Protection Warrior Numerical Analysis: 26530

Battle for Azeroth Items and Classes
1 2 3 10 Next
I wanted to add to the analysis of protection warriors discussion by looking at some of the mathematics of our baseline 110 mitigation, especially in relation to druids as they are another primarily mitigation based tank. Most of the discussions on these forums have been primarily qualitative analysis; disliking the general feel or flow of the class, and/or commenting on the various talent lines. The goal of this post is to present a numerical overview of our mitigation as it stands on the beta (build 26530), because I think from a mitigation standpoint we are significantly weaker than other mitigation based tanks (namely guardian druid).

Assumptions:
  • Baseline 110 warrior and druid on BFA Beta
  • Build 26530

Warrior Stats:
  • Baseline Physical Mitigation: 47.49%
  • Dodge + Parry: 23.13%
  • Block Chance: 29.46%
  • Critical Block Chance: 39.60%
  • Block Fraction: 30%

Effective Warrior Mitigation

If one then calculates the average damage taken (reducing the incoming DPS by the P/D chance and then adding in block and mitigation), the baseline warrior health takes 35.51% of the incoming physical damage. If we add on 100% uptime on Bolster as we already have too much rage, this becomes 31.54%. Thus, with bolster the average mitigation of a protection warrior is 68.46%.

  • No Bolster: 64.49% effective mitigation
  • Bolster: 68.46% effective mitigation

Another calculation worth doing is our damage reduction in our active mitigation (shield block). This gives us 100% chance to block and at 110 baseline a 69.06% chance to critically block. Thus the results are as follows:

  • No Bolster: 80.07% effective mitigation
  • Bolster: 82.30% effective mitigation

If we then amortize the active mitigation calculation by assuming 33% uptime (6s with 18s recharge), the average mitigation of a protection warrior is as follows:

  • No Bolster: 69.69% effective mitigation
  • Bolster: 73.07% effective mitigation

This does not take into consideration the change in any additional cooldowns, this is merely a base mitigation calculation.

Effective Druid Mitigation:

It's worth comparing this against a baseline druid on the beta who can maintain 100% uptime on IF (ironfur). We'll look at 100% uptime 0, 1, and 2 stacks of IF. The baseline druid has 21% dodge.

  • No Ironfur: 52.42% effective mitigation
  • 1 Stack IF: 60.90% effective mitigation
  • 2 Stack IF: 66.43% effective mitigation

Assuming a druid can keep 100% uptime on 2 stacks of IF, the protection warrior with 100% bolster uptime (easily maintained) manages 2.03% higher mitigation without the use of shield block. With consistent shield block usage this gap increases to 6.64% higher mitigation.
Class Comparison

At this point the effective differences between the class may be stated:
  • Druid: +29.57% more health than warrior
  • Druid: Self heal of 0.69%/sec (averaged FR)
  • Warrior: Magic mitigation of 15% for 5s every 25s
  • Differing DR cooldowns

The effectiveness of 29.57% more health cannot be overstated, especially when a difference of 6.64% mitigation is all that separates the two classes. While the heal on FR has varying importance based on damage intake rates, the overall health difference clearly suggests warriors are a suboptimal choice. On the topic of FR, the self-heal could either be incredibly important or relatively unimportant. If we rephrase the damage calculations from before in terms of DPS as a fraction of health we can see what the "effective mitigation" of FR is, since it is strictly additive. Here we average out the healing of FR assuming an infinitely long fight with no overhealing. Simply reduce these values by the fraction of effective healing you expect if you would like to consider overhealing.

Frenzied Regeneration Effective Mitigation (Infinitely Long Fight):
  • DPS of 5% Total HP/S: 13.89% effective mitigation
  • DPS of 10% Total HP/S: 6.94% effective mitigation
  • DPS of 20% Total HP/S: 3.47% effective mitigation
  • DPS of 30% Total HP/S: 2.31% effective mitigation
  • DPS of 40% Total HP/S: 1.74% effective mitigation
  • DPS of 50% Total HP/S: 1.39% effective mitigation

Assuming high damage output (say ~30% total HP/s taken) and an infinitely long fight, the 2 IF stacking druid finds a total effective mitigation of 68.74%, only 4.33% less than the Protection Warrior using shield block and bolster optimally. The point here is that the trade is effectively 4.33% less damage for 29.57% less HP. This point can be modified by taking Indomitable instead of Bolster, and thus the warrior takes 0.95% less damage for 17.79% less HP.

