Are my mitigation numbers off ?

[Freyyy]

This has already been discussed a while ago. The only times where PLD were the "mitigation king" was before 2.1. After the WAR fix, it ended. Yoshida is pretty much the only one still believing that PLD is the "mitigation king" or something like that. All of your calculations are good and have been done a while ago. That's why we insist that PLD needs a buff. They're the worst tank in every domain, literally.

[Exodus_Kenpachi]

A huge thing you are missing on the WAR end is Bloodbath, Bloodbath+Berskerk+maim, Equilibrium+Berserk, and other self healing abilites. Over the duration of a fight the amount of self healing WAR provides is a little obscene. If you look at mitigation as "HP that needs to be healed back" instead of "total damage reduced" the gap widens even more between WAR/DrK and PLD.

Additionally if you look at eHP as the same as mitigation (think Defince vs Shield Oath) you left out ToB+Convo which roughly translates to a 20 seconds of 18% mitigation.

[Lyrica_Ashtine]

I'll nitpick here:
passive mitigation variation with Shield Oath/Grit in. Warriors do get increased healing what brings their eHP close to other tanks combined with Storm's Path, but we're talking about "mitigation" here (also see below)
Warrior GCD/stack uses is wrong. Yes, you get 16 stacks after 24 GCDs. No, it's not 16 stacks in practice after 24 GCDs (Inner Beast also consumes a GCD). It's actually 8,5 GCDs for 1 Inner Beast on average and Fracture isn't even taken in consideration here. Berserk and Vengeance can be used as ways to get stacks as well and Unchained also consumes stacks - Although for paper theory, you could leave out Unchained while it's not the case in practice.
Storm's Path is actually for both tanks for content where two tanks are required
-10% stat reduction is roughly -10% damage dealt that's associated with the corresponding stat. There's been some virus testing data somewhere on the healer forum. Although it's kind of buried with all the facepalm topics in there nowadays.

[Reynhart]

Average mitigation isn't everything. PLD uses cooldowns individually and gets the full benefit of them. WAR and DRK stack Inner Beast and Dark Mind with other cooldowns which means that they're less effective. Sentinel itself is 40%, Inner Beast+Vengeance is only 44%, not 50%.

Yes, but for each Rampart, WAR can use one Inner Beast.
And, unless you have two big hit closer than 15s, for each Sentinel, WAR can use Inner Beast+Vengeance.

Since Shield Oath already applies a 20% damage reduction, Sentinel is also not really -40% from "base" damage.

[Dante_V]

The problem with calculating mitigation in this manner is that the actual would tend to be widely variant and per situation and fight based. It doesn't take into account player skill or knowledge and views these classes completely in a vacuum which negates natural synergy with other classes and per fight mechanics. For instance if two Pallys enter their own instance with the same gear and both MT the fight, the way cooldowns are used, team strategies, and just general skill will have varying results in terms of mitigation as a whole making the above wholly unreliable as a means of objective assessment.

Also mathematically speaking this assessment falls short for a few other reasons, the first one being user differential in using cool downs (people don't just blow cooldowns because they come available, rather they tend to use them when it is advantageous in mitigating larger amounts of damage or saving them when necessary making the cooldown time assessment unreliable as a mean of mitigation measurement.) It doesn't account for tank stance which increases mitigation base for pally and drk while increasing parry and hp for warrior. (ehp) It isn't properly accounting for the ultimate cooldowns (which can't be judged within the limited context of these calculations, yet any damage acquired at "1hp" while under the affects of holmgang or living dead is 100% mitigated and depending on the fight this could vary widely with outgoing damage and cooldown usage.) Whether or not foresight is available to all tanks or not it should be considered in the grand scheme of things as the warriors foresight is superior than on other tanks. This also doesn't account for parry/block nor does it account for healer synergy created by things like convalescence which while not mitigation wholly increases survivability of the tank. Last but not least it doesn't account for raid wide mitigation (which would be per fight and is utility) like for instance in final coil, dragon kick (or delirium blade now) was/is one of the highest raid wide mitigation abilities in the game due to the prevalence of magic wide aoe attacks in high level content.

