[7.2 Guide] Protection - Back with Vengeance

[Lysozyme]

@Emancptr

Yes, haste to 30% is too much on a lot of bosses. I've been thinking about it and I'm going to change it to "haste to 100% effective uptime" rather than "30%" in the summarised stat priority.

However the vast majority of incoming damage is blockable damage; the stuff on logs is after mitigation; after armour and blocks etc.

I do still think that gearing to 30% is a pretty good idea in general if not only for DPS reasons but also for bosses where you are tanking the whole time etc.

Magic and bleed damage is almost never relevant. On Ursoc, the bleed's not going to kill you on its own - it's only an isuse if you take melee damage at the same time (which doesn't happen if you're doing it properly).

@Yellowfive and @Emancptr

Yes, you can get 'good' coverage without much haste. With absolutely perfect play and never having to use shield block at suboptimal times you can get just over 70% uptime with 0 haste (less in practice, of course). But until you have 100% effective uptime, haste is the best. How much haste you need for that varies a lot depending on the encounter. For someone who does a lot of dungeons, and only wants one gear set, I'd absolutely recommend going for 30%. For someone who only raids, or at least only cares about gearing for raids, I'd still probably go for something like 20 to 25% haste.

I do think that it's better to have "not enough haste" on some bosses than "too much haste" on others. But the other benefits to 30% haste - i.e. what if you have to use shield block at a suboptimal rotational time, because you need to use it for survivability purposes immediately? Or what if you are tanking more of the time than you ought to be? And DPS considerations - make it a good limit.

I'm not suggesting that you should forego e.g. a better trinket, or higher item level gear, for haste until 30%. I'm not saying that haste until 30% is loads better than the other stats. Just that until you have 100% effective uptime, which in many cases is 30% haste (or slightly higher, really, but), haste is the best.

- - - Updated - - -
@Yellowfive

Your sim is broken. It's showing vers reducing the damage taken by 8.2%. This would only be true if you had no other damage reductions, but you do (i.e. armour, block, other things). 8.2% DR-from-vers (which 6700 is roughly equal to) should mitigate 8.2% of the damage remaining after all other damage reductions are accounted for (which is how all damage reductions work).

[Deleted]

So which stat priority is correct for tanking mythic dungeons? On here you're saying haste > crit > vers. whilst on another "recommended" thread it's saying vers > mastery

[Swol]

[MENTION=1192695]
Your sim is broken. It's showing vers reducing the damage taken by 8.2%. This would only be true if you had no other damage reductions, but you do (i.e. armour, block, other things). 8.2% DR-from-vers (which 6700 is roughly equal to) should mitigate 8.2% of the damage remaining after all other damage reductions are accounted for (which is how all damage reductions work).

It doesn't say that 8.2% Vers gives an 8.2% damage reduction - it says it gives 8.2% of the NPS, which isn't exactly the same thing.

We do realize that is a bit confusing, and a detailed "white paper" of sorts on NPS is forthcoming. When Yellowfive is awake again, he could give a more detailed explanation, since he did most of the really detailed work on tanking for the simulator and designed the NPS metric.

[Bobby1291]

Don't play prot but thinking about it in the near future, but can someone explain to me the exact reason we use revenge within the last GCD before Shield Slam becomes available, like the math behind it? I know the reason is because it's more of a gain to fish for SS resets but I mean the exact math behind it? I'm sure it's in here somewhere but lol at finding that. If someone can at least point me in the direction of it that'd be great too thanks in advance if you can.

[AanvilGT]

It doesn't say that 8.2% Vers gives an 8.2% damage reduction - it says it gives 8.2% of the NPS, which isn't exactly the same thing.

We do realize that is a bit confusing, and a detailed "white paper" of sorts on NPS is forthcoming. When Yellowfive is awake again, he could give a more detailed explanation, since he did most of the really detailed work on tanking for the simulator and designed the NPS metric.

