Survival Breakdown and Evolution

[Satorri]

This started in a different form and has evolved into a new pet project for me. I'm working on a way of intelligently constructing a breakdown to compare the value of each survival tool in the big picture with the hopes of creating a fairly easy but not too reductive system for tanks to consider the relative values of their choices on survival talents and stat distributions. This will take lots of time to refine, but I appreciate feedback and input so long as it is constructive.

Negativity and un-supported Op-Ed comments about how X class is broken, or Y class is overpowered only serve to derail the continued education of our community as tanks. Please take those threads elsewhere or stop posting them altogether.

Here is the starting point for this which I hope to maintain as the backbone of the evolution. This first breakdown is based on the following components:

• Simple Boss Melee Swings
• Baseline Passive Survival Stats (Health, Mitigation, Avoidance)

This is not meant to be all-encompassing, but this is the easiest portion to generalize about. This should also be a very big portion of the picture, though not at all the complete and tangled web. While this does not express the whole, the comparison in terms of changing a stat within a particular exemplar is meaningful.

**This is not a comparison of the classes. While similar in level, the four tanks presented have different gear and have made their own choices on how to set it up. All 4 classes have been added and expressed as such to be sure this is accessible to all 4 classes as it evolves.**
================================================== =====

So after playing with chardev.org and having some wonky errors with a couple of the sample characters, I tried Rawr, and I'm liking the numbers I'm getting there. I came up with the following fully raid buffed stats. I will use the following breakdown to create a table including the final net gain from raid buffs. The buffs included are as follows and assumed equal for all tanks:

Stoneblood Flask, Well Fed buff (I just used Fish Feast, each tank would likely use something different depending on what they wanted, hit, exp, agi, strength, etc), Prayer of Fortitude, Gift of the Wild, Blessings of Kings, Sanctuary, and Might, Strength of Earth/Horn of Winter, Commanding Shout, Devotion Aura, Scorpid Sting/Insect Swarm, and a de-haste (Tclap, Frost Fever, JotJ, Infected Wound). AP reducing abilities will be left out because I'm not 100% on how to apply it correctly.

<<<<<<<<<EDIT: Original pre-3.2 values have been bumped to the bottom of the page. 3.2 details will follow here. >>>>>>>>
Here are the raw stats first (note, some gear has changed, updated ilvls will be posted):

                      Warrior       Paladin         Death Knight       Druid
Health                 48,938        49,501           49,357          56,319
Armor                  30,333        29,550           30,818          32,744
Miss                    8.82%         8.69%            9.53%           7.33%
Dodge                  25.78%        24.49%           27.69%          45.41%
Parry                  18.32%        17.97%           19.51%           0.0%
Block (%)              18.35%        17.75%            0.0%           45.00%
Block (V)              1,773         1,605              0             1,941
==============================================================================
Armor Rdx (%)          64.58%        63.98%           64.94%          66.31%
Total Avoidance        55.92%        54.15%           59.73%          55.74%
Total Dmg Rdx          12.7%         14.2%            12.5%           14.6%
Average iLvl           233.4         227.8            231.1           226.8

*New adjustments with new gear! Pally libram will approximate +2.5% dodge on average uptime/magnitude.

The new survival pie:

Warrior (avg ilvl = 233.4)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         55.92%      11,020    46.6%
Armor Rdx         64.58%       3,903    23.7%
Misc Dmg Rdx      12.70%       3,407    1.65%
Block Rdx         738/sw       3,272    0.45%
===============================================
Total Inc      17.8k/2.4s      3,272    89.1%
Hit String     44.08%(1)   19.43%(2)   8.56%(3)
Health            48,938

Paladin (avg ilvl = 227.8)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         54.15%      11,463    45.1%
Armor Rdx         63.98%       4,129    24.4%
Misc Dmg Rdx      14.20%       3,543    1.95%
Block Rdx        1672/sw       3,223    1.07%
===============================================
Total Inc      16.9k/2.4s      3,223    89.3%
Hit String     45.85%(1)   21.02%(2)   9.64%(3)
Health            49,501