TLR I don't think this is an adequate means of assessing tank mitigation and such things are more emphasized by the particular fight mechanics/boss damage/player ability than just analyzing tank cooldowns in a vacuum. Analyzing in a vacuum is nice when looking at certain things but when trying to assess actuals it comes up extremely short. Player ability should not be understated as active mittgation is 100% player reliant.

[Jpec07]

So then the better way to evaluate it would be to do so on a few levels:

• Tankbusters Only - Assume incoming regular damage at ~200% max HP once a minute, and all other damage inconsequential. This is roughly the way the current tanking meta works, so evaluating each of the tanks' ability to soak this damage (or, to achieve at least 50% total mitigation). This would look at stacking buffs and their interplay to see who can do it best. This damage should be assumed to be magical most (if not all) of the time.
• Consistent High Physical Damage - Since the majority of this game's content consists of medium to high amounts of incoming physical damage, evaluating the tanks' ability to mitigate this damage should also be a focus. For this one, buff stacking would generally be kept to a minimum, and the calculation would need to account for Parry, Block, and self-healing. Incoming damage could be said to be 20% HP/sec. This would basically mimic big trash pulls.
• Mixed Physical/Magic Damage - This evaluation would be similar to the previous, only it would include saving big cooldowns for regular magic damage. So assume 10% HP incoming damage per second, with a once-a-minute magic nuke of 75% HP. This would probably be close to what you'd see in a dungeon context on some of the harder bosses (looking at Stone Vigil for inspiration here).

The amounts I've picked are all ballpark estimates, but would provide a good baseline for evaluating a tank's ability to mitigate incoming damage across a variety of contexts. You could theoretically also do calculations encounter-to-encounter, to evaluate mathematically which tank is best for each, but that would be crazy amounts of maths.

[Reynhart]

A huge thing you are missing on the WAR end is Bloodbath, Bloodbath+Berskerk+maim, Equilibrium+Berserk, and other self healing abilites. Over the duration of a fight the amount of self healing WAR provides is a little obscene. If you look at mitigation as "HP that needs to be healed back" instead of "total damage reduced" the gap widens even more between WAR/DrK and PLD.

Yeah ! More to WAR ! Again...

Yes, you get 16 stacks after 24 GCDs. No, it's not 16 stacks in practice after 24 GCDs (Inner Beast also consumes a GCD). It's actually 8,5 GCDs for 1 Inner Beast on average and Fracture isn't even taken in consideration here.

Yes, it's not really 16 stacks, but you could use one CD for the missing stack every minute. And you can use Inner Beast without breaking your combo, meaning you can probably build one stack immediatly after Inner Beast.
And no, Fracture, is not included since the focus is mitigation. Again, if PLD is supposed to be the best, WAR shouldn't be able to top it, even by sacrificing some DPS.

The problem with calculating mitigation in this manner is that the actual would tend to be widely variant and per situation and fight based. It doesn't take into account player skill or knowledge and views these classes completely in a vacuum which negates natural synergy with other classes and per fight mechanics.

I don't see any reason why a very good WAR would suffer more than a very good PLD or DRK...yet, I see plenty of reasons why a bad WAR would...but we won't make "bad" players the standard.

Also mathematically speaking this assessment falls short for a few other reasons, the first one being user differential in using cool downs (people don't just blow cooldowns because they come available, rather they tend to use them when it is advantageous in mitigating larger amounts of damage or saving them when necessary making the cooldown time assessment unreliable as a mean of mitigation measurement.)

And WAR is probably the one tank that can always have a cooldown ready, if they only focus on mitigation. In the current meta, CD duration is frequently irrelevant, and only recast matters...

It isn't properly accounting for the ultimate cooldowns (which can't be judged within the limited context of these calculations, yet any damage acquired at "1hp" while under the affects of holmgang or living dead is 100% mitigated and depending on the fight this could vary widely with outgoing damage and cooldown usage.)

