[Warrior] Critical Block Mechanics

[Kojiyama]

I just made a post with this on the official beta forums and figured I would cross-post it here for those who cannot post there:

Being curious about how critical block currently works on beta compared to live, I sat myself in front of a low-level NPC and did about 30 minutes of testing to get an idea of how things currently function.

Obviously, it's very hard to get definitive answers with a limited set of data, but I have gathered enough data to make some general observations.

The current mastery description states that we have a 10% chance to Critically Block (without Hold the Line) and thus if it were a one-roll system with Critical Block I should see roughly a 10% Critical Block rate.

I performed four tests each against a level 50 melee mob, some Dodge itemization was removed to roughly compensate for the higher miss rate of lower level NPCs:
1) No Hold the Line talent, weapon unequipped to lower avoidance to very low levels, 10 minutes
2) No Hold the Line talent, weapon equipped adding Parries to the test, 10 minutes
3) Same as 2), with 2 points in Hold the Line, 10 minutes
4) Same as 3) with 3 NPCs over a 5 minute period

I used Shield Block on cooldown to attempt to determine if it was one-roll or two-roll by filling up the combat table with Blocks as much as possible. If it was one-roll, we should observe roughly 10% Critical Blocks while if it was two-roll, we should observe roughly 10% of the full Block count as Critical Blocks.

Test 1 - Low Avoidance, No Hold the Line:
Total Attacks: 292
Total Hits: 91 (31.16%)
Total Misses: 39 (13.36%)
Total Parries: 0 (0%)
Total Dodges: 28 (9.59%)
Total Blocks: 134 (45.89%)
Total Normal Blocks: 120 (41.1% / 89.55% of Blocks)
Total Crit Blocks: 14 (4.79% / 10.45% of Blocks)

Test 2 - Higher Avoidance, No Hold the Line:
Total Attacks: 244
Total Hits: 55 (22.54%)
Total Misses: 33 (13.52%)
Total Parries: 30 (12.3%)
Total Dodges: 22 (9.02%)
Total Blocks: 104 (42.62%)
Total Normal Blocks: 92 (37.7% / 88.46% of Blocks)
Total Crit Blocks: 12 (4.92% / 11.54% of Blocks)

Test 3 - Hold the Line Talented:
Total Attacks: 300
Total Hits: 53 (17.67%)
Total Misses: 36 (12%)
Total Parries: 41 (13.67%)
Total Dodges: 23 (7.67%)
Total Blocks: 147 (49%)
Total Normal Blocks: 117 (39% / 79.59% of Blocks)
Total Crit Blocks: 30 (10% / 20.41% of Blocks)
Hold the Line Hits: 161 (53.67% / 55.1% of Blocks)

Test 4 - Hold the Line, 3 NPCs
Total Attacks: 360
Total Hits: 106 (29.44%)
Total Misses: 43 (11.94%)
Total Parries: 59 (16.39%)
Total Dodges: 43 (11.94%)
Total Blocks: 109 (30.28%)
Total Normal Blocks: 91 (25.28% / 83.49% of Blocks)
Total Crit Blocks: 18 (5% / 16.51% of Blocks)
Hold the Line Hits: 349 (96.94% / 98.17% of Blocks)

The numbers seem to indicate that it is still a two-roll system or that the numbers are implemented in a strange fashion. I only reached 10% of the total hits with Hold the Line having over 50% uptime. In both other tests, the total number of Critical Blocks was less than 5%, or around 10-11% of the total number of Blocks.

In the final test, despite the very high Hold the Line uptime, my Critical Block count was extremely low. I am unsure how to explain this other than it being a display about the volatility of a two-roll system when both rolls can fluctuate a fair amount.

Conclusions
Being apparently a two-roll system as on live, we can draw a few conclusions with Critical Block in its current state.

1) Hold the Line offers a fair increase in Critical Block levels, but a low increase in survival. At base Mastery levels, it had a major impact on the rate of Critical Blocks, however it still only offers roughly 1.5% damage reduction for 2 talent points, or 0.75% damage reduction per point. This is low in the new talent system.