Death Knight (avg ilvl = 231.1)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         59.73%      10,068    49.8%
Armor Rdx         64.94%       3,530    21.8%
Misc Dmg Rdx      12.50%       3,088    1.47%
Block Rdx           0/sw       3,088    0.0%
===============================================
Total Inc      18.4k/2.4s      3,088    89.7%
Hit String     40.27%(1)   16.22%(2)   6.53%(3)
Health            49,357

Druid (avg ilvl = 226.8)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         55.74%      11,065    46.5%
Armor Rdx         66.31%       3,728    24.5%
Misc Dmg Rdx      14.60%       3,184    1.81%
Block Rdx         1973/sw       2,820    1.21%
===============================================
Total Inc      15.3k/2.4s      2,820    90.6%
Hit String     44.26%(1)   19.59%(2)   8.67%(3)
Health            56,319

================================================== ==========================================
================================================== ==========================================
================================================== ==========================================

<<<<<<<<<<<<<<<<< Pre-3.2 values are below, preserved for comparison and historical reference. >>>>>>>>>>>>>>>>>>>
Here are the raw stats first:

                      Warrior       Paladin         Death Knight       Druid
Health                 46,518        47,911           49,685          53,469
Armor                  28,734        29,463           30,829          31,924
Miss                    9.0%          8.6%             9.1%            7.6%
Dodge                  30.00%        27.92%           29.49%          48.95%
Parry                  19.19%        18.19%           17.49%           0.0%
Block (%)              17.64%        14.48%            0.0%           40.30%
Block (V)              1,400         1,484              0             1,665
==============================================================================
Armor Rdx (%)          63.33%        63.91%           64.95%          65.74%
Total Avoidance        61.16%        57.71%           59.08%          59.55%
Total Dmg Rdx          12.7%         14.2%            12.5%           14.6%

These are the values I want to use for some work going forward, so I'll keep them here for posterity. I want to build some more practical illustrations of boss survivability, even if it loses the human factor (which is one of the most important and should have nothing to do with how the classes are laid out in game mechanics).

Again, for our first example, we have a boss doing melee swings only. He hits for 60,000 before mitigation on a 2.0 base swing timer.

Small stuff adjusting from the base above. The protadin's Redoubt will be averaged out into his average block chance (+12.5% = 26.98), his tome will also be factored in assuming 100% Holy Shield uptime (+272 bv = 1756). The warrior's block value will be adjusted using crit chance to create an average block value (final bv with 30% crit chance = 1820). The DK's relic is assumed at about a 75% uptime (so average of +1.62% dodge). Again, the Bear block numbers above are derived from the Bear's crit chance, swing speed, and the swing speed of the boss to get an average factor per swing.

The survival pie breaks down like this:

Warrior (avg ilvl = 227.8)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         61.16%       9,710    51.0%
Armor Rdx         63.33%       3,561    20.5%
Misc Dmg Rdx      12.70%       3,109    1.51%
Block Rdx         827/sw       2,975    0.45%
===============================================
Total Inc      18.4k/2.4s      2,975    90.1%
Hit String     38.84%(1)   15.09%(2)   5.86%(3)
Health            46,518

Paladin (avg ilvl = 225.2)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         57.71%      10,573    48.1%
Armor Rdx         63.91%       3,816    22.5%
Misc Dmg Rdx      14.20%       3,274    1.81%
Block Rdx         1120/sw       3,076    0.66%
===============================================
Total Inc      17.5k/2.4s      3,076    89.7%
Hit String     42.29%(1)   17.88%(2)   7.56%(3)
Health            47,911

Death Knight (avg ilvl = 222.5)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         60.70%       9,825    50.6%
Armor Rdx         64.95%       3,444    21.3%
Misc Dmg Rdx      12.50%       3,014    1.44%
Block Rdx           0/sw       3,014    0.0%
===============================================
Total Inc      18.4k/2.4s      3,014    90.0%
Hit String     39.30%(1)   15.44%(2)   6.07%(3)
Health            49,685

Druid (avg ilvl = 223.7)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         59.55%      10,113    49.6%
Armor Rdx         65.74%       3,465    22.2%
Misc Dmg Rdx      14.60%       2,959    1.69%
Block Rdx         1659/sw       2,679    0.17%
===============================================
Total Inc      15.9k/2.4s      2,679    91.1%
Hit String     40.45%(1)   16.36%(2)   6.62%(3)
Health            53,469

More will come with time, particularly as I develop new metrics and find an effective way to add other encounter mechanics to this consideration. For those who think this is too reductive to be meaningful, consider for a moment that in a given night in Ulduar 25, the above breakdown applies to 65-75% of your incoming damage depending on what fights you're doing and what gear/buffs you use for mitigation of and resistance to magic effects.

