3.2 Tank Tests

[Xav]

3.2 Tank Testing & Comparisons

First post in this thread will contain just the information. Replies will contain everything else and the discussion.

Patchwerk (PTR Tank Test) swings every 1 second, does not dual wield, and hits for about 40k on a tank class before buffs/debuffs. After appropriate buffs/debuffs swing damage is roughly 25k. It is approximately as hard as an Immortal Guardian at 100% HP in Yogg-Saron 25, but swinging twice as fast. This version of the Patchwerk Tank Test is extremely different from the most recent version most players saw.

All buffs/debuffs that contribute to a tank doing more damage or taking less damage were up between 99 and 100% of the time, unless otherwise noted. This includes devotion aura, stoneskin, scorpid sting/insect swarm, demoralizing shout, leader of the pack, blood poisoning, curse of weakness, etc, etc.

Notes will be edited in as they're made more clear (exceptions/mistakes made during the testing, for example)

All gear profiles are very close, but not exact, there may be minor gem differences.

Each Tank Trial began with the tank getting a single misdirect from a hunter, Patchwerk approaching the tank, and the tank using a survivability cooldown to soften the impact while crucial raid buffs and debuffs are applied.

Tanks were trialed for 5 minutes, after that the logs were cut. The logs generally may be within a few seconds of the 5 minutes, and often actually 5 minutes starting from the first time Patchwerk or the tank performs an attack in melee range, similarly Threat screenshots are not always exact due to the timer being started at different times due to user error (!) and/or screenshot being delayed for the same reason. However, the overall effects of the duration for the purposes of threat calculation is extremely minor due to the size of the numbers being dealt with (millions).

The tanks were performing maximum threat rotations to the best of their ability while trying to rotate in survivability cooldowns as soon as they come off cooldown. All rotations are not 100% perfect but they fairly accurately represent what a tank can be reasonably expected to do in a real raid situation.

The logs have been parsed and uploaded by two different sources as well as uploaded to two different hosts, you have your choice of which you go to for perusing:


StasisCL - http://premoguild.com/pwtest/
If using premoguild's stasis (pwtest) link, Try 1,2,9 & 10 are Xav's tank trials in order, Try 3,4,13, & 14 are Melaar's trials in order, Try 5,6,11 & 12 are Devium's trials in order, Try 7,8,15, & 16 are Fayz' trials in order.

World of Logs - http://www.worldoflogs.com/guilds/6280/reports/

Information on Damage Taken

Xav (Warrior)

Talents/Glyphs: http://ptr.wowhead.com/?talent#LAM00...idczsGo:dTcoMm
Gear/Spec: http://profiler.wowhead.com/?profile=13034658
Raid buffed stats: http://premoguild.com/publictesting/all4stats.jpg

Warrior Tank Test 1:

Damage taken: 2,050,000~
DPS Taken: 6950
Damage blocked: 138,000~
Damage absorbed: 409,000~
Average absorb size (SS/PW:S): 5900
Average hit taken (counts all Absorbs): 22000
Average hit after block: 19696
[60.1 Total Avoid] Missed 10.9%, Dodged 31.5%, Parried 17.7%, Blocked 23.8%, Absorb 27%
Cooldowns: 2 Shield Wall, 2 Heart of Iron, 4 Monarch Crab, 8 Shield Block, 33% Mongoose Uptime

Warrior Tank Test 2:

Damage taken: 2,350,000~
DPS Taken: 7900
Damage blocked: 161,000~
Damage absorbed: 420,000~
Average absorb size (SS/PW:S): 5050
Average hit taken (counts all Absorbs): 22000
Average hit after block: 19700
[55.3 Total Avoid] Missed 11.7%, Dodged 23.4%, Parried 20.2%, Blocked 28.6%, Absorb 33%
Cooldowns: 2 Shield Wall, 2 Heart of Iron, 3 Monarch Crab, 8 Shield Block, 58% Mongoose Uptime

Warrior Tank Test 3:
Note: 3 Healers were used on this rather than 4. Reason for this is successive trials using 3 healers resulted in two wipes in a row early on due to insufficient HP/S to keep a tank alive if they were hit 4 times in a row, while trialing Warrior. Due to time constraints we moved to 4 healers to prevent wasting our opportunity to test (We had a limited number of pulls, after a certain point, and no way at all to get them back, due to a literal quantity of mobs)

No priest was originally going to be used on the 3-healer trials, so this trial has only Sacred Shield absorbs, no Power Word Shields. For most damage-taken analyizing, this parse is useless due to it being an unfair comparison. It is here for sake of completeness, and the Threat section.