I pointed this out to Marok back when he did his sims using AMR. Armor and versatility damage reduction is calculated first, and then blocks and crit-blocks (and all of your other active abilities) are taken off second. So if you sim your character with zero versatility and then add 4000 versatility and do the sim again, you will see a 5% reduction in overall damage taken, all of which is "assigned" to versatility, but all of your other mitigation abilities, like block and shield wall and demo shout, will look like they got 5% worse. If you reversed this, then arguably it would just reverse the problem instead of solving it. The detailed breakdown of "assignments" of what each ability is doing for you is interesting, but it doesn't work when comparing stat values.

However, these "assignments" of damage reduction or NPS value, should not affect your topline damage reduction, or NPS total. If you want to use the AMR sims for stat values, you need to ignore the assignments, run your character with several different builds, and look at the topline damage reduction.

[Brewschii]

@Yellowfive, @Swol:

Thanks, your posts are helpful. Can you explain why the simulator shows 4.6m less NPS from armor on a log where Armor is 5996 (56.7m NPS, http://www.askmrrobot.com/wow/simula...7b477207ce9935) compared to one where Armor is 5877 (61.3m, http://www.askmrrobot.com/wow/simula...22079e2ff01c65)?

In theory it could have something to do with Parry or Dodge but, eyeballing it, the differences look small and they mostly cut the wrong way. Is Darkening Soul getting counted before Armor? Is there something else I'm missing? Thanks.

[Sal the Shieldhog]

So which stat priority is correct for tanking mythic dungeons? On here you're saying haste > crit > vers. whilst on another "recommended" thread it's saying vers > mastery

This is the newest info that we've gotten ("we" as in Marok, Macrologia, and myself), so go with this. If it's the post/thread I'm thinking of, it's very old and pre-nerf IP.

Don't play prot but thinking about it in the near future, but can someone explain to me the exact reason we use revenge within the last GCD before Shield Slam becomes available, like the math behind it? I know the reason is because it's more of a gain to fish for SS resets but I mean the exact math behind it? I'm sure it's in here somewhere but lol at finding that. If someone can at least point me in the direction of it that'd be great too thanks in advance if you can.

The reason is it's a DPS increase (very slight, but so). Marok made a post on it in the Protection Legion thread a while ago, and I don't have a link to it. It's a very, very, very marginal increase (we're talking like a few hundred, it's not significant), so don't feel like you have to do it.

[Lysozyme]

It's a survivability increase too; it's a free 5 rage. Pretty minor.

[Bobby1291]

Ok thanks for the replies, i was just asking because of being used to SS Rev Dev rotation for a long time, guess I COULD keep doing it because on my warrior I don't plan on doing anything else besides Lfr lol

[vanquishedValiant]

In that you mean its a small increase over not using revenge at all? And spamming dev > slam is still the #1 priority?

It's been kind of annoying and wearing out my devastate finger, but I can't really deny the feel of results. I do kind of wish that when SSlam had been buffed back up during the IP nerf that they had given it back equally to revenge, so like 15/10, so Revenge wasn't this weird key that lights up all the time but is rarely useful. Just doesn't feel good.

[AanvilGT]

350 mastery is required for 1% AP from mastery.

The main function of mastery, however, is not to increase the size of Ignore Pain; it’s to reduce your damage taken by giving you critical block chance. You need about 233 mastery for 1% chance to critically block instead of blocking.

Could you take a few moments to explain the conversion rates you are using for mastery here?

I think...

the chance to block is 350 points per percentage.
The chance to critblock is 233 points per percentage.
The contribution to AP is 350 points per percentage? This is the part that I can't get to match up with my character sheet on AMR.

[Zortorai90]

Jesus, 30% haste? How are you guys getting that, the gear from EN is worthlessly stuffed with Crit. I'm at 15% Haste, 50% mastery, 6% Versa, and 15% crit (without buffs).

[Alaexaender]

When you block, you take 30% less damage. A critical block on top of that means that you go from taking 70% of the damage to 40% of the damage. This is a damage reduction of 1 – 0.4/0.7, which is roughly 43%.