[Satorri]

<<I've removed the previous speculation post about going into 3.2 and bumped up topics from below>>

I want to start sketching out the next step now, though I will wait to fill in numbers as the dust settles and certain elements shift into their new homes with patch 3.2. The next level I want to add to the breakdowns is on-use and proc survival abilities.

In general we will split the abilities into the following 3 categories (I'll use my own terminology so I can short-hand:

1.) Protection Buffs. These are abilities that reduce incoming damage by a percentage for a duration. These abilities stack though each effect (including the passive buffs present from long-term buffs and stances already included) diminish each individual effects actual benefit. In other words if you have 40% reduction from one buff and 60% from another, you will not become damage immune. Each component is applied one after the next so if the buff effects were P1, P2, and P3, the final reduction would be 1-[(1-P1)(1-P2)(1-P3)]. Or as a direct example, 50% reduction from armor, 10% reduction from stance, and 40% reduction from Shield Wall would total as follows Ptotal = 1-[(1-0.50)(1-0.10)(1-0.40)] = 0.73 = 73% reduction total.

2.) Survival Mechanic Buffs. Each class has its own short CD, or high uptime buffs that they rely on to benefit their survival, as well as longer CD effects that can improve their passive survival tools. Particular examples are Shield Block for Warriors, and Holy Shield for Paladins. These buffs are made to be used frequently, but they are balanced around the fact that the tank must actually activate them. Doing so smartly and well-timed they will provide a greater benefit than if they are used blindly or foolishly.

3.) Supported Survival Buffs. These special buffs work by improving the tank's survival in a different fashion than just increasing passive survival traits. These are the hardest to factor in a metric like this, but I will be working towards finding a smart way to do it. This category includes effects that increase the tanks incoming heals, or heal the tank directly.

I will first enumerate the buffs for each class (and please let me know if there is something I appear to have missed). In order to illustrate the values of these abilities, I want to sketch the nature of the buff, and sketch limits of sorts. Maximum benefit (i.e. maximum possible uptime) is important, though it is hard to measure actual use in general and as we move towards the most specific model the application will become extremely encounter specific. The way I will present the abilities is as follows:

Ability Effect (with duration, CD, and specific effect)
Maximum Value (essentially using it on CD against the melee-only boss model thus far)
Synergy/Inhibition (what other survival mechanics improve the effect, which diminish it)

This breakdown will go by class, and will take time to develop, but I will keep updating here as I do the work. For abilities with glyphs and talents to change the effect, I will illustrate each of the permutations. I will not include gear set bonuses yet, as they may come and go and you will not likely use a lower level set bonus thanks to the heftier survival bonuses from later sets. I will also not be factoring in Spell damage survival mechanics yet, these will come with a later iteration that handles magic and stacking multiple damage input considerations (i.e. boss melee and special encounter mechanics).

Warrior

• Shield Block: Increases Block Chance and Block Value by 100% for 10 sec on a 1 min CD. Talents reduce the CD by 10/20 sec.
• Shield Wall: Reduces all incoming damage by 60% for 12 sec on a 5 min CD. Talents can reduce the CD by 30/60 sec, and the glyph makes the reduction 40% and reduces the CD by an additional 2 min.
• Last Stand: Increases total health by 30% for 20 sec on a 3 min CD. Glyph reduces the CD by 1 min.
• Glyph of Blocking: Increases Block Value by 10% for 10 sec after each Shield Slam. It will be assumed that in the grand average a Warrior Tank will Shield Slam more than once per 10 sec, so this will have a 100% uptime.
• Enraged Regeneration: Restores 30% of the Warrior's health over 10 sec, on a 3 min CD, and while active it blocks other enrage effects (+10% damage and/or bonus rage generation). Glyphing increases the total heal by another 10% of the Warrior's max health.