Damage taken: 2,475,000~
DPS Taken: 8350
Damage blocked: 160,000~
Damage absorbed: 127,000~
Average absorb size (SS): 3540
Average hit taken: 25000
Average hit taken (counts all Absorbs): 22549
[57.5 Total Avoid] Missed 13.8%, Dodged 23.4%, Parried 17.8, Blocked 25.9%, Absorb 14.2.%
Cooldowns: 3 Shield Wall, 2 Heart of Iron, 3 Monarch Crab, 8 Shield Block, 33% Mongoose Uptime

Warrior Tank Test 4:
Also fairly useless for tank damage taken comparison, as the paladin who subbed in to heal this attempt didn't know to keep up sacred shield 100% of the time.

Damage taken: 2,165,000~
DPS Taken: 7300
Damage blocked: 149,000~
Damage absorbed: 268,000~
Average absorb size (PW:S): 4800
Average hit taken (counts all Absorbs): 24380
Average hit after block: 20253
[59.9 Total Avoid] Missed 9.7%, Dodged 34.4%, Parried 15.8%, Blocked 24.3%, Absorb 22.3%
Cooldowns: 3 Shield Wall, 2 Heart of Iron, 2 Monarch Crab, 8 Shield Block, 49% Mongoose Uptime


Melaar (Druid)

Gear/Spec: http://chardev.org/?template=234937
Raid buffed stats: http://premoguild.com/publictesting/all4stats.jpg

Druid Tank Test 1:

Damage taken: 1,935,000~
DPS Taken: 6470
Damage absorbed: 495,000~
Average absorb size (SS/PW:S/SD): 5633
Average hit taken (counts all Absorbs): 18777
[58.5 Total Avoid] Missed 8.5%, Dodged 50%, Absorb 35.8%
Cooldowns: 4 Barkskin, 3 Heart of Iron, 33% Greatness Uptime (7 procs), 53.5% Mongoose Uptime

Druid Tank Test 2:

Damage taken: 1,848,000~
DPS Taken: 6170
Damage absorbed: 554,700~
Average absorb size (SS/PW:S/SD): 6186
Average hit taken (counts all Absorbs): 18100
[59.3 Total Avoid] Missed 8.9%, Dodged 50.4%, Absorb 35.5%
Cooldowns: 4 Barkskin, 2 Heart of Iron, 30% Greatness Uptime (7 procs), 30% Mongoose uptime

Druid Tank Test 3:

Damage taken: 2,000,000~
DPS Taken: 6750
Damage absorbed: 500,000~
Average absorb size (SS/PW:S/SD): 5559
Average hit taken (counts all Absorbs): 18800
[56.7 Total Avoid] Missed 11.4%, Dodged 45.3%, Absorb 36.3%
Cooldowns: 5 Barkskin, 3 Heart of Iron, 32% Greatness Uptime (7 procs), 55% Mongoose uptime

Druid Tank Test 4:

Damage taken: 1,577,000~
DPS Taken: 5300
Damage absorbed: 504,000~
Average absorb size (SS/PW:S/SD): 6105
Average hit taken (counts all Absorbs): 18000
[64.5 Total Avoid] Missed 11.4%, Dodged 53.1%, Absorb 32.7%
Cooldowns: 5 Barkskin, 3 Heart of Iron, 35% Greatness Uptime (7 procs), 50% Mongoose uptime

Devium (Death Knight)

Spec: http://ptr.wowhead.com/?talent#jcEGM...0AbIoczbuzhZ0L
Gear: http://profiler.wowhead.com/?profile=13034597
Raid buffed stats: http://premoguild.com/publictesting/all4stats.jpg