Accounting for shield block uptime, the formula is therefore:

Total Damage Reduction = 1 – (Uptime*critblockchance*(0.4/0.7) + uptime*0.7*(1-critblockchance) + (1-Uptime)((1-blockchance)*1 + blockchance*critblockchance*(0.4/0.7) + blockchance*(1-critblockchance)*0.7))

Could you explain how you arrive at 0.4/0.7 as the damage taken from a crit block? It should be purely 0.4, as a crit block is a straight 60% damage reduction. If you were to calculate damage reduced in addition to the damage reduction from the normal block, then this makes sense, as you would be taking 0.57% of the damage you take from a normal crit. However, in the equation for total damage reduction, it looks as if you calculate the chance of a crit block, and the chance of a normal block which isn't a crit. Essentially, you removed the reduction of a normal block during the crit block, twice.

Also, I'd like to note that this math doesn't seem to take into account scales of the earth or dragon scales, which are both fairly significant.

[Emtee1986]

Weird, wish i gave this stat prio over a month ago to the sound of every theorycrafter telling me i was wrong because they can open calculator and change the default rotations and stat values in AMR and simcraft.

That's because the simcraft results are completely wrong, they use the wrong rotation and they don't use ignore pain intuitively. Our rotation is based around fishing for a dragon scales proc which the simcraft does not do, nor does it gain rage correctly by prioritising revenge over shield slam.


Haste > Mastery > Vers > Crit

People importing their character to simcraft and smacking the go button need to stop it. It's a tool for theory crafting not for cementing opinion based on a program you can pretty much alter to prove you're right no matter what you say.

Oh wait, i did. Can you guys stop theorycrafting around broken sims at the start of an expansion, never has an expansion been released and had a simcraft tool that's been anywhere near correct, ever. We're almost 2 months in and you're only now getting the correct information because of stubbornness and utter faith in numbers being calculated incorrectly.

[Llarold]

Weird, wish i gave this stat prio over a month ago to the sound of every theorycrafter telling me i was wrong because they can open calculator and change the default rotations and stat values in AMR and simcraft.

I'll largely defer to @Lysozyme with regards to SimCraft, but you're certainly (as I also have) failing to account for its grossly inaccurate damage taken and thus Rage values. Since the damage taken values are inaccurate, that well is basically poisoned until further notice.

I'm also not certain what Dragon Scales has to do with anything. You can't actively force it to proc beyond simply having Shield Block active as much as possible when tanking. Since that is how Warrior has been played since 2004, that's certainly not breaking ground.

[Emtee1986]

I'll largely defer to @Lysozyme with regards to SimCraft, but you're certainly (as I also have) failing to account for its grossly inaccurate damage taken and thus Rage values. Since the damage taken values are inaccurate, that well is basically poisoned until further notice.

I'm also not certain what Dragon Scales has to do with anything. You can't actively force it to proc beyond simply having Shield Block active as much as possible when tanking. Since that is how Warrior has been played since 2004, that's certainly not breaking ground.

Sorry i'm not sure i portrayed the correct level of sarcasm. My responses were all in regards to people pushing numbers on simcraft to inform us of the stat prio early on in the expansion. I'm aware that everything it does is wrong, i'm an advocate of that. Dragonscales i could have removed from the quote, it wasn't relevant - it was in a time where everybody was screaming for versatility which inherently lowered your proc chance of Dragon Scales.

Just to validate once more, i'm in general annoyed that the obvious math for our stat prio was there from day one - you can find my other posts in the legion prot warrior thread if you want, i'm basically quoting myself to be a little bastard in the hope that people will stop simming tank classes and actually start using common sense.

[Llarold]

using common sense.

Well, there's certainly never enough of that on the internet.

[Lysozyme]

Could you explain how you arrive at 0.4/0.7 as the damage taken from a crit block? It should be purely 0.4, as a crit block is a straight 60% damage reduction. If you were to calculate damage reduced in addition to the damage reduction from the normal block, then this makes sense, as you would be taking 0.57% of the damage you take from a normal crit. However, in the equation for total damage reduction, it looks as if you calculate the chance of a crit block, and the chance of a normal block which isn't a crit. Essentially, you removed the reduction of a normal block during the crit block, twice.