The problem with Living Dead and Holmgang is that if your HP stays above 1, you basically wasted your CD. That's why I didn't take it into account. But were I to put Hallowed Ground aside too, it would, again, only put PLD even lower.

Last but not least it doesn't account for raid wide mitigation (which would be per fight and is utility) like for instance in final coil, dragon kick (or delirium blade now) was/is one of the highest raid wide mitigation abilities in the game due to the prevalence of magic wide aoe attacks in high level content.

Storm's Path mitigatess at least the same as Dragon Kick/Delirum and Rage Of Halone combined.

Tankbusters Only - This damage should be assumed to be magical most (if not all) of the time.

For that part, there's no need for calculations. PLD loses so much against magical tankbusters that it's not even a competition...which is the actual state of the game, yeaaah...

[bounddreamer]

I took my fresh 60 PLD into EX roulette yesterday with one of my healers and felt so much squishier than WAR. I spent the whole run around 50% health or less. I rotate and use all available cooldowns including Hallowed. Unless they make some changes to the job, I won't be playing it much methinks.

[Februs]

It is, so you'd need to include that in the calculation as well (since the interplay between block rate and parry rate is complicated). So effectively, a parry rate of 25% would only affect those physical hits that couldn't be blocked, so (100-{blockRate})% * {parryRate}%. Assuming 25% for both, that's 25% of 75%, or 18.25% (times .2 is only 3.65% mitigation, versus the 5% it would have been otherwise).

Correct me if I'm wrong, but doesn't Awareness also skew this number? Using it typically increases the efficiency of Bulwark because Crits can't normally be blocked or parry'd... Though I'm not sure how the heck you'd work that into the math...

[Lyrica_Ashtine]

Yes, it's not really 16 stacks, but you could use one CD for the missing stack every minute. And you can use Inner Beast without breaking your combo, meaning you can probably build one stack immediatly after Inner Beast.
And no, Fracture, is not included since the focus is mitigation. Again, if PLD is supposed to be the best, WAR shouldn't be able to top it, even by sacrificing some DPS.

Using Vengeance out of the blue to get a stack back is far from practical, thus that leaves only Berserk. Even if you'd decide to use both to get stacks, disregarding mitigating any tank busters that will come sooner or later, you'll still lose two GCDs out of 24 on IB/Fell Cleave. Considering you're not even taking Oaths, Grit or Defiance in consideration, you might as well just assume the stacks are used on Fell Cleave instead. Which has no mitigation potential at all.

Since you're still not considering Oaths/Grit, I'll do the numbers here for you. To start with, here are the ones you have without any other modifiers:

PLD:

• Rampart : 20*(20/90) = 4.44%
• Sentinel : 40*(10/180) = 2.22%
• Bulwark : (Let's be generous and consider 100% Block rate for the 15s) 30*(15/180) = 2.5
• Sheltron : 30*(3/30) = 3% (3s is close enough to cover "one hit")
• Rage Of Halone : 10% (I thinks its a little lower, but can't find the proper number)
• Hallowed Ground : 100*(10/420) = 2.38%

Total active mitigation : 24,55%

WAR:

• Storm's Path : 10%
• Vengeance : 30*(15/120) = 3.75%
• Inner Beast : 20*(6/15) = 8% (With a GCD on 2.5 seconds, you can do 24 GCD during each Infuriate, so 8 combos. Without any additionnal CD, 8 combos gives you 16 stacks of Wrath, so at least 3 Inner Beast + 1 from Infuriate. So one each 15s on average)
• Raw Intuition : 20*(20/90) = 4.44%
• Since Holmgang is not really a "mitigation" skill, I won't inclue it

Total active mitigation : 26,19%

DRK:

• Shadowskin : 20*(20/90) = 4.44%
• Shadow Wall : 30*(10/180) = 1.67%
• Reprisal : (Since Dark Dance has a 60s recast, and you can at least have one parry on its duration, I'll count one Reprisal per minute) 10*(20/60) = 3.33%
• Dark Dance : (A bit tricky, but with a proper set, you can probably parry half the hits during its duration, so, around 10s of mitigation) 20*(10/60) = 3.33%
• Dark Mind : (All with Dark Arts, since mitigation is the main goal) 30*(10/60) = 5%
• Delirium : 10% (See Rage Of Halone above)
• Walking Dead : (See Holmgang)

Total active mitigation : 27,78%

I won't use the exact same list as some effects can be on nearly permanently and not all mitigation types affect the same attacks:

Paladin
Shield Oath and Rage of Halone included.
Rampart: 36% magic (8%), 42.4% physical (9.42%)
Sentinel: 52% magic (2.89%), 56.8% physical (3.16%)
Bulwark: 0% magic (0%), 49.6% physical (4.13%)
Sheltron: 0% magic (0%), 49.6% physical (4.96%)
Hallowed Ground: 100% magic (2.38%) 100% physical (2.38%)
Active magic: 10.89%
Active physical: 21.67%

Warrior
Vengeance: 28% magic (3.5%) 28% physical (3.5%)
Inner Beast: 19% magic (7.6%) 19% physical (3.5%)
Raw Intuition: 0% magic (0%) 28% physical (6.22%)
Foresight: 0% magic (0%) 17.2% physical (2.15%)*
Active magic: 11.1%
Active physical: 19.47%

*I think it was 8%, someone correct me on that if it's wrong

Dark Knight
Shadowskin: 42.4% magic (9.42%) 36% physical (8%)
Shadow wall: 49.6% magic (2.76%) 44% physical (2.44%)
Reprisal: 35.2% magic (11.73%) 28% physical (9.33%)**
Dark Dance: 0% magic (0%) 36% physical (6%)
Dark Mind: 49.6% magic (8.27%) 0% physical (0%)
Active magic: 32.18%
Active physical: 25.78%

** It's impossible to use Reprisal twice/minute on average, but I'll humor you here.

Tally:
Magic
Dark Knight 32.18%
Paladin 13.27%
Warrior 11.1%

Physical
Dark Knight 25.78%
Paladin 24.05%
Warrior 19.47%

Now that I've finished this pointless task, take this with a grain of salt because:
2 Reprisal/minute is impossible
Convalescence isn't even taken in consideration (also: Mantra)
Awareness isn't taken in consideration
All stacks are used for Inner Beast, not a single one used on Fell Cleave
Storm's path counts for both tanks when the warrior is off-tanking, but not considered in the list
Self healing isn't taken in consideration (IB, Bloodbath, Storm's path, Souleater, Clemency)
Passive mitigation from blocking/parrying isn't taken in consideration
Performance loss for mitigation isn't considered

Performance loss? Yes, performance loss. Warriors lose mitigation ability when they use their stacks on Fell Cleave. Their "active mitigation" drops significantly when it is not spend on Inner beast on anything but tank busters - Which is perfectly fine and what they should do, in my opinion.
Using Dark Arts on anything but Soul Eater or Carve and Spit reduces the Dark Knight's damage output. Using it every 60 seconds possible on Dark Mind eats away on it significantly and the amount of tank busters that requires Dark Mind to be used that frequently is rare.
While Dark Mind is still the subject: Dark Mind does not work on nearly all auto attacks. The only exceptions I can think of at the moment are Ramuh and Manipulator.
Staying in Grit for an extended duration denies the use of Blood Weapon for Dark Knight.

Paladin CDs are pretty much "press this and do that". The major annoyance would be their ability to change oaths, which eats away on a GCD unless they simply toggle off Shield Oath and not go into Sword Oath like Dark Knights do (although that's a bit senseless). Warriors, while they have it easier to switch stances, are stuck in their stances for at least 15 seconds upon switching.

tldr; There are an impossible amount of factors involved. The most important one would be their output potential in terms of damage which Paladin is undeniably lacking compared to the other tanks. But in terms of effective active mitigation, paladins have it the easiest as it's a single button action without the use of resources.