Death Knight Tank Test 1:

Damage taken: 1,530,000~
DPS Taken: 5160
Damage absorbed: 463,000~
Average absorb size (SS/PW:S): 6013
Average hit taken (counts all Absorbs): 17500
[64.6 Total Avoid]Missed 15.4%, Dodged 30.1%, Parry 19.1%, Absorb 31.3%
Cooldowns: 3 Icebound Fortitude, 2 Unbreakable Armor, 2 Heart of Iron, 2 Monarch Crab

Death Knight Tank Test 2:

Damage taken: 1,561,000~
DPS Taken: 5260
Damage absorbed: 455,000~
Average absorb size (SS/PW:S): 6500
Average hit taken (counts all Absorbs): 18300
[65.6 Total Avoid]Missed 14.6%, Dodged 32.4%, Parry 18.6%, Absorb 28.3%
Cooldowns: 3 Icebound Fortitude, 2 Unbreakable Armor, 2 Heart of Iron, 3 Monarch Crab

Death Knight Tank Test 3:

Damage taken: 1,561,000~
DPS Taken: 5950
Damage absorbed: 455,000~
Average absorb size (SS/PW:S): 6110
Average hit taken (counts all Absorbs): 19500
[63.6 Total Avoid]Missed 15%, Dodged 30.4%, Parry 17.8%, Absorb 28.4%
Cooldowns: 3 Icebound Fortitude, 3 Unbreakable A2mor, 2 Heart of Iron, 2 Monarch Crab

Death Knight Tank Test 4:

Damage taken: 1,570,000~
DPS Taken: 5285
Damage absorbed: 385,000~
Average absorb size (SS/PW:S): 6316
Average hit taken (counts all Absorbs): 19000
[66.9 Total Avoid]Missed 17.3%, Dodged 29%, Parry 20.6%, Absorb 24.2%
Cooldowns: 3 Icebound Fortitude, 3 Unbreakable A2mor, 2 Heart of Iron, 3 Monarch Crab

Fayz (Paladin)

Spec: http://ptr.wowhead.com/?talent#sVZV0...ufdxfM0o:0Gpmc
Gear: http://www.chardev.org/?template=234558
Raid buffed stats: http://premoguild.com/publictesting/all4stats.jpg

Paladin Tank Test 1:

Damage taken: 1,720,000~
DPS Taken: 5768
Damage blocked: 153,000~
Damage absorbed: 372,000~
Average absorb size (SS/PW:S): 6072
Average hit taken (counts all Absorbs, 100% blocked): 20475
[66.3 Total Avoid] Missed 13.3%, Dodged 31.7%, Parried 21.3%, Blocked 33.7%, Absorb 24.5%
Cooldowns: 2 Divine Protection, 3 Royal Seal of King Llane, 5 Monarch Crab, 71% Mongoose Uptime

Paladin Tank Test 2:

Damage taken: 2,161,000~
DPS Taken: 7254
Damage blocked: 200,000~
Damage absorbed: 424,000~
Average absorb size (SS/PW:S): 5659
Average hit taken (counts all Absorbs, 100% blocked): 19670
[55.8 Total Avoid] Missed 12%, Dodged 24.9%, Parried 18.9%, Blocked 44.2%, Absorb 30.1%
Cooldowns: 3 Divine Protection, 3 Royal Seal of King Llane, 5 Monarch Crab, 42% Mongoose Uptime

Paladin Tank Test 3:

Damage taken: 1,580,000~
DPS Taken: 5283
Damage blocked: 150,000~
Damage absorbed: 395,000~
Average absorb size (SS/PW:S): 5848
Average hit taken (counts all Absorbs, 100% blocked): 19400
[67.5 Total Avoid] Missed 12.4%, Dodged 34.5%, Parried 20.5%, Blocked 32.5%, Absorb 26.5%
Cooldowns: 2 Divine Protection, 3 Royal Seal of King Llane, 5 Monarch Crab, 70% Mongoose Uptime

