M+ Balance, DPS Per Spec 9.1

Community General Discussion
1

DPS per spec for +15 dungeons as of 9/2/21

Spec 25th Avg 1st Avg % RIO DOS HoA MoTS NW PF SD SoA ToP
Windwalker 9,317 12,166 100 2884 9,468 10,496 10,284 10,256 8,438 10,527 7,565 7,508
Marksmanship 8,973 12,054 96 2845 8,638 9,774 9,947 9,669 8,093 10,541 7,648 7,476
Elemental 8,951 11,299 96 2831 8,785 10,447 10,067 10,121 8,033 10,146 7,188 6,824
Fury 8,754 11,249 93 2790 8,377 9,617 9,803 9,861 7,944 9,872 7,184 7,374
Havoc 8,699 11,500 93 2884 8,254 8,991 9,555 9,609 8,188 10,115 7,272 7,612
Frost Ma 8,695 11,134 93 2779 8,359 9,745 10,000 9,201 7,865 10,170 7,120 7,100
Arms 8,508 10,597 91 2792 8,105 8,954 9,482 9,407 7,718 9,497 7,362 7,546
Affliction 8,367 10,854 89 2815 8,070 10,034 9,555 8,946 7,446 9,303 7,017 6,569
Fire 8,348 11,428 89 2853 8,379 9,557 9,570 8,815 7,256 9,133 7,212 6,869
Arcane 8,333 10,809 89 2863 7,858 8,886 9,589 9,043 7,132 9,453 7,464 7,245
Subtlety 8,193 10,143 87 2826 8,011 9,002 9,102 8,957 7,266 9,719 6,929 6,559
Outlaw 8,051 9,719 86 2810 7,598 8,256 8,939 8,605 7,704 9,032 7,006 7,275
Balance 7,967 10,175 85 2891 7,982 9,437 9,120 8,088 6,974 8,791 6,877 6,467
Unholy 7,924 9,953 85 2715 8,059 9,050 8,738 8,564 6,918 8,772 6,517 6,777
Beast Master 7,910 10,494 84 2793 7,695 8,435 8,763 8,349 7,406 8,677 6,934 7,023
Shadow 7,757 10,057 83 2754 7,653 8,611 8,924 7,964 6,943 8,563 6,761 6,642
Survival 7,743 10,343 83 2764 7,925 8,104 8,646 8,222 7,129 9,075 6,427 6,416
Destruction 7,595 10,276 81 2754 7,407 8,176 8,551 7,909 6,993 8,166 6,752 6,807
Frost DK 7,464 9,546 80 2725 7,535 8,201 8,291 7,965 6,422 8,489 6,230 6,579
Retribution 7,320 8,853 78 2740 6,981 7,747 8,219 8,152 6,437 8,062 6,307 6,658
Assassination 7,224 8,624 77 2742 6,776 7,607 8,040 7,439 6,873 8,118 6,286 6,657
Enhancement 7,097 9,104 76 2680 6,728 7,289 7,852 7,536 6,794 8,116 6,073 6,393
Demonology 6,864 8,321 73 2531 6,432 7,557 7,874 7,188 6,451 7475 5,892 6,046
Feral 6,743 8,072 72 2600 6,476 6,966 7,546 6,952 6,281 7,457 6,053 6,219

Data comes from Raider IO and warcraftlogs as of 9/1/21 from pages like these:

`https://www.warcraftlogs.com/zone/rankings/25#metric=dps&partition=2&class=Monk&boss=12291&bracket=15&spec=Windwalker

25th avg is the average of the 25th best parse from each dungeon at +15 ( stable )

1st avg is the average of the 1st best parse from each dungeon at +15 ( outliers )

% is the percentage of the best 25th avg ( 9,317 DPS )

RIO is the highest IO score for that spec.

Dungeons are the 25th best parse for that Dungeon. Check my math, I’d be very surprised if there isn’t a mistake in this somewhere.