Paladin

• Holy Shield: Increases Block Chance by 30% for 10 seconds, or 8 blocks. On 8 second CD.
• Divine Protection: Reduces all incoming damage by 50% for 12 sec on a 3 min CD. Talents can reduce the CD by 30/60 sec.
• Glyph of Divine Plea: So long as Divine Plea is active, the Paladin takes 3% less damage.
• Glyph of Hand of Salvation: When the Paladin uses Hand of Salvation on himself, he will take 20% less damage. (Hand of Salv reduces the targets total threat by 2% per second for 10 sec, on a 2 min CD).
• Lay on Hands: the Paladin heals the target for an amount equal to the Paladin's maximum total health, and restores 1950 mana. Talents can reduce the CD by 2/4 min, and will grant the target 10/20% damage reduction for 15 sec, while the glyph will reduce the CD by 5 min.
• Ardent Defender: Any attack that takes the Paladin below 35% health is reduced in damage. The damage portion that takes the Paladin's health below 35% is reduced by 10/20/30%. In addition, when the Paladin receives damage that would kill him, he instead is pushed up to 10/20/30% health (assuming 540 def tank), on a 2 min CD.

Bear

• Barkskin: Reduces all incoming damage by 20% for 12 sec on a 1 min CD.
• Survival Instincts: Increases the Bear's maximum health by 30% for 20 sec, on a 3 min CD. Glyph increases the health increase by an additional 15% of max.
• Frenzied Rejuvenation: The Bear converts 10 rage per sec for 10 sec, into health, on a 3 min CD. The exchange is 0.3% health per point of rage, or 3% per 10 rage. Glyphed this ability increases all incoming heals on the Bear by 20% for the duration.

Death Knight

• Icebound Fortitude: Reduces all incoming damage by an amount based on your defense (we'll assume 540 def, or 40%) for 12 sec on a 2 min CD. Talents can increase the duration by 3/6 sec (Frost).
• Death Strike: In addition to dealing damage, heals the DK for 5% of his maximum health per disease on the target. Talented from Blood this becomes 7.5% max health per disease, and Unholy can talent for a third disease. This is only frequently used for Blood tanking, but is accessible to other tanks.
• Vampiric Blood (Blood): Increases the DK's health by 15% and increases all healing affects on the DK by 35% for 20 sec, on a 2 min CD. Glyphed the duration is increased by 10 sec.
• Rune Tap (Blood): Instantly heals the DK for 10% of his total health on a 1 min CD. Talenting reduces the CD by 10/20/30 sec, and increases the healing by 3/7/10% of max health. Glyphing increases the heal on the DK by 10% (so to 11/14/19/22% of max), and causes it to heal the DK's group for 10% of their max health.
• Mark of Blood (Blood): The target heals anyone it hits for 4% of their max health with every hit. Lasts for 20 sec or up to 20 hits, on a 3 min CD.
• Imp Blood Presence (Blood): The DK heals himself for 2/4% of his damage dealt.
• Will of the Necropolis (Blood): Damage that would take the DK below 35% health is reduced. The damage portion that reduces the DK's health below 35% is reduced by 5/10/15%. This effect cannot be trigger by a hit smaller than 5% of the DK's max health, and the proc has a 15 sec CD.
• Unbreakable Armor (Frost): Reduces all incoming hits on the DK by an amount equal to 5% of the tank's armor and increases the DK's strength by 25% for 20 sec, on a 2 min CD. The Glyph increases the damage absorbed by 20% (increasing the value to 6% of the DK's armor total).
• Frigid Dreadplate (Frost): Increases the DK's chance to be missed by 1/2/3%.
• Imp Frost Presence (Frost): Increases the passive damage reduction of Frost Presence by 1/2% (to a total of 6/7%).
• Bone Shield (Unholy): Activates a shield around the DK with 4 charges. So long as one charge remains the DK will take 20% less damage from everything, and each time the DK takes damage a charge will be consumed (no more than one per 2 sec). Lasts 5 min or until charges are gone, on 2 min CD. Glyphed this shield has 2 extra charges.

