There WILL be a TL;DR version at the bottom, of course.

As many people are no doubt aware, Patch 4.2 will bring with it a change to Holy Shield for Protection Paladins.

Holy Shield has been redesigned. This talent is now an activated ability off the global cooldown. It grants 20% increased block amount to a paladin's shield blocks for 10 seconds, with a 30-second cooldown.

Simple math indicates that this is an overall nerf.

Instead of 40% damage reduction with a 100% uptime, you have 50% damage reduction with a maximum 33.33% uptime.

Look at it like this. Imagine you have a 100% block chance, and are facing an enemy that will deal 1000 damage every second to you.

Under current conventions, the damage stream will be flat - 600 damage per hit.

With the change, however, this alters the flow. For 10 seconds every 30 seconds, the damage will be 500 damage per hit. At all other times, it will be 700 damage per hit.

Over short-term, this is a net reduction. However, at the 20 second mark, it is equivalent. Beyond that, it decreases, and at the 30 second mark - where the cooldown is available again - it is markedly worse.

Over 30 seconds, a 4.1 Prot Paladin would receive 18000 damage. Over 30 seconds, a 4.2 Prot Paladin would receive 19000 damage.

Okay, so it's a nerf. And nerfs happen all the time. It's a fact of life in this game.

Well, the question then becomes "why?"

Paladins and Warriors are two of the three shield-using classes in the game. The other class, Shaman, isn't a tank-capable class (or at least, isn't designed with that goal in mind), and so the natural inclination is to draw a parallel between Warriors and Paladins.

Death Knights and Druids use other abilities to simulate the damage reduction granted by shields, in order to balance the classes. These abilities are difficult to draw parallels to, and so we'll take Death Knights and Druids off the table, and compare Warriors to Paladins.

Protection Warriors and Protection Paladins both use mastery to increase their shield block chance.

For Protection Paladins, upon learning Mastery, you are granted 18% shield block chance, and an additional 2.25% for each point of mastery you have. (More accurately, you are given 8 mastery to start with, which accounts for the 18% shield block chance, as 2.25x8 = 18)

When Protection Warriors learn Mastery, they are granted 12% shield block chance, 12% critical block chance, and 1.5% chance of each for each point of mastery you have. (Again, they start with 8 mastery, 8 mastery = 12% chance.)

Each shield-using class (including Shamans) is also given a base 5% chance to block.

In addition to this, Protection Warriors have a 'baked in' 15% additional chance to block, from from the 'Sentinel' passive ability.

The long and short of all this, is that at 20 Mastery (base 8 + 12 from gear), both classes will have a 50% chance to block attacks. I believe this is the point to which scaling was intended.

Prior to reaching this, Warriors will have a greater chance to block than a Paladin in identical gear, and afterwards, a Paladin will have a greater chance to block than a Warrior in identical gear.

Since the 50% mark seems to be the 'magic number', we'll work with that. It's at that point where gear benefits both classes to an equal amount. Prior to/subsequent to that point, scaling works a little differently.

When comparing the two classes, it is important to identify and illustrate the differences in the respective mastery effects.

Without the use of cooldowns (or gear, such as the 1% block value meta gem), a Protection Paladin in 4.1 - provided they are keeping Holy Shield up, which is done by using Shield of the Righteous or Inquisition (or, with appropriate talents, Word of Glory) - will block for 40% of incoming physical damage upon a successful block.

This is a static value, and will only vary if the Paladin's rotation changes, at which point, it becomes 30%. For the purposes of this experiment, let us assume that the Protection Paladin's rotation is constant, and the uptime is 100%.

In other words, assuming no other avoidance stats (which doesn't happen in a real setting, but for the purposes of simulation, we will use this), there's a 50% chance that an incoming physical attack will hit for 100% damage, and a 50% chance that an incoming physical attack will hit for 60% damage.

A critical block is essentially a 'double strength' block. The Warrior's chance to critically block - before the benefits of Shield Block or the Hold the Line talent - is equal to the chance to block granted by mastery. Since we're working with the 20 mastery value (12 + 8 base), the chance to critically block is 1.5% per point of mastery, or 30%.

Critical blocks operate on a 'second roll' system, so a Warrior must first successfully block an attack, before the chance to critically block is applied.