The end result of this analysis is the following, protection warriors offer little additional mitigation as compared to other mitigation based tanks, at the detriment of significantly reduced health pools and lack of burst self-healing. Hopefully Blizzard will take note of this and a self-heal and/or reactive active mitigation tool will be added to protection warriors to bring them in line with other tanks.

Please point out any errors in these calculations, as this is intended to help in the tuning of the protection warrior!
This is very helpful and points out where Warriors are still lacking.
The Answer to this would be a recovery mechanic which would cover our reduced health pool, doesn't even need to be massive, something like Victory Rush coming active at low HP or a version of IP with a Fixed raged cost (20) that was stronger the lower our HP was
Idea - at 100% HP it just gives us a 1% healing buff for 4s, at 20% Health it gives us an absorb for 20% of our health and increases healing by 25%, descaling as we gain HP back.

I think Increased incoming healing combined with an absorb is the way to go with Warriors to balance us out. Just make sure the absorb never reaches IP levels.
Given the nature of prot warriors, they could bring back Second Wind and restructure it as a tanking failsafe... like when the warrior drops below 20% (or death) the warrior gets 20% HP back and another 50% over 6 seconds, 3 min CD. Kind of like cauterize but no DoT.

They could also implement blood craze where a chance on hit is we regen a % of our health back, but given warrior's low HP it doesn't really fit as we'd need snap healing.

Or they could play off the class fantasy of a warrior defying death and fighting harder the closer they are to death with something like increasing haste but reducing dmg taken the lower our health gets.... but again due to our low health that gets dangerous.

Ugh just buff warrior health, I feel like keeping it low just limits or makes coming up with solutions harder.
05/02/2018 01:27 PMPosted by Searik
Ugh just buff warrior health, I feel like keeping it low just limits or makes coming up with solutions harder.


HP cap means absolutely nothing once you're past being killed in a snap shot (mechanic + melee hit)

We have nothing that scales off of our HP, so that isn't an answer to anything.

We need a recovery mechanic, its what we're lacking that other tanks already have, be it an absorb, or heal, we need a recovery mechanic that brings us out of a hole, hell it could just be our critical block % goes to 100% when we're below 30% health.

I don't even want to be self sufficient, I just want to have a button I MYSELF can press to save MYSELF, you know actually play my warrior, and survive, rather than just being a meat punching block that healers throw green numbers at. This is what is lacking, if I'm at 20% health, hitting more mitigation will do NOTHING FOR ME. Warriors need a short CD, "save my butt" like all the other tanks, Last stand isn't an answer.
All of this changes *completely* once you get to 120 vs a 122 or 123 mob. It's hard to draw any kind of conclusion at 110 before any stats start diminishing as you gain levels.

Obviously it's much harder to get that data on your own time since you have to level to 120 in the first place (or get scaled to 120), but it should be easy to obtain during one of the raid tests.
05/02/2018 01:19 PMPosted by Ikarikaze
This is very helpful and points out where Warriors are still lacking.
The Answer to this would be a recovery mechanic which would cover our reduced health pool, doesn't even need to be massive, something like Victory Rush coming active at low HP or a version of IP with a Fixed raged cost (20) that was stronger the lower our HP was
Idea - at 100% HP it just gives us a 1% healing buff for 4s, at 20% Health it gives us an absorb for 20% of our health and increases healing by 25%, descaling as we gain HP back.

I think Increased incoming healing combined with an absorb is the way to go with Warriors to balance us out. Just make sure the absorb never reaches IP levels.
The problem is that while this works fine for low incoming damage, at high levels of content, a serious concern is simply not getting one-shot. On fights like Argus or Kil'jaeden, no amount of self-healing is going to alleviate a mechanic that strictly requires a set amount of effective health for you to not die from felclaws or scythes. In Legion, DKs managed by having gigantic health pools to offset their low mitigation, while monks stagger to increase their effective health for example. Without buffing warriors' mitigation or health, their categorically lower effective health will make them strictly at a large disadvantage for these types of mechanics.
Disclaimer: this is sort of analysis is necessarily rough--beyond these basic estimates, talents and many other things will affect mitigation and overall tank performance in any given context.

Disclaimer 2: I didn't review your calculations or assumptions, but the approach seems reasonable.

Even based on your own conclusions, you're greatly downplaying the Warrior's mitigation advantage in your example. "Only 6.64%" is a very large difference when the Druid is only taking 33.57% of incoming damage in the first place. The Warrior (who's taking 26.93% of incoming damage) is taking 19.8% less damage than a Druid. 19.8% is a very large mitigation advantage--it amounts to a 25% increase in effective health, which is very close to the 29% higher health the Druid has in your example.
05/02/2018 04:16 PMPosted by Sigma
Disclaimer: this is sort of analysis is necessarily rough--beyond these basic estimates, talents and many other things will affect mitigation and overall tank performance in any given context.