Also, I'd like to note that this math doesn't seem to take into account scales of the earth or dragon scales, which are both fairly significant.

I didn't take into account SoE or dragon scales, no. I don't think they'd change very much, though. The value of haste versus mastery/vers changes depending on how high the damage you're taking is. The higher the damage you're taking is, the better shield block is relative to ignore pain; and so the better haste is relative to mastery/vers.

You're right, I made an error in the TDR formula. It could be done with 0.4/0.7s but would need tweaking; easier to change those to 0.4s. I will amend this now.

I expect that this may put mastery slightly ahead, as opposed to slightly behind, haste, for total damage reduction. Even if this is the case, I would still advocate going for haste (up to 100% effective uptime) before mastery, for defensive purposes (and offensive purposes beyond 100% effective uptime, up to 100% total uptime), because it's vastly more consistent.

I have now edited my post to reflect the difference. Mastery is indeed better on average than haste at reducing total damage taken, on average. However the consistency of haste still makes it the superior defensive stat (up to 100% effective uptime).

Could you take a few moments to explain the conversion rates you are using for mastery here?

I think...

the chance to block is 350 points per percentage.
The chance to critblock is 233 points per percentage.
The contribution to AP is 350 points per percentage? This is the part that I can't get to match up with my character sheet on AMR.

The chance to block is not 350 points per percentage, and it is affected heavily by diminishing returns. The formula for block is found in my post (including the hidden passive that increases block chance).

The contribution to AP is 350 points per percentage, with no diminishing returns.

[AanvilGT]

Thank you for that clarification. Looking at Mastery vs. Versatility, I think most of us have found that whether we want it or not, we have far more mastery than versatility. So, if I have 3500 mastery and 0 versatility, I would be getting 18% increased AP from mastery, and would need 1.18*350=413 mastery for my next 1% increase to Ignore Pain or to DPS, whereas for versatility going from zero to 1% would only require 400. At this point, versatility would contribute more to Ignore Pain or to DPS. As you pointed out though, mastery's contribution to block and critblock substantially outweighs (but does not double) the contribution of versatility to preventing physical damage.

So, if you are looking at a specific fight, such as progressing on Xavius.

https://www.warcraftlogs.com/reports...taken&source=4

A majority of the damage you take after mitigation is magical, and that magical damage is occurring at the same time as the physical damage and seems quite relevant to the danger of being killed. If you already have enough haste to block all or almost all of the melee swings, and you have substantially more mastery on your gear than versatility (in excess of 3500). For just this one fight, wouldn't versatility become more valuable than mastery at some point around a difference of 3500?

[Sal the Shieldhog]

Sorry i'm not sure i portrayed the correct level of sarcasm. My responses were all in regards to people pushing numbers on simcraft to inform us of the stat prio early on in the expansion. I'm aware that everything it does is wrong, i'm an advocate of that. Dragonscales i could have removed from the quote, it wasn't relevant - it was in a time where everybody was screaming for versatility which inherently lowered your proc chance of Dragon Scales.

Just to validate once more, i'm in general annoyed that the obvious math for our stat prio was there from day one - you can find my other posts in the legion prot warrior thread if you want, i'm basically quoting myself to be a little bastard in the hope that people will stop simming tank classes and actually start using common sense.

You're forgetting two things: 1. Ignore Pain was hit twice by nerfs, which lowered the value of Versatility immensely, and 2. Versatility does nothing to lower the chance of Dragon Scales proccing. Versatility was so strong for us, and Warriors were so strong in general, that Exorsus used two Prot Warriors stacking Versatility. Haste has been strong in general, yes, but this current stat priority comes after a series of nerfs that hit Ignore Pain multiple times.

Now that Ignore Pain isn't so grossly overpowered, it's led to a change in the stats we want to prioritize.