Note 1: I've left out Magic-oriented CDs and effects for the time being, as they will be factored in with a later layer.
Note 2: I've included itemized passive bonuses for non-uniform procs and some standard passive items for DK's since each tree will function differently as a tank.
Note 3: Self-heals have been included as they will become very important for particular class-specs, though hard to discuss in a general breakdown of survival. This will be a later layer as well.
================================================== ================================================

The first and easiest to incorporate is the Paladin Holy Shield buff. Essentially, the only way to not have 100% uptime on this buff is to take more than 1 hit per sec, which will never happen against a boss which is our current measure. Furthermore magic damage items won't interact with or consume charges so this will remain static as that layer is added. So, we can assume, given a diligent protadin, that this will be 100% uptime of 30% additional block chance. We can make the same assumption of Glyphed Divine Plea as it should always be active provided the Protadin activates it and pays attention, that will be a permanent 3% damage reduction.

Shield Block for Warriors is not quite as easy. When it is active it will double block value (2.6 x actual sheet block value if you account for crit chance), and will make every hit that is not avoided a block (effective block chance becomes 100%-Avoid%). With standard Warrior tank talents, the ability will have a 40 sec CD with a 10 sec uptime, which, if used on CD is 25% uptime. That said, most warriors will not pop it on CD, they will use it strategically (often for threat). So, the maximum effect would be 25% uptime, but the actual effect will depend on use. I will treat this as an on-off condition, and examine the relative and average effects later on.

For the standard protection abilities, the classes break down as follows, and I will add in some granted effects that healers/raid can provide as well (I'll provide stats consistent with our exemplar tanks' talents and glyphs):

• Warrior: Shield Wall = 40% dmg rdx, 12 sec duration, 2 min CD
• Paladin: Divine Protection = 50% dmg rdx, 12 sec duration, 2 min CD
• Paladin: Glyphed Hand of Salv = 20% dmg rdx, 10 sec duration, 2 min CD
• Death Knight: Icebound Fortitude = 40% dmg rdx, 12 sec duration, 2 min CD
• Bear Druid: Barkskin = 20% dmg rdx, 12 sec duration, 1 min CD
• Inspiration/Ancestral Fortitude: (crit heal from talented Priest/Shaman healer) = 10% physical dmg rdx, 10 sec duration
• Pain Suppression (Disc Priest) = 40% dmg rdx, 8 sec duration, 3 min CD

Considering these abilities, the 3 plate wearers have a maximum of 10% uptime while the bear has a maximum of 20% uptime. The Paladin will also have the Hand buff available at a 8.3% uptime. Again, like Shield Block these abilities are often saved for encounter specific peaks, but we will gauge that as the model develops.

Situational damage reductions (specifically Will of the Necropolis and Ardent Defender) are hard to work into a general model of an average. They will become meaningful when they save the tank and reduce damage, but the conditions of triggering them are a little too tricky to generalize about. I will attempt to demonstrate some of their value, if I can, later on.

Similarly health maximum bumps can be hard to gauge smartly. I can express them in terms of average maximum health, but that's not a particularly meaningful number. I will find a way to illustrate these as an on-off scenario as well.

Tank self-healing can become a very serious value, but it needs to be handled with care and a great deal of marked assumptions and conditions. In general, tank self-healing can amount to functionally appear as less damage taken, but it needs to be factored in terms of over-healing and other player/human-related factors. I will create a specific special consideration for this aspect as it becomes particularly meaningful for Blood DKs, and somewhat less so for Bears and Warriors. Interestingly the one tank that can heal while tanking, Paladins, also relies the least on their own healing to survive. In tank self-healing I will also discuss the effects of Imp Leader of the Pack and Judgment of Light, which I have excluded to this point as they can be available to all of the tanks.
================================================== ======

So, first as a baseline, let's refresh the base stats, including the following assumptions that will buff the baseline stats:

Warrior = crit block chance causes actual block value to average 1.3xbase

Paladin = 100% Holy Shield and Divine Plea uptime equate to +30% block (actually included in the first post) and 3% dmg rdx always. Based on boss swing speed (2.4 sec) we're assuming Redoubt is in effect roughly 41.7% of the time, multiplied by the increased block chance will average +12.5% block over time.

