Analysing the benefits of Dodge/Parry and Mastery（Editing）

[kakalu2000]

First of all,we must know how to calculate the Dodge/Parry/Mastery on your character sheet

Parry = 5%+ parry(after DR)
Dodge = 3.9% +dodge(after DR)
Block = 5%+15%(Sentinel)+Mastery*1.5%+0.44%(shield enchant)

DR math : 1/y = 1/c +k/x
x= Dodge/Parry Rating adds Dodge/Parry chance before DR
eg: Dodge rating 1384 of adds 7.84% Dodge chance,so x = 7.84

c=65.63 k=0.956

In order to calculate easily,i ignore the 2 factors
1 Hold the Line in the talent
2 ShieldBlock in the skill

1.jpg


some explanations for this table

Parry = parry on your character sheet after add 176.72 parry rating
Before DR = increase 176.72 parry rating adds parry chance before DR
DR = increase 176.72 parry rating adds parry chance after DR
1%parry = the increase reduction when add 176.72 parry rating(1%parry =1reduction)

eg：the row of Parry = 15.58, 1%parry=0.75%
it's mean that 14.79%→15.58%,add 176.72 parry rating can take the 0.75% reduction


Mastery = mastery on your character sheet after add 179.28% mastery rating

Mitigation = the matery you had gain the reduction,In fact, it's a Mathematical expectation
math = block*(critBlock*0.6+(1-critBlock)*0.3)

1%Mastery =the increase reduction when add 179.28 mastery rating
1%Mastery(actual) = add 176.72 converse 0.9852 mastery' effect in actual
becase 176.72 = 1% Dodge/Parry(before DR) =0.9852 Mastery

it's easy to find that
Parry：13.23%→14.02% 1%Parry =0.79%
Mastery : in the same row 17→18 Mastery 1%mastery(actual)=0.77%

Parry：14.02%→14.79% 1%Parry =0.77%
Mastery : in the same row 18→19 Mastery 1%mastery(actual)=0.78%

so,it's mean when your character sheet
parry=14.79 mastery=19
Regardless add the points from gems or refoged
add the same points, the reduction from parry < mastery

however, when parry<15,mastery<19
you should choose the parry, because parry's reduction >mastery

2.jpg

from this picture ,wo can know the rate of change from this 2 attributes

X-axis is mean adding X% parry before DR

in the picture,the cross-point is close to 11(before DR)，so 5%+after DR(11) =14.59


At the last, for the easy to remember,the conclusion is that

1.if your character sheet(buffed) Dodge/Parry<14/15 and Mastery<19,you should reforged or gems to Dodge/Parry is better than Mastery
It's gain the better reduction that maintaining the blance between Dodge and Parry as soon as possible

2.if your character sheet(buffed) Dodge/Parry>14/15 and Mastery>20 , stack the mastery util your unhittable =102.4%

note :
parry14.59≈ 15 because 15 is more remmbered
dodge have the similar relationship when parry-1.1 ,so dodge =14

[kakalu2000]

my English is pool ,so i'm srroy to let everyone watch my article hardly

[kakalu2000]

e.... i forget the title cann't be edited in this forum

[Airowird]

But I love some anal zing!!

Jokes aside, I don't think you can just cast aside Shield Block. It GREATLY improves Mastery rating.
Also, you don't really explain what I should do with Dodge/Parry > 14-15 & Mastery < 19 or the other way around. Its probably not a case that will happen that often, but just for the people that like to know everything

[kakalu2000]

because Shield Block can be casted whenever parry15/mastery20 or parry 16/mastery19

and we change variable only = 1% ，△Reduction Expectation is too small, but it let me construct the math model become complex

so,for this reason, i decide to ignore the ShieldBlock factor,and same to Hold the line

[Booi]

Am I interpreting this properly?
Each parry step is 176.72 parry rating and each mastery step is 179.28 mastery rating - then you are comparing the values directly across...

So when you are at 10 pre-DR parry and 18 mastery, you are actually comparing:
1767.2 parry rating against 1792.8 mastery rating?

Sure it's only 1.4% difference, but your conclusions are also based on percentage point differences... Surely you should be comparing ratings directly.

[kakalu2000]

But I love some anal zing!!

Jokes aside, I don't think you can just cast aside Shield Block. It GREATLY improves Mastery rating.
Also, you don't really explain what I should do with Dodge/Parry > 14-15 & Mastery < 19 or the other way around. Its probably not a case that will happen that often, but just for the people that like to know everything

for this model: parry =14.02,mastery =18
if you get the 176.72 2nd Attribute's points from reforged

1.you converse to the parry , then your parry increase to 14.79, =your reduction +0.77% = the Boss Damage -0.77%
(+1%parry =-1%damage)

2 you converse to the mastery ,then your mastery increase to 18.9852(176.72 =0.9852mastery) = your reduction +0.79% = the Boss Damage -0.79%

so, it is easy to see, in the situation ,mastery is better then the parry

On the contrary , we find the excel table, when parry14.02/mastery18,1%parry>1%mastery(actual)

So,conclusion is deduced for these

[kakalu2000]

Am I interpreting this properly?
Each parry step is 176.72 parry rating and each mastery step is 179.28 mastery rating - then you are comparing the values directly across...

So when you are at 10 pre-DR parry and 18 mastery, you are actually comparing:
1767.2 parry rating against 1792.8 mastery rating?

