. IIRC anyway.

And yet you put Shadow Priest at the top, debunking yourself quite handily. Impressive.

Are you able to read or no?

Numbers simply do not agree with you.

Learn to read before making asinine comments.

1 Like

I love this, I remember seeing an info graphic thing awhile ago that did a similar thing where they noted how often specs were top and the safest bet was warlock, and I found that interesting. This sort of confirms/builds on that.

The other interesting corellation is that DH are almost never the very tippy top, which makes me wonder about the nerf DH meme.

1 Like

Just to be clear, these are rankings of what raid and mythic leaders wanted to see in their raids rather than say, how much punishment a given spec to deal to a training dummy? If so, then perhaps this has to do with raid encounter design and how world first guilds respond to those encounter designs rather than DPS design.

Most likely a combination but to look at DPS without an analysis of the raid encounters and what has become standard mechanics on the part of guilds that may be more ingrained than carefully analyzed.