2) The Critical Block portion of our Mastery is currently low in value, offering less than 1.5% damage reduction. Oddly enough, the normal Block increase of 10% is much more substantial in benefit, offering twice the value of the Critical Block portion. This seems strange, given that the Mastery is named after the Critical portion.

3) At level 82 it takes roughly 80 Mastery Rating to gain 1.25% Block and Critical Block from Mastery. Quick napkin math puts this at roughly 0.42% damage reduction. However, for the same amount of Dodge Rating I would gain nearly 1% Dodge. This makes the Critical Block Mastery in its current form fairly redundant and inefficient of a stat, only viable at high ranges of Dodge/Parry diminishing returns.

4) In addition to this budget shortfall, in a two-roll system Critical Block may scale inversely with Avoidance. Avoidance pushes Block off the table and as Critical Blocks are a percentage of the total blocks, with Shield Block active it would stand to reason that the number of Critical Blocks would decrease at higher avoidance levels.

5) For Critical Block to become attractive, it will probably require either having much higher numbers than 10% or be preferably converted to a one-roll system with its own entry on the combat table. Shield Block would then ideally increase both your Critical Block and Block chance by 100%.

[Ohhnose]

Did they actually say at any point it was going to be a 1 roll system?
I think that'd work better but havent seen any blue posts about it just people assuming so based on the mastery description (which I dont think you can do).

That aside, responding to some other points:
hold the line isnt just a survival boost talent it also increases threat/dmg so you cant compare it to other survival talents as a %
if it stays 2 roll then the 2 parts of the mastery will scale off each other. By that i mean the more block you get the better crit block chance is and vice versa. That plus diminishing returns on avoidance might explain why at base values it seems underwhelming.

Shield block being garaunteed crit blocks would be too op with it as it is now (60% phys dr for 10sec every 30?) + devalue the mastery bonus even more.

[Kojiyama]

It currently is a two-roll system on live, but this is often not an issue due to the large number used (60%) and the fact that it is a static talent.

I think people assumed it would be converted to one-roll when it became a variable stat-sink, but it currently seems to function more or less as it does on live--just with much smaller numbers.

The fact that Critical Block as a Mastery stat has increasing returns seems very hard to balance, so it would not appear to be a good long-term solution.

Also, the suggestion on Shield Block increasing Critical Block chance by 100% would of course need to be a multiplicative effect--e.g. if you had 10% Critical Block you would have 20% Critical Block with Shield Block active.

[Jimboe]

I believe he means shield block "increase by 100%" as in doubling (2x block% + 2x crit block%) or something similar.

I'm not a fan of double-rng being core to prot warrior mastery. It just feels wrong to me. How would one eyeball it against dodge/parry?

[Kojiyama]

Yeah, it does make it hard to eyeball.

You basically get 1.25% * 30% per 1 Mastery from the Block portion, and 1.25% * 30% * Block% for the Critical Block portion.

It probably only overtakes Dodge/Parry at both high DR levels of each and high enough amounts of Mastery to have a large chunk of base Block%. So, it's a strange mechanic.

[Petninja]

From the sounds of it then, until things change, we're going to want to avoid Mastery on gear since it's basically wasted stats at this point. Dodge and parry sound far superior.

[Airowird]

* In other words, Critical Block is nothing more but a (baseline) 10% extra block amount over a theoreticly infinite amount of time?
* Assuming Mastery works for both the block chance and Critical Block chance, it does work exponentially, meaning it's a bit like ArP now; if you don't have the possibility to get X amount of it, don't bother going for it.
* On Hold The Line: It works better with higher gear levels. The higher your parry chance, the better your overall block and crit chances. Thus, especially against single target, it is a talent a starting tank might want to skip untill end-heroic level. (Atleast if the current situation would go live as it is)

[mistersix]

Do you have a link for your thread on the forums? I'd like to follow the thinking there as well. Thanks.

[Kojiyama]

http://forums.worldofwarcraft.com/th...96528&sid=2000

[jere]

Neat thread Koji. Sounds like the % should be bumped up. Some questions:

1) Hold the Line offers a fair increase in Critical Block levels, but a low increase in survival. At base Mastery levels, it had a major impact on the rate of Critical Blocks, however it still only offers roughly 1.5% damage reduction for 2 talent points, or 0.75% damage reduction per point. This is low in the new talent system.