The order is based off the 25th average because at that point the DPS seems stable and the difference between adjacent parses is in the 10’s, not 100’s or 1000’s

11 Likes
2

This seems like good work, but why not do percentiles instead? If there are 50 feral parses for a dungeon and 3000 balance parses for a dungeon (made up numbers), the feral amount is coming from a very average player and the balance amount is coming from one that is an extremely good player. 25 out of 50 vs 25 out of 3000, I do not believe that these two should be doing equally well anyway. I think it would be more proper to use the 2nd feral log and 25th (or well, 60th to be more accurate … with these fake numbers) balance log.

Also, DPS is not the only thing that matters. If you’re interested to see how specs are doing, this website runs statistical regression on all runs that are level 16 or higher (where the meta starts to matter):

https://mplus.subcreation.net/all-affixes.html

7 Likes
3

Interesting you do 15s. Let’s see 20s.

3 Likes
4

What is the purpose of this thread? I’m not trying to be mean I’m genuinely asking.

5 Likes
5

It’s the data that’s available. You can take it as you see fit.

By the 25th best parse the DPS is pretty close to one another so on average you could assume that’s a sample from the best players of the spec with the best gear. Maybe a good bench mark for the potential of the spec.

6

I don’t disagree that it’s a better choice to choose the 25th log over the highest log, especially since this data can be tampered with. But perhaps it would be even better to use the 75th percentile DPS for your comparison.

3 Likes
7

It’s not actually. The mobs lower health and the fact you don’t need mass pulls to time in 15s actually limits many dps. Again do 20s

1 Like
8

Thanks for informing us, the numbers don’t lie and anyone who develops a game with that amount of extreme variance doesn’t have a leg to stand on when confronted with the facts.

The fact is certain classes are gimped while others are not. I’ve seen these kind of stats change, hinging on a single whim. One huge reason many folks do not play poorly designed games.

Might as well just play chess but let each tourney have different rules some as whimsical as if you allowed pawns to move 3 squares but only on tuesdays when the moon is waning.

That’s how whimsical many of us view the changes made to the game are. You can only imagine my opinion of folks who get paid to produce this content.

2 Likes
9

If you cherry pick your stats you can get basically whatever you want. Marksmen is #2 there in 15s because the trash dies before their cooldowns end. In 20s they plummet.

4 Likes
10

Interesting ti see it laid out like that. One bias that it many people talk about is that the best players will always play the best specs. So will perform even higher rthan the best of a lower rated class.

If in 9.2 feral is the out and out highest dps then the best dps players will reroll feral so a good number of those top 25 windwalkers would then be the top 25 feral for example.

I’m not trying to justify soch a big spread or anything but it’s an interesting statistical thing Imo

2 Likes
11

I’m trying to be agnostic and just present the data.

There may be balance changes with 9.1.5, it will be interesting to see if they improve things are not.

Also you can see things like a very strong correlation between DPS performance and highest IO from the spec.

1 Like
12

Not nearly as bad as I thought it was gonna be tbh

13

Ya, it’s closer than I thought it would be as well.

But 72% of best performance is still over a 2k difference, that is a lot.

14

We get it, you wanna see 20s so you can complain about meta.

3 Likes
15

Yeah, but even demo sitting in the bottom 2 I know my spec enough to acknowledge that while our overall aoe isn’t great, we make up for it by absolutely cranking on ST priority targets.

While this leads us to be lower with the Aoe, our profile means we are far more valuable on tyrannical for example.

Though that may have skewed a bit since all the bosses are nerfed heavily atm.

2 Likes
16

It’s actually decent. You have to take damage profiles into account.

17

That would probably be better to judge how the average player is doing with a spec. It would be interesting to see if it’s produces drastically different results in the relative orders.

But I think the 25th best parse is a better measure of the specs potential. That may be the case it may not be.

18

Thanks for posting this.

…so I guess what you’re REALLY implying is that we DON’T need the AOE cap removed? :wink:

19

Ik this is probably sarcasm but the issue with the aoe cap is it prevents us from pulling half the dungeon. That is whats fun.

20

You can get a rough idea of how much impact this has by looking at the individual dungeons and how much those orders deviate from the average.

The devil is deeply buried in the details on that issue. Could improve balance if done right. Could make it much worse.