Since the Warrior has a 50% chance to block with 20 mastery, and a 30% chance for that to be a critical block, an incoming attack has a 35% chance of being blocked, a 15% chance of being critically blocked, and a 50% chance of going through uncontested. (Again, we're not dealing with other avoidance stats here.)

Looking at the combat table again, for both classes, with 50% block chance, and no other avoidance:

Protection Paladin - 50% chance for 60% damage, 50% chance for 100% damage

Protection Warrior - 15% chance for 40% damage, 35% chance for 70% damage, 50% chance for 100% damage.

Let's put some numbers into this.

Assume you are fighting a boss. This boss fight lasts 3 minutes. The boss swings every 2 seconds, and hits for 1000 damage. This results in 90 swings. For the purposes of easy math, let's extend this fight then, to 3 minutes and 20 seconds - this results in 100 swings.

You have no other avoidance to speak of. (The reason for this, is because avoidance such as miss, dodge and parry* is equal among both classes - one class doesn't treat it differently than the other, and so they effectively cancel each other out.)

* - yes I realize that Parry affects Hold the Line. More on that later.

Let's also remove any consideration from armour, trinkets, meta gems, or personal cooldowns, or proc effects that add/alter avoidance.

Let's also create a 'control group', so we can show what a 'naked' target would receive.

The damage would read (approximately) as follows:

Control Group:

100 attacks hitting for 1000 damage each, for 100000 damage

total of 100000 damage

Protection Paladin:

50 attacks hitting for 1000 damage each, for 50000 damage

50 attacks hitting for 600 damage each, for 30000 damage

total of 80000 damage

Protection Warrior:

50 attacks hitting for 1000 damage each, for 50000 damage

35 attacks hitting for 700 damage each, for 24500 damage

15 attacks hitting for 400 damage each for 6000 damage

total of 80500 damage

And thus, you see that - all things being equal - a Protection Paladin (provided they keep Holy Shield up) takes less physical damage in 4.1 than a Protection Warrior.

Of course, it's not as cut and dry as that for Warriors, as there are abilities such as Shield Block, that increase one's chance to block by an additional 25% - and if your chance to avoid an attack/block an attack exceeds 100%, the excess is added to your critical block chance.

There is also the "Hold the Line" talent, that adds 10% critical block chance following a successful parry.

Shield Block is a defensive cooldown, and can be used on command; considering its intricacies, we will remove it from consideration by saying that its benefit is cancelled by the Prot Paladin's Divine Protection - a generous concession, as my personal belief is that Shield Block is superior to Divine Protection.

Hold the Line, on the other hand, increases the Warrior's chance to critical block by 10% for 10 seconds (assuming 2 points in the talent) following a successful parry.

This makes it a difficult ability to number crunch for, since we haven't taken into consideration parry for these numbers.

Let's do that now.

If we add 15% parry to each class, the damage output changes.

Each class takes 15k less damage, as 15 of the 'full strength' attacks are removed.

So a Prot Paladin takes 65000 damage, and a Prot Warrior takes 65500 damage, right?

Wrong.

Out of 100 swings over 200 seconds, 15 of them will be parries. With a proc time of 10 seconds, and no internal cooldown, this is a fairly significant uptime for Hold the Line.

Since we're working with 'best case scenario' already - Prot Paladins being able to keep their rotation 100% perfect, and later on, keeping Holy Shield up - let's assume that Hold the Line won't eat its own procs. So over 200 seconds, 150 seconds of that is with Hold the Line active. That's a 75% uptime. And since Hold the Line increases critical block chance by 10% - well, it's pretty easy. With a 75% uptime, you're basically given a 7.5% critical block chance increase over the entire period.

With a 30% critical block chance with 20 mastery, with Hold the Line granting its 'overall' increase of 7.5% critical block chance, the Prot Warrior has an overlal 37.5% critical block chance. This means that with 50% block chance, 18.75% of all attacks would be 'critical blocks'. The remaining 31.25% would be regular blocks.

35 attacks hitting for 1000 damage each, for 35000 damage

15 attacks parried for 0 damage each, for 0 damage

31.25 attacks hitting for 700 damage each, for 21875 damage

18.75 attacks hitting for 400 damage each for 7500 damage

total of 64375 damage

So, as you can see, even in 4.1, and without the use of cooldowns - the strength of which can vary, and is another topic altogether - a Protection Warrior will take slightly less physical damage than a Protection Paladin.