Paladin Tank Test 4:

Damage taken: 1,865,000~
DPS Taken: 6270
Damage blocked: 168,000~
Damage absorbed: 397,000~
Average absorb size (SS/PW:S): 4859
Average hit taken (counts all Absorbs, 100% blocked): 20200
[62.9 Total Avoid] Missed 16.1%, Dodged 28.2%, Parried 18.5%, Blocked 37.1%, Absorb 32.3%
Cooldowns: 2 Divine Protection, 3 Royal Seal of King Llane, 5 Monarch Crab, 54% Mongoose Uptime

[Xav]

Threat / DPS

The threat was calculated simply by looking at the ingame provided threat numbers, which is parsed and easily recognizable by Omen. It is as accurate as Blizzard allows to be possible. Total threat is divided by total time in combat with the tank test mob, very slight variations of a few seconds (1-3) may be in place, contrary to what timestamps may indicate.

Warrior: http://premoguild.com/publictesting/xavthreat.jpg

Druid: http://premoguild.com/publictesting/melaarthreat.jpg

Paladin: http://premoguild.com/publictesting/fayzythreat.jpg

Death Knight: http://premoguild.com/publictesting/deviumthreat.jpg
Be aware, the first 2 Tank Tests, the Death Knight was using incorrect keybindings and thus the threat/dps readings are highly skewed and not useful for this. Tests 3 and 4 are fine.

[Xav]

To start with, I'll say this: I haven't looked indepth at the numbers and compared them to the other classes. I've simply gotten all the data together and made it easy to sift through.

I noticed the obvious right off the bat, RNG (random number generator, pure luck) with avoidance makes the biggest difference more than any other factor.

The limited number of parses makes any sustained judgement of a tank's advantages or weaknesses when tanking a boss difficult. This, combined with spoiled parses due to mistakes (healing differences, etc) makes it HIGHLY ADVISABLE not to use these damage taken numbers as 100% repeatable, guaranteed results. Classes randomly swinging 10-15% avoidance from one attempt to the next is a huge difference.

However, average damage taken per swing is fairly clear-cut and different, with druids and death knights taking a noticeable amount less damage per swing (1000-2000), the Paladin and Warrior regularly took ~19500-20000 damage a swing after block.

These are purposelly brief and semi vague conclusions because of the limited number of attempts and the nature of random avoidance.

On the other hand, DPS/TPS numbers are something that are very reliable and don't fluctuate nearly as much as Avoidance. The differences in DPS/TPS between the classes are striking, and the pictures should really speak for themselves:

Warrior: Without Vigilance, sustaining around 7.2-7.8k TPS, ~3200 DPS
Paladin: Sustaining 9-10k TPS, ~3700 DPS
Druid: Sustaining 7.2-7.8k TPS, ~3500 DPS
Death Knight: Sustaining 7.5k TPS, ~3000 DPS

If you account for a DPS class sustaining around 5k TPS, you can have the warrior gaining 500 or 750 TPS from it, depending on the Glyph or not, bringing the warrior to around 8-8.3k sustained TPS, a bit higher than the Druid and Death Knight. However, the DPS differences between the classes is all over the place.

Warriors and Death Knights do comparable DPS while tanking, the druid is a good 10% or so more than them, and the paladin is 20% or more ahead. We've known paladin threat, both burst and sustained, has been insane for a while, and it's still here.

[Xav]

I also forgot to mention / link this.

When we were done with our testing, we had 2 Tank Patchwerks left, not enough to do any real significant individual tanking trials with. So, since we hadn't yet seen the paladin proc Ardent Defender in any situation the entire night (and we're not even sure the Paladin even procced the 30% damage reduction effect much either), we decided to have Fayz tank the last two Tank Patchwerks at the same time.

Remember, these swing every 1 second for ~25k damage.

This is the result of that test:

http://www.worldoflogs.com/reports/Q...TJHtd/spell/0/
http://premoguild.com/pwtest/sws-pat...525/index.html

http://wowwebstats.com/1c3hki5bru1tq...R+Tank+Test%22

Ardent Defender procced twice on Fayz, and he died after the 2nd proc, tanking for over 8 minutes.