Disclaimer 2: I didn't review your calculations or assumptions, but the approach seems reasonable.

Even based on your own conclusions, you're greatly downplaying the Warrior's mitigation advantage in your example. "Only 6.64%" is a very large difference when the Druid is only taking 33.57% of incoming damage in the first place. The Warrior (who's taking 26.93% of incoming damage) is taking 19.8% less damage than a Druid. 19.8% is a very large mitigation advantage--it amounts to a 25% increase in effective health, which is very close to the 29% higher health the Druid has in your example.


Right, overall health is not our issue right now, recovery from large hits is where we are lacking.
We're still lacking a 2nd AM, no other tank has to talent into a rotational mitigation.
Our talents as I've said, rather, I HOPE, are largely unfinished works in progress, right now, we might be close EHP, but we need something to save us at low health outside of a single 3minute cooldown, Every other tank has this self recovery option on a short cooldown.
Lets also consider, our low health pool is still a hinderance for burst magic damage. With spell reflect heavily nerfed, a bear is still much more likely to survive a big bad magic hit over a warrior.

Also, our block is also still tied to MELEE hits not physical, so there are many other types of big attacks that it doesn't affect. You can't use block/crit block for bleeds or big unblockable hits. Bolster does help with this at least and is something that iron fur has helped bears with for a long time. That's something.

But TL/DR, can't forget that the warriors strength is also the warriors weakness. We're still the most niche tank that often feels like a one trick pony, but we quickly fall to bottom when mechanics our block is worthless for come into play.

It's really what's made bears so strong for so long, even post ursol removal. At end of day they are still the most versatile "good for any situation" tank.
05/02/2018 04:16 PMPosted by Sigma
Disclaimer: this is sort of analysis is necessarily rough--beyond these basic estimates, talents and many other things will affect mitigation and overall tank performance in any given context.

Disclaimer 2: I didn't review your calculations or assumptions, but the approach seems reasonable.

Even based on your own conclusions, you're greatly downplaying the Warrior's mitigation advantage in your example. "Only 6.64%" is a very large difference when the Druid is only taking 33.57% of incoming damage in the first place. The Warrior (who's taking 26.93% of incoming damage) is taking 19.8% less damage than a Druid. 19.8% is a very large mitigation advantage--it amounts to a 25% increase in effective health, which is very close to the 29% higher health the Druid has in your example.


4.43% would be a better value (see the summary section). That leads to a 16.08% damage mitigation lead by the warrior, but again with a 30% smaller health pool and no effective burst healing/recovery mechanic. If we take the Indomitable example now we're looking at warriors only having a 3.3% damage mitigation lead (relative again) with a 18% smaller health pool! Either way you cut it, warriors are effectively behind by ~10% in EHP.
One thing I would like to point out is that the bear will not be able to keep 100% uptime on 2 IFs while tanking(barring the explicit idea the fight works out to let them start tanking with a full rage bar or their rage generation has increased this build compared to the last time I played them last build). I think the 1 stack IF is a more realistic number of damage reduction. Using the last numbers you had for with-bolster mitigation minus the one stack if mitigation, we're talking a 12.17% physical overall mitigation advantage in almost all instances.

Of course these numbers are conjecture since we'd like to round the IF stack to 1.5 and the value of shield block goes to 0 when it is not up so realistic damage intake comparisons will really depend on many factors that are encounter specific. EHP considerations are always dependent on the factor of the rest of balance. Are healers swimming in mana? What is damage the DTPS number and how many instances over time does it occur?

I would really enjoy a more quantitative analysis of tank mitigation/EHP but it is a very slippery situation with many variables and plugging in past encounters is not something I like to do. Hopefully we can get our hands on some logs from this week and really dig in and see how the damage intake model looks. From that I would definitely love to paint a picture of quantitative strength and weakness between tanks(with regards to potential optimal things going on within that specific encounter).
Protection Warrior destined to be rock bottom of all the tanks again.
Honestly, I have to agree with Ultrasmooth. There's factors that make it hard to maintain 2 stacks of IB,but there's also fundamental failings to Shield Block. On things that it can work on,its great. Phenomenal, even. But, every other tank can reduce non-blockable damage with their AM. If a fight's major danger (will key in on later) is in magic damage, or non-blockable physical, warrior will not only be less effective at mitigating it than every other tank, but combined with our low health pool, we're a downright liability. An active hindrance, compared to another tank.