However, I'm okay with things as they are now. The incoming damage for a Prot Paladin is more predictable than for a Prot Warrior. This is a characteristic of the class, and I'm okay with the difference between the two variables. You sacrifice a little overall 'power' for 'reliability'.

Now let's plug in the numbers from 4.2, for the Prot Paladin.

Since Holy Shield is an activated ability, with a cooldown, it no longer has an 'always up' status. As such, the strength that Paladins had over Warriors - reliability, at the cost of overall reduction - is effectively eliminated.

Over 200 seconds, the most you can use Holy Shield is 6 times. You can start it at 0, or you can start it at 20s, either way you slice it, you can only use it 6 times. Since if you started it at 20s, its final use would fall at 200s, the intelligent way to use it would be to use it as early as possible, but no later than 10s. Beyond 10s, and you will be clipping the final 10s of its duration.

So you use it as efficiently as possible, and get the full 6 uses.

Now, you could fall into a mathematical trap here.

If you get 6 full uses of Holy Shield, that's 60s of uptime, right?

And if the boss swings every 2 seconds, that's 30 swings.

And since we block 50 of 100 swings, and 30 of them are during Holy Shield, 20 of them are not, right?

And using that information, here's how it breaks down:

35 attacks hitting for 1000 damage each, for 35000 damage

15 attacks parried for 0 damage each, for 0 damage

20 attacks hitting for 700 damage each, for 14000 damage

30 attacks hitting for 500 damage each for 15000 damage

total of 64000 damage

But wait! 64000 is less than 64375! This is a buff, right?

Wrong. We're oversimplifying. We've overlooked something important.

We're assuming that each and every attack during the uptime of Holy Shield will be a blocked attack.

If just TWO of those 30 swings during the uptime of Holy Shield are parried, then this would be the result:

35 attacks hitting for 1000 damage each, for 35000 damage

15 attacks parried for 0 damage each, for 0 damage

22 attacks hitting for 700 damage each, for 15400 damage

28 attacks hitting for 500 damage each for 14000 damage

total of 64400 damage

Bear in mind that over a period of time, the amount of blocks during that period is predictable - but whether or not they occur during the Holy Shield buff is not. The first example above assumes that the maximum number of blocks - 30 - occur during the duration of Holy Shield. 30/50 is 60%.

In reality, since Holy Shield has a max 33.3% max uptime, it is fair to say that 33.3% of blocks will occur during its uptime.

Of course, RNG could tilt this number either way - but it would have to be an exceptional incident. Even if 50% of all blocks occur during Holy Shield's duration, this would only yield the SAME performance that Prot Paladins currently enjoy.

(Quick math: 25 attacks hitting for 700 damage for 17500, 25 attacks hitting for 500 damage for 12500, is the same as 50 attacks hitting for 600 damage; 30000.)

So the realistic outcome is this:

35 attacks hitting for 1000 damage each, for 35000 damage

15 attacks parried for 0 damage each, for 0 damage

33 attacks hitting for 700 damage each, for 23100 damage

17 attacks hitting for 500 damage each, for 8500 damage

total of 66600 damage

This is a gain of 1600 damage over what Prot Paladins currently take.

Thus, it is a nerf.

We've already established that this change to Holy Shield is a nerf; we've compared it to its current incarnation, we've compared both it and the current incarnation in action, versus a Protection Warrior - and we've come up with the inescapable conclusion that, even in its present state, Holy Shield doesn't give a Prot Paladin an advantage over a Prot Warrior; why then is it being nerfed?

Well, I can't answer that question. However, what I can do, is propose a solution.

It's easy to complain about a change on these forums, but the real task is to identify a problem, and propose a solution - and back it up, if at all possible.

The change I propose is simple. Instead of the 'new' Holy Shield offering a 20% bonus to damage reduction upon successful block, make it 30%.

This puts it on par with the Warrior's 'critical block'. Instead of being based on random chance, a Paladin's 'critical block' analog would be based on cooldown - just like every other ability in the Prot Paladin's arsenal.