Death Knight = Based on an assumed 40% turnover rate from melee swings to RS's, and an average weapon speed of 3.0 sec after buffs, the sigil will be assumed to have a 66.7% uptime for an average buff of 1.47% dodge.

Bear Druid = Based on the Bear's raid buffed crit chance (~45%), attack speed (2.39), and the boss's incoming swing speed (2.4), the Savage Defense proc will be roughly a 44.88% block chance, with an absorb value 25% of the bear's raid buffed AP (1941).

The adjusted raw data will be as follows:

[Satorri]

(reserved 1)

[Satorri]

(reserved 2)

[TomHuxley]

Fantastic work so far, I'm looking forward to the rest of it.

[junkilo]

nice work Satorri!

I'm sure this is coming, but can you post a little more info on the gear? aka how many pieces of tier, what types of non-tier gear was used, which trinkets etc. It sounds like you used real armory's for this work no?

[Satorri]

Sorry, should've mentioned this before, here are profiles of the cloned gear/talent sets:
Sample Warrior
Sample Paladin
Sample Death Knight
Sample Druid

These are all based on real high-end tanks, but I ask that even if you can figure out who the toon belongs to that you not discuss it here. Respect people's privacy, especially since I didn't ask half the list if they minded me using their setups. =)

It's also inconsequential who the people are...

[Eetabeetay]

The warrior you chose is a very poor example. Gemming expertise is a poor choice for progression, and in most cases warriors shouldn't be wearing 4 peices of T8 because only one of those pieces is actually BiS (I said most cases because the 4pc bonus is very situational and should be used as such). More common is one or two peices depending on what other pieces you have to choose from, but I find it hard to believe that someone who has done a few hard modes couldn't get at least a better helm by now. I would highly suggest using Xav as your example instead of this person.

edit: I fixed the major gearing problems while keeping the iLvls the same. chardev.org v6 ~ a World of Warcraft character planner

[jere]

I might have missed it, but how did the warrior's block value drop to 840? Is that adjusted for uptime, crit, and so on I take it?

[Irat]

Great post again Satorri, looking forward to the rest of it.

It might be worth keeping in mind glyphs too, I’m not sure about the others but:

http://www.wowhead.com/?item=45745

Is basically compulsory for Pallys and provides a massive boost to DR.

[Tengenstein]

I might have missed it, but how did the warrior's block value drop to 840? Is that adjusted for uptime, crit, and so on I take it?

I think you're comparing the average block reduction per swing to the block value.

for 3.1 the case-warrior's BV is 1400, and its dropping the damage on each swing by 827 on average

in 3.2 the case-warrior's BV is 1920 and its dropping the damage on each swing by 840 on average


Or maybe not, i've had to take a couple of intuitive leaps to keep up with this so i could be quite wrong.


EDIT: yeah i'm definetly going wrong somewhere as i can't make those BVs line up with the avg swing reductions satorri has

[Satorri]

Eeta, you're very confident that you have it right, but the warrior tank exemplar I'm using has tanked through the current content, hard modes included. I'm happy standing by the choices of the exemplar. If you'd rather use someone else's set/spec or a custom created character, go to town, you can use my method to do the math out yourself. My goal is to use real tanks in who's judgment I trust. Also, gemming for expertise is anything but a bad idea if you care about threat as a Warrior, and the survival value doesn't hurt at times either.

Jere, sorry I did the explanation in a different thread. Here's how I computed Block for this metric:

Since I'm looking at the average effect of block per swing (so as to keep this statistical and general, not a simulation or time frame limited), I want the average block value per incoming hit. This means, in plain English, we're averaging the amount blocked over every hit that lands so the Block per swing is also a function of avoidance. Here is the formula:

Block Value x Block Chance x 100% = Block per Swing
 (100% - Total Avoidance)

Also, the Bear's Block "chance" and "value" are based on the melee auto-attack speed of the bear, and the fully raid buffed melee crit chance, and the boss's incoming attack rate. The value may be slightly conservative because of special ability crits but given the relatively small % contribution of Block I'm not going to squibble over a few hundredths of a percent.