Now, I want to clarify what I mean by danger. In a fight, Auto-Attack will very possibly be the damage that hits you for the most damage over the course of a fight. That doesn't make it the most dangerous thing, though. Auto-attacks are consistent and reliable, which drastically reduced their danger, but helps them to pull that high overall number. Spikes tend to be consistent and/or reliable too, but their very nature makes them the danger. Its not Argus' auto-attacking that makes him dangerous, its the nukes of his scythe. Its not Hasabel's punching that scares you, it's Reality Tear. Imonar's melee are largely a joke, but bombs (if using a soak strat), the bleeds, and the slight nuke of static hurt (that one barely).

Shield Block alone is insufficient of an AM for these reasons. There's too many tank dangers that it can't help with. Bolster can help with some. However, every other tank has two main AM effects. One that reduces damage (Marrowrend, Ironfur, Ironskin Brew, Shield of the Righteous, Demon Spikes), and one that heals it (Death Strike, Frenzied Regeneration, Purifying Brew, Hand of the Protector, Shear). Warrior has a way to reduce damage baseline (sometimes), but no heal.

Personally, I would suggest a change to Impending Victory, to make it our second form (with some rebalancing), and replace its talent. It would fit with the warrior 'fantasy', in having to force your way past the pain, fighting into a second wind. Plus, its been a core of warrior for so long (Wrath I wanna say, my memory isn't always the best), that it would be a perfect fit for what we've always stood for.
I was about to say, at 114 those mitigation numbers are off by over 20% for me (I have ~68% mitigation w/ shield block + bolster up, item level 222).

Edit: To the post immediately above me, Purifying brew does not heal you. That is incorrect information. Brewmaster, without its artifact, currently suffers from also not having a recovery mechanic when it dips low. Additionally, using purify more often than not on beta requires you to drop your ISB (active mitigation), which Brewmaster is balanced around having 100% uptime on.
05/02/2018 07:39 PMPosted by Mmaadder
Protection Warrior destined to be rock bottom of all the tanks again.


Havent people figured out we are still being punished for being the only tanks in vanilla, and 90% of tanks in TBC?
05/02/2018 04:16 PMPosted by Sigma
The Warrior (who's taking 26.93% of incoming damage) is taking 19.8% less damage than a Druid. 19.8% is a very large mitigation advantage--it amounts to a 25% increase in effective health, which is very close to the 29% higher health the Druid has in your example.


This is only for blockable physical damage. for magic, its even worse. To echo what everyone else has said, bolster should be baseline, and still only helps on physical. with IP gone, and spell reflection nerfed, we basically have no way to mitigate magic damage.

We need an active mitigation that reduces magical damage taken, and as pointed out above we are the only tanks without a self heal.

Without a self heal we have no way to save ourselves after a big spike or if healers are overwhelmed.

Possible solutions: 1) clone bolster but have it affect magical damage instead of physical. this enables us to keep both rolling at the same time (maybe make the CD longer on them to make it so we cannot have 100% uptime but close, similar to ironfur) and use SB for adds and boss auto attack 2) Make spell reflect stay active for the total 5s, reduce the CD to 15-20sec and adjust the % of magical damage reduced to be inline with other classes. 3) Bring back enrage regen on a a 20-25sec CD 10% instant heal with another 10% over 5 sec. which then lets you talent for impending victory if you REALLY want more heals.

No matter what, we need a heal and a magic AM. wthout the we are crippled.
05/03/2018 12:08 AMPosted by Barlar
We need an active mitigation that reduces magical damage taken


Since spell reflect is only 15% they could make it a 10s CD.

Also, thus far, nearly all heavy melee damage is unblockable, meaning we're about half the EHP as Druids currently.
05/03/2018 04:21 AMPosted by Ikarikaze
05/03/2018 12:08 AMPosted by Barlar
We need an active mitigation that reduces magical damage taken


Since spell reflect is only 15% they could make it a 10s CD.

Also, thus far, nearly all heavy melee damage is unblockable, meaning we're about half the EHP as Druids currently.


15% for 5 sec, yea deff has to be closer to 10 sec cd, or a higher %
Only slightly lower than druids in physical blockable damage, when we have no or nearly no self healing, when in all other situations we are massively lower than druids, IS NOT A BALANCED SCENARIO.

Join the Conversation

Return to Forum