It would have been fun to try the same test with the other tanks, but we didn't have any chance to, and frankly I find it extremely unlikely the tanks would've lived more than 2-3 minutes each.

Edit: After looking for a while, seems like I will have to upload every parse, again, to WowWebStats just so I can use the LogBrowser, since the world of logs log browser is really poor/time limited.

And in order to get it to parse correctly WWS I have to rename every single "Patchwerk (PTR Tank Test)" in the log to "Patchwerk PTR Tank Test", so this'll take even more time!

0:08'53.625 Patchwerk PTR Tank Test #2 melee swing hits Fayz for 25647 Physical. (1817 Blocked)
0:08'54.847 Patchwerk PTR Tank Test #1 melee swing hits Fayz for 25101 Physical. (1817 Blocked)
0:08'54.847 Patchwerk PTR Tank Test #2 melee swing hits Fayz for 25330 Physical. (1817 Blocked)
0:08'56.077 Patchwerk PTR Tank Test #1 melee swing hits Fayz for 15034 Physical. (1817 Blocked) (10641 Absorbed)

Living through stuff like this is why Will of the Necropolis was so good, and it's why Ardent Defender is good. Everyone else dies, but you're saved.

Anyway, if you scroll through that log of Fayz tanking both, you can see a bunch of times when he's hit at least 4 times in a row, and the AD damage reduction saves him (or procs when even that isn't enough to save him)!

[Xav]

Some responses to the drooling masses of people who can't read on the WoW Tanking forums:

We got the tanking stats/gear between the tanks as similar as we could. Sorry but we don't all have identical gear, and druids can't wear plate. Also, the different classes can't all wear each other's sets.

No tank was gemmed or geared for threat in any way.. the threat stats that we had were simply what was on our gear, after we tried to equalize our avoidance/armor/hp by swapping around pieces here and there.

The sample size is very small, so you can only draw a limited number of conclusions.

[Zaroua]

In reference to tanking 2 of the Patchwerks at once: for any other tank except Paladins (or maybe Blood DK to a lesser extent), living would've just been a RNG fest on whether or not 2 swings connect back to back without enough healing inbetween. From accurate guesstimations, it felt like Fayz on his test would've died 10+ times without Ardent Defender's DR% (not the built-in Guardian Spirit) from taking 2 swings nearly simultaneously.

[Xav]

Part 2: Tank Tests vs Grand Widow Faerlina (Frenzied)

Due to our limited testing time/sample size on the uber-Patchwerk tank mobs, we decided to go in to Naxxramas25 and try something else. A few guildmates and tank forum posters came to help out.

The setup: Grand Widow Faerlina, one tank on Faerlina. Parsing begins when Faerlina Frenzies, and continues for 9 minutes. The frenzy is not dispelled.

Standardized stats between the 3 plate tanks (literally, all total avoidance values between the 3 plate tanks were within .5% of one another, Armor from gear from the plate tanks was within 200 [if the shield tanks unequipped their shield, they had within ~200 armor of the armor-from-gear as the death knight] - this means that when all 3 plate tanks were against the boss, they had nearly identical gear, and all the differences in armor at that point are due to class differences (armor modifiers, stances, etc, basically only Frost Presence I think).

The profiles for the tanks are the same as the last post, except we modified gems and some other minor gearing choices elsewhere to make all defensive attributes within the very, very small margins we wanted.

Threat stats were also very closely equalized, but we ended up completely ignoring all threat/dps related parts of these tests. We had to do this because we simply could not field the raid necessary to fairly account for all proper raid buffs and debuffs which impact a tank's TPS/DPS. However, for Melaar's trials, we did make sure to give him all appropriate +attack power, crit, etc buffs. Although, I think we were still missing 3% melee crit on the mob for Melaar, meaning his savage defense uptime could potentially have been very marginally higher.

In short: THESE TESTS CANNOT BE UTILIZED AT ALL FOR TANK DPS/TPS COMPARISONS. Most of the tanks didn't do much of a threat rotation at all. I personally know I half-heartedly mashed devastate and shield slam while I rotated defensive cooldowns.