For Tengenstein:
Sheet BV = 1400 (3.1), 1477 (3.2) *only a single piece with block value on it
Crit adjusted average (30% of all blocks at double value) = 1820 (3.1), 1920 (3.2)
Block Chance = 17.64% (remains the same)
Avoidance = 61.16% (3.1), 59.69% (3.2)

So, 3.1 BV/swing = (1820*0.1764)/(1-0.6116) = 827
And, 3.2 BV/sw = (1920*0.1764)/(1-0.5969) = 840

The increase in block value increases the amount, the reduction in avoidance reduces it slightly. You'll notice the big loss of avoidance for the Bear actually reduces the average block value per swing by quite a lot, though the block value remains the same. That's purely an matter of how I am expressing it here, as if every swing not avoided is being blocked, essentially, just with a value adjusted down.

[Satorri]

Also, since 3.2 went through today and tank choices and gems will likely change, I'm going to be updating this in the next couple days with the re-adjusted values from the tanks current gear sets. I will likely stick with these gear sets for a while though until they have all had the chance to pick up some t9 level gear so that the levels don't get too diverse.

[Eetabeetay]

Eeta, you're very confident that you have it right, but the warrior tank exemplar I'm using has tanked through the current content, hard modes included. I'm happy standing by the choices of the exemplar. If you'd rather use someone else's set/spec or a custom created character, go to town, you can use my method to do the math out yourself. My goal is to use real tanks in who's judgment I trust. Also, gemming for expertise is anything but a bad idea if you care about threat as a Warrior, and the survival value doesn't hurt at times either.

I realize this isn't the venue for discussing how to gear as a warrior, but I feel like it will affect the outcome of the numbers if you are comparing one person who gears correctly to one who doesn't. Just because someone has tanked through hard mode content doesn't mean they know how to gear properly. I know he probably is good at tanking, and by that I mean picking up mobs, holding aggro and being positioned right etc. Tanking is different than dps though, in that gear choice doesn't make an easily noticeable difference. The reason I dislike gearing for expertise for hard mode ulduar is that you shouldn't need it. If your rogues and hunters are doing their jobs in keeping MDs on CD and your tank can build aggro properly and your raid knows how to use their threat management tools correctly, then there is no reason that expertise should ever be prioritized over survivability. Since you are presenting this to the tankspot community, I think it would be proper to use a tank that the tankspot community trusts, not just you. Again, I do not wish to derail this thread. If you or anyone else wishes to discuss this issue further, we could take the discussion to a new thread or to PMs. I appreciate all the work you have done on these forums, and I always look forward to reading your posts.

[Satorri]

I realize this isn't the venue for discussing how to gear as a warrior, but I feel like it will affect the outcome of the numbers if you are comparing one person who gears correctly to one who doesn't. Just because someone has tanked through hard mode content doesn't mean they know how to gear properly. Since you are presenting this to the tankspot community, I think it would be proper to use a tank that the tankspot community trusts, not just you.

You're using loaded language. "Gear correctly" suggests there is a right way and a wrong way. You've already parted trails with me. That said, bearing in mind that I only trust this warrior (and you do as well, promise), gearing for Expertise is not a bad idea at all for a warrior doing hardmode work. Expertise has a very positive effect on your threat/damage (the latter of which becoming more important in hardmodes), and still has some small survival buffing. If you would make a different choice, that's fine, but I wouldn't call the exemplar into question for everyone because of it.

Besides, you assume that I'm not using tanks who are prominent community members, not just well-geared? Part of why I trust a tank as an example is BECAUSE they talk the talk, not just show up with good gear.

Like I said though, who the tank actually is, is not of consequence. If you disagree with their choices, that's fine, you can plug in your own values to see how you fair next to them on the breakdown. The tank just happens to be an excellent choice for comparison with the other exemplars.