If we look at the math, you can see that this change will yield a more balanced approach:

35 attacks hitting for 1000 damage each, for 35000 damage

15 attacks parried for 0 damage each, for 0 damage

33 attacks hitting for 700 damage each, for 23100

17 attacks hitting for 400 damage each, for 6800 damage

total of 64900 damage

This produces a net reduction of 100 damage from what Prot Paladins currently take.

I daresay that it leaves them roughly equal to where they sit in 4.2, while at the same time, adding a measure of RNG to the class's incoming damage.

(Which, if that wasn't the goal - why do it in a way that basically forces it?)

You can plug any sort of numbers into it - account for dodge, parry, miss, different block chances - and you'll still come up with similar ratios.

So, in conclusion, I think that this change needs a little more work, and a slight buff to prevent this from being a nerf.

TL;DR:

4.2's Holy Shield change is a significant nerf.

If Holy Shield becomes a +30% block value buff instead of a +20% block value buff, it will be only a TINY buff (and can even STILL be a nerf, if RNG wills it), and add a little RNG to the mix.

I think that is what should be done.

What do you think?

06/05/2011 05:21 PMPosted by Atrea4.2's Holy Shield change is a significant nerf.

It is a nerf in TDR, but TDR isn't what kill tanks.

Even using it as a 'when needed' cooldown, instead of as soon as the cooldown is available, the loss of 10% overall physical damage reduction is steep. Giving an additional 20% 'on demand' isn't enough of a recompense. 30% however, would make it roughly equal to the current incarnation - which would make everyone happy, both the "I like it as a 'for when I need it' cooldown" crowd, and the "I want to use it on cooldown" crowd.

Edit: Just realized I posted this as one of my alts. Oh well.

Anyhow, I don't see why they don't just make Paladin and Warrior mastery the same and make Holy Shield = Shield Block (We'd have to get the extra 15% too I'd guess). I know it's homogenization, but we're different in so many ways that I don't think it would matter.

Seriously, it resolves the situation of Paladins capping mastery and it makes shield tank balancing quite a bit simpler. Also, it would free them up to work on fixing DKs.

Edit: Good math, the post was very informative. I like your idea if they are determined to make sure Paladins and Warriors aren't balanced against each other, but as I mentioned, I would prefer them to just give us the same mastery.

It's all a matter of variety.

I'm not against them being identical either, but I think if they wanted that, it would have been that way from the beginning.

Oh, and I downed H Nef with Void. :P

P.S. - Well if you decide to raid on that Paladin again then come on over to Summit and bring any Resto Shaman, Rogues, Mages, or any type of Priest ya know with you!!

We're already behind Warriors in overall damage mitigation from the block mechanic, this change only opens the gap even wider.

A simple additional +10% buff would correct this issue, and leave us at (more or less) the same place we are now, while also giving us a measure of RNG - plus added depth.

(Not that we need it, because I agree that we already have too many cooldowns as it is.)

06/06/2011 12:39 PMPosted by TankeehCouldn't have said it better myself, I was coming to the forum to blast this nerf and propose that they at least extended the duration out to 20 seconds but your post put mine to shame. I hope blizz listens!

If they extend it to 20 seconds, they'd have to reduce the amount blocked, otherwise it would be crazy OP. 50% physical damage reduction 2/3 of the time? Yes please but not going to happen.

If they extend it and reduce the amount blocked, then it loses it's point Just put it back to 10% extra all the time.

Prove me wrong.

PS. nerf like this make me want to play the new KOTOR more than wow....

PSS. im going to click sticky on this and hope the blues will relize things and get the designers to stop this nerf but i doubt it will we all can hope tho lol

I think you are not seeing the forest from the trees. Or to put it in more mathematical terms you are not looking at enough data to draw meaningful results. If i put one foot into liquid nitrogen and the other into magma on average im fine. You need other things like mode, median, and standard deviation as well as r value to know if you got meaningful results.

The majority of the time im getting hit boss dps is below healer hps and healers can spam their low cost heal and nearly regen all the mana they spend before the 2nd cast. A paladin tank taking in slightly more damage during this time is meaningless. But when the boss gets a frenzy and healers wake up i can use HS and now i have 10% more block value than i did pre-4.1.

The question is will HS be up for the entire boss burst time or will it be up for all boss burst. The answer to this will change on an encounter by encounter basis.

____________

セリンダラシャド