Like before, you can find these logs at:

http://premoguild.com/pwtest/
http://www.worldoflogs.com/guilds/6280/reports/
And the 8/1 testing can be downloaded at http://premoguild.com/publictesting/

Information on Damage Taken

Xav (Warrior)

Warrior Tank Test 1:

Damage taken: 1,980,706
DPS Taken: 3670
Damage blocked: 203,429
Average block value: 2311
Average hit after block: 13293
[59.7% Total Avoid] Missed 11.1%, Dodged 28.4%, Parried 20.3%, Blocked 24.1%
Cooldowns: 5 Shield Wall, 5 Heart of Iron, 9 Monarch Crab, 13 Shield Block, 36.8% Mongoose Uptime

Warrior Tank Test 2:

Damage taken: 1,995,175
DPS Taken: 3701
Damage blocked: 188,898
Average block value: 2361
Average hit after block: 13213
[59.3% Total Avoid] Missed 9.1%, Dodged 28.8%, Parried 21.5%, Blocked 21.5%
Cooldowns: 5 Shield Wall, 5 Heart of Iron, 9 Monarch Crab, 14 Shield Block, 29% Mongoose Uptime


Melaar (Druid)

Druid Tank Test 1:

Damage taken: 1,874,252
DPS Taken: 3480
Damage absorbed: 180,936
Average absorb size (SD): 1675
Average hit taken (counts all Absorbs): 12014
[58.1% Total Avoid] Missed 9.1%, Dodged 48.9%, Absorb 29%
Cooldowns: 8 Barkskin, 5 Heart of Iron, 30% Greatness Uptime, 48.9% Mongoose Uptime

Druid Tank Test 2:

Damage taken: 1,881,469
DPS Taken: 3494
Damage absorbed: 554,700~
Average absorb size (SD): 1658
Average hit taken (counts all Absorbs): 11825
[57.6% Total Avoid] Missed 5.3%, Dodged 52.3%, Absorb 31.5%
Cooldowns: 8 Barkskin, 4 Heart of Iron, 30% Greatness Uptime, 41.5% Mongoose uptime


Devium (Death Knight)

(Blood) Death Knight Tank Test 1:

Damage taken: 2,179,112
DPS Taken: 4044
Average hit taken: 14723
[60.2% Total Avoid]Missed 9.1%, Dodged 30.6%, Parry 20.4%
Cooldowns: 5 Icebound Fortitude, 4 Vampiric Blood, 5 Heart of Iron, 9 Monarch Crab

(Blood) Death Knight Tank Test 2:

Damage taken: 2,090,828
DPS Taken: 3884
Average hit taken: 14127
[60.2% Total Avoid]Missed 10.5%, Dodged 30.4%, Parry 19.4%
Cooldowns: 5 Icebound Fortitude, 4 Vampiric Blood, 5 Heart of Iron, 9 Monarch Crab

(Unholy) Death Knight Tank Test 3:

Damage taken: 2,028,391
DPS Taken: 3759
Average hit taken: 13338
[59.2% Total Avoid]Missed 9.7%, Dodged 30.6%, Parry 19%
Cooldowns: 5 Icebound Fortitude, 4 Bone Shield, 4 Heart of Iron, 9 Monarch Crab

(Frost) Death Knight Tank Test 4:

Damage taken: 2,099,235
DPS Taken: 3889
Average Absorb size: 1570
Average hit taken (after Absorb): 13456
[57.8 Total Avoid]Missed 13%, Dodged 27.3%, Parry 17.6%, Absorb 6.5%
Cooldowns: 5 Icebound Fortitude, 5 Unbreakable Armor, 4 Heart of Iron, 8 Monarch Crab


(Frost) Death Knight Tank Test 5:

Damage taken: 2,123,916
DPS Taken: 3936
Average Absorb size: 1570
Average hit taken (after Absorb): 13528
[58% Total Avoid]Missed 12%, Dodged 29.7%, Parry 16.3%, Absorb 4.8%
Cooldowns: 5 Icebound Fortitude, 5 Unbreakable Armor, 4 Heart of Iron, 8 Monarch Crab