How are you calculating the 1.5% number?

2) Oddly enough, the normal Block increase of 10% is much more substantial in benefit, offering twice the value of the Critical Block portion.

I don't follow this. If you critical block for 10% of your blocks, then you get (on average):
0.9*block_chance*damage_after_block + 0.1*block_chance*damage_after_block*2 = 1.1*block_chance*damage_after_block.

Given enough hits, this should become a 10% increase, not a 5% increase. Am I missing something?

4) In addition to this budget shortfall, in a two-roll system Critical Block may scale inversely with Avoidance. Avoidance pushes Block off the table and as Critical Blocks are a percentage of the total blocks, with Shield Block active it would stand to reason that the number of Critical Blocks would decrease at higher avoidance levels.

Block shouldn't be pushed off the table unless you get enough avoidance (as you pointed out). I was under the impression that the devs said we wouldn't be reaching those levels in Cataclysm, so it should remain static instead of scaling inversely. Do you anticipate us reaching the 102.4% barrier again in Cata?

[Kojiyama]

Hits get pushed off the table with Shield Block active.

Also, I clarified the 1.5% thing in the thread in a later post. That's actually a best-case scenario. In realistic single-target scenarios, Hold the Line is actually only worth about half that. Blurb quoted below:

Actually, I overstated the value of Hold the Line a bit in that part due to the nature of the first test results.

As for calculating it, presuming it is two-roll as it appears, Hold the Line will convert an additional 10% of your Blocks multiplied by the uptime into Critical Blocks.

Against a single target, the uptime is around 55%. Against multiples, it is around 100%.

If your Block Rate is 45% against a single target and the uptime is 55% of Blocks, you gain 2.5% Critical Block, presuming the combat table is kind to you.

Each percent of Critical Block is only worth 30% damage reduction per point, meaning in that scenario Hold the Line is only giving you 0.74% damage reduction, or 0.37% per point.

The 0.75% approximation per point is presuming a best-case scenario of uptime.

The problem with Hold the Line is that with the innate volatility of a two-roll system, the rate at which you Block will have a massive impact on how many Critical Blocks you get.

For instance, if you notice my lower Block rate against the pack of three, you'll see that the gains from Hold the Line are next to nothing, whereas with my higher Block rate against a single target the gains were not huge, but at least noticable.

(Hold the Line also will wildly fluctuate in value against a single target depending on if you align your Shield Block cooldowns with Hold the Line being active or not. This is less of an issue against multiples given the high uptime of Hold the Line.)

As for the statement about the block increase vs. critical block: given that the bonus from Mastery increases both Block and Critical Block in equal amounts, and given that it is a two-roll system, the Critical Block segment will always be the weaker of the two effects. With the default version being +10% Block, +10% Critical Block, we would have 25% Block and 2.5% Critical Block. In realistic scenarios you will have a much higher Block% due to Shield Block uptime, but even in the 40-50% Block range--which is basically optimal in a single-target scenario--10% Critical Block will still only occur on 5% of your total hits.

Obviously the 10% Block is worth more than 2.5% (or 5%) Critical Block. Due to the effect of it being two-roll, the Critical Block segment is actually always less powerful than the normal Block segment with realistic numbers. (This isn't a 'problem', per se, it's just odd considering the Mastery is called 'Critical Block'.)

[jere]

10% Critical Block will still only occur on 5% of your total hits

Yeah, but 10% critical block is still an average case increase of 10% more block value, regardless of what your block value is. Or are you using 5% to mean more like block chance?

[Kojiyama]

The way you phrased it may be confusing terms. It's a 10% increase to your current Block Value. 10% Block Value is a 10% increase in Block Value%.

10% Block is 3% damage reduction (.1 * .3) whereas 10% Critical Block with optimistically normal values (~45% block rate) is 1.35% damage reduction (.1 * .45 * .3). (Since it's two-roll, you would have always Blocked anyway, so you only gain an additional 30% reduction for the hit.)