[geros]

So this is your druid (pre-avoidance nerf):

Druid (avg ilvl = 223.7)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     0.0%
Atk Spd Rdx      60k/2.4s     25,000    16.7%
Avoidance         59.55%      10,113    49.6%
Armor Rdx         65.74%       3,465    22.2%
Misc Dmg Rdx      14.60%       2,959    1.69%
Block Rdx         1659/sw       2,679    0.17%
===============================================
Total Inc      15.9k/2.4s      2,679    91.1%
Hit String     40.45%(1)   16.36%(2)   6.62%(3)
Health            53,469

Here's a druid geared/socketed with the intention of keeping ~50k hp. I didn't fill out your %'s, and I got a bit lost at the end with the misc and block. I arrived at 1716 per swing by taking the blocked amount 1889 * the uptime of 90.858% determined in rawr.

The philosophy being that ~50k is enough hp to keep you alive under almost all scenarios tanking. Once you've hit the EH plateau, work avoidance into the set. I think it sort of highlights why the avoidance nerf is coming. At some point we were going to get near sunwell good again. Not immune to damage good, but ridiculously hard to kill with physical damage good.

Druid (avg ilvl = 223.7)

                  Factor      iDPS      % rdx
Unmitigated      60k/2.0s     30,000     
Atk Spd Rdx      60k/2.4s     25,000    
Avoidance         63.936%      9,016    
Armor Rdx         70.998%       2,614    
Misc Dmg Rdx      14.60%       2,232    
Block Rdx         1716/sw       1,932    
===============================================
Total Inc      15.9k/2.4s      1932    
Hit String     36.06%(1)   13.01%(2)   4.69%(3)
Health            49,197

If I did my math correctly, 3 consecutive hits would be 44,373 to the druid. so 49,167 is more than enough to survive it. 53k isn't enough to survive 4 consecutive swings.

I can't save out to a profiler atm, but this is the jist of it.

Head    Garona's Guise (austere, solid, stalwart prot)
Neck    Mark of the Unyielding
Shoulders    Shoulderpads of the Intruder ( solid, solid dragon, ins of glad)
Chest    Embrace of the Gladiator (glinting, solid, shifting, powerful stats)
Waist    Death-warmed Belt (solid, shifting, solid)
Legs    Conqueror's Nightsong Legguards (solid, solid dragon, frosthide)
Feet    Runed Ironhide Boots (solid, greater fort)
Wrist    Solar Bindings (glinting, major stam)
Hands    Conqueror's Nightsong Handgrips (solid drag, major agi)
Finger1    The Leviathan's Coil
Finger2    Fate's Clutch (vivid)
Trinket1    Heart of Iron
Trinket2    Darkmoon Card: Greatness
Back    Drape of the Faceless General (shifting, glinting, major agi)
MainHand    Dreambinder (glinting, solid, mongoose)
Ranged    Idol of the Corruptor

[Tengenstein]

Cheers, i totally forgot to factor in the affects of parry and dodge.

[Nahmette]

Satorri, thank you for the excellant job so far.I happen to agree with your model warriors Gearing choices....and as you've said, gear choices are just that...choices that every individual can do differently.

A question did occur to me...will you be adjusting anything to account for current glyphs?....i believe that those have a huge impact on survivability also but i do realize it would be difficult to incorporate.

[TomHuxley]

FWIW I'll gem for expertise is I'm below the dodge ("soft") cap, as it has a noticeable effect on my threat gen and of course is also decreasing parries. I don't like to do it, but sometime that's just the way the numbers fall.

Anyways, I had one warrior-centric mechanic I wanted to point out: you can macro a Last Stand/Enraged Regeneration combo that results in what is functionally a stronger CD (you get the LS buffer plus extra healing since you have more life at the time of calculation) but of course at the expense of less overall CD uptime. On some fights this is a powerful combination...it's one of my standard "cooldowns" when eating a Mimiron plasma blast, for example.

I don't know if other classes have synergistic CDs like this, but if so they can allow for greater survivability against specific burst encounters.

[Satorri]

Nahmette, yes, I'm including defensive-oriented glyphs in the next section I'm developing. I haven't been posting updates to bump the thread, but I probably will when I finish an increment. =) You'll see the above sections change and expand.

And Tom, explain this, "dodge soft cap" to me. =) I've heard lots of people mention it, but I find the terminology misleading. People seem to think there is some magical turning point on the curve where dodge suddenly becomes less valuable...