Fayz (Paladin)

Paladin Tank Test 1:

Damage taken: 1,813,802
DPS Taken: 3358
Damage blocked: 256,164
Average block value: 1778
Average hit taken (100% blocked): 12586
[60.9% Total Avoid] Missed 12.5%, Dodged 28.7%, Parried 19.8%, Blocked 39%
Cooldowns: 4 Divine Protection, 4 Royal Seal of King Llane, 9 Monarch Crab, 49.2% Mongoose Uptime

Paladin Tank Test 2:

Damage taken: 1,993,873
DPS Taken: 3697
Damage blocked: 287,188
Average block value: 1783
Average hit taken (100% blocked): 12384
[57% Total Avoid] Missed 8.3%, Dodged 28.6%, Parried 20.1%, Blocked 43%
Cooldowns: 5 Divine Protection, 5 Royal Seal of King Llane, 9 Monarch Crab, 56.3% Mongoose Uptime

[Xav]

Seeing as how fighting Faerlina this way is almost half the damage of the last tank test, (in terms of individual swing power), it makes sense to see the Block tanks doing a lot better.

Death Knights appear weak nearly all the way around, but I would have been interested to see how it changes with hits dealing significantly more individual damage.

Warriors as usual the spikiest, blocking for the most per hit, but not as reliably. Overall damage taken was a bit below the DK, with the Paladin very close as well.

Druids did great, Paladin appears to have been about middle of the pack (one slightly low avoidance run, one slightly high, can only assume the average between them, to match the ~59% avoidance elsewhere)

While these parses are a fair bit more useful due to the longer test durations, and the much more closely standardized gear, it still doesn't paint the clearest picture.

If you look at only the "average hit taken" figure (which is mostly unaffected by avoidance RNG), there's actual clearer differences between the classes:

(averages)
DK at ~13500
Warrior ~13200
Paladin ~12500
Druid ~11900

It may be possible to work out appropriate damage reduction/scaling had the hits been for more, since all relevant data is there - individual swing damage (thus you can work out her unmitigated damage swing, make it larger, apply it to the perceived damage reduction each tank had in each trial, apply their block value, etc)

[Xav]

OK, so I was going to sleep, then I got curious, so I tried to start that "reverse engineering" project I mentioned. Basically I took the numbers, accounted for the blocks, absorbs, etc, stances, raid buff (ancestral fortitude pretty much, the 10%).

Then I did what I think is "close enough" math. There's a good chance I messed some of this up since I'm kind of terrible at math.

WARRIOR
2336 avg block by warrior
13253 avg hit after block
55% of all hits taken were blocekd
2336 * .55 = 1284, + 13253 =
14537 average hit

51k base * .9 *.9 * .35 = 14458


DRUID
1666 avg absorb by druid
11919 avg hit after absorb
13585 avg hit

72% of all hits taken were absorbed
1666 * . 72 = 1200 + 11919 =
13118 average hit

51k base * .88 * .9 * .328 = 13248



PALADIN
1780 avg block by paladin
12485 + 1780 = 14265

51k base * . 94 * .97 * .97 * .35 = 14208


BLOOD DEATH KNIGHT
14425 average hit

UNHOLY DEATH KNIGHT
13338

Taking base damage from 51k to 100k results in

25740 base hit for Druid
28350 base hit for Warrior
27860 base hit for Paladin
28350 base hit for Blood DK

After block/absorbs:

Druid: 25740-1200 = 24540 avg hit
Warrior: 28350-1284 = 27066 avg hit
Paladin: 27860-1780 = 26080 avg hit
Death Knight: 28350

Not doing it for all the DK specs atm. Just using a spreadsheet to fiddle with numbers and seeing if they make sense.

Possible flaws and stuff is if the druid's armor wasn't actually ~34500. Melaar is sleeping so I don't know how close that is!
Warriors/Paladins/DK's all have armor very close to each other at 31k.