If it was a one-roll scenario, the Critical Block portion would be worth twice the normal Block portion (.1 * .3 vs. .1 * .6) rather than having to be multiplied by the percentage of attacks that actually are Blocked.

[jere]

10% Block is 3% damage reduction (.1 * .3) whereas 10% Critical Block with optimistically normal values (~45% block rate) is 1.35% damage reduction (.1 * .45 * .3). (Since it's two-roll, you would have always Blocked anyway, so you only gain an additional 30% reduction for the hit.)

That's why I don't understand then. You are comparing a guaranteed 10% block value (.1*.3) to an average critical block (.1 *.3 *.45)

10% block value isn't a 3% damage reduction. It is also subject to the 45% block chance you listed (normally). If all attacks are always blocked then 10% block value would be 3% damage reduction, but only a portion of the attacks are blocked. Regular block and critical block are subject to the same block chance since critical block is on a two roll system.

It should be (.3 * .1 * .45) VS (.3 * .1 * .45)

[Kojiyama]

I think you're comparing apples to oranges a bit.

The +10% Block is actually changing the combat table, the +10% Critical Block is not changing the combat table, it is simply converting a portion of the Blocks on the combat table to Critical Blocks.

Therefore, the bonus of Critical Blocks will always be a sub-set of Blocks. If you increase both Critical Block% and Block% by the same rate, the Critical Block% increase will always be lower than the Block% increase, as one is dependant on the other to function.

Example:

25% Block = .25 * .3 = 7.5% reduction
35% Block = .35 * .3 = 10.5% reduction
35% Block + 10% Critical Block = .35 * ((.3 * .9) + (.6 * .1)) = 11.55% reduction

As you can see, the additional 10% Critical Block secondary roll portion is only granting an addition 1.05% damage reduction, while the initial primary combat table addition of 10% Block is granting 3% damage reduction.

Being two-roll, the contribution of the Critical Block portion increases (or decreases) with the amount of Block you have, but as it is dependant on your Block% to fill up the combat table entries in the first place will never exceed the equal quantities of Block% in power until the combat table is already completely full.

[Airowird]

I think that what Koji is trying to explain is this:

Dodge, Parry, Block are all 1 roll table. Probabilities(=avg reduction) are thus calculated as (A OR B OR C)
Critical Block is a second roll in a specific case, and has to be considered as a (D AFTER C) probability.
In other words: 10% Block gives you 10% chance to get 30% reduction or 10% * 30% = 3% overall
10% Critical Block gives you 10% chance to get +30% reduction if you block. Chance = 10% * 30% * Block chance.
Therefor Critical Block chance will always offer less reduction than normal Block chance, unless you have no avoidance and capped block out.

But what makes Critical Block the Mastery a bit better is this:
The base 10% improves your 25% Block to 35%, of which 10% are Critical.
Total increase in reduction is 4.05%
Adding enough Mastery for another 10%, you would get 45% Block, of which 20% will be Critical.
.45 * ((.3 * .8) + (.6 * .2)) = 16.20% or a 4.65% increase.
That's a 14.8% increase in efficiency

[Kojiyama]

Yes, as WarTotem says, the returns increase over time. It's very much like Armor Penetration is on live right now for DPS classes.

Early on, Mastery is a lot less efficient than Dodge or Parry rating (almost half as efficient, last I checked, in terms of damage reduction) with initial values. However, depending on scaling, it will likely eventually be better once you can stack a ton of it.

[mistersix]

What sort of break point for our mastery stat would we be lacking at? Too early to tell?

[Kojiyama]

Too hard to tell, because it would also depend on the DR curve on avoidance. When we can get up to level 85 and know the final ratings conversions, I'll try to do some math on it.

[jere]

I think you're comparing apples to oranges a bit.

The +10% Block is actually changing the combat table, the +10% Critical Block is not changing the combat table, it is simply converting a portion of the Blocks on the combat table to Critical Blocks.

Then that's what it was, it looked like you meant 10% block value. From this statement I see you mean 10% block chance. 10% block can mean a couple different things. I just misunderstood what you were talking about.