Sure it's only 1.4% difference, but your conclusions are also based on percentage point differences... Surely you should be comparing ratings directly.

i edited the size of picture , now ,the picture is normal

now,you look the colunm K 1% Mastery(actual)

1% Mastery(Actual) = 176.72*0.9852

[Martie]

Unless I'm reading this wrong, you are looking solely at the pure damage reduction benefits. Since mastery prevents less damage but over a larger percentage of hits, it also functions to flatten out damage graphs. While this may not be the intent of your analysis, it is worth noting, as it (in my opinion) gives mastery slightly more value.

It's the same reasoning why armor was valued so much higher than avoidance pre 4.0.1 - since it was easier to predict, it was better.

[kakalu2000]

Unless I'm reading this wrong, you are looking solely at the pure damage reduction benefits. Since mastery prevents less damage but over a larger percentage of hits, it also functions to flatten out damage graphs. While this may not be the intent of your analysis, it is worth noting, as it (in my opinion) gives mastery slightly more value.

It's the same reasoning why armor was valued so much higher than avoidance pre 4.0.1 - since it was easier to predict, it was better.

yes,you r right,but we also know that ,in the low parry,the DR effect is also low
parry 14% +176.72 = 0.77%parry =0.9852*mastery
parry 30% +176.72 = 0.33%parry =0.9852*mastery

In defferent time,in the same rating converse rate are different

the conclusion is only dodge/parry = 14/15 , we only need few rating from gems or reforged can be achieved the Dodge/Parry to these Criital points

if 3parry vs 4mastery, i think i could choose 3parry,because parry is avoidance,but block is not

for the smooth out the damage ,
if Boss attack =60K, it have 30% chance direct hit you
+3% vs 6%block

continual direct double hit
3%parry =9% chance,and the 6%block =7%
the former have 2% less than the latter,it is mean latter have relative 2% smooth more than former,but parry is aviodance,block only mitigation

[Bigbad]

I like the graph and the table but to make it realistic you should include shieldblock and Hold the Line.

Hold the Line is only complicated if you want to do it accurately. 1% parry increases the uptime for HtL around 2-3% depending on the swingtimer of the boss and current amount of parry. Average amount of block is between 55-60% including shieldblock so HtL will turn an additional 5,5-6% blocks into crit blocks for an additional 30% damage reduction. 2,5%*6%*30%= 0,045%
I would just add a static 0,035-0,045% into the 1%parry column to account for HtL.

Not sure how you calculate the mitigation from mastery, should be %block*average block. Where average block is 0,3*(1+crit block).
Mitigation from 1 mastery: 1,5% crit block*0,3*old block percentage + 1,5% block*average block if you rewrite this you get:
1,5% crit block*0,3*old block percentage + 1,5% block*(0,3*(1+critblock)
0,45%*old block percentage+0,45%*(1+critblock)
0,45%*old block percentage+0,45%+ 0,45%*critblock
0,45% + 0,45%*(old block percentage + crit block)
In the last formula its pretty easy to add in shieldblock just add 50/3=16,67 for 25% critblock + 25% block divided by 1/3 uptime
0,45% + 0,45%*(old block percentage + crit block + 0,1667)

One last thing that changes the value of mastery a little is the 10% crit block from Hold the Line. When i did some calculations a while back i just added average 5% crit block for a 50% uptime on HtL.

Hope this helps

[Booi]

I am still opposed to averaging in shield block for damage reduction in stat comparisons.

You don't die with shield block up. You are unhittable and crit block is over 50%; It just doesn't happen. The marginal difference between stat priorities with shield block up, isn't worth the cost of an effective comparison while you are at your weakest.

[Bigbad]

Impact of shieldblock on the value of mastery is easy to seperate if you dont want to average it. 0,5*0,45%=0,225% extra damage reduction during shieldblock for 1 point of mastery. You can divide it by 3 and add it up to the total or just keep it separate. its significant enough that it shouldn't be left out.

[Booi]

Impact of shieldblock on the value of mastery is easy to seperate if you dont want to average it. 0,5*0,45%=0,225% extra damage reduction during shieldblock for 1 point of mastery. You can divide it by 3 and add it up to the total or just keep it separate. its significant enough that it shouldn't be left out.

Absolutely, I have no issues with it's separation - just trying to avoid failures we've run into in other threads.

[kakalu2000]

Why i ignore the Hold the Line(HtL) and ShieldBlock(SB)

cause
if your sheet parry10,mastery19, they can be casted
if your sheet parry11,mastery18, they can be casted

now,△parry/mastery = 1, the △benefit=1masteryBenefit(cause SB）-1parryBenefit(cause HtL) is too small

if i import them,my math will become complex,but △result is too small,and you know,more complex more may go to wrong
i wander to get the dim value like 14、15 etc, not 14.897、15.999
because nobody likes to remember 14.xxx%、15.xxx

[Katzazi]

Sure, math will become more complex with them, but they do actually change the outcome. Especially SB. It's up either 1/3 of the time or when needed most. (You have other big CDs as well that can be used at SB downtime if you go low there, so it's not totally wrong to assume, that one of the two above assumptions is true for a prot warrior who knows her class and the encounter.)

SB pushes both, survivability and dmg/threat. I doubt many prot warriors will allow SB to have a much longer downtime than 2/3 of the encouter. If you want to be on the save side, just calculate it's uptime around 1/4 of the time. That would be far off of what it's for most tanks but it's much better than to ignore it completely.

Same for HtL. Well you can chose to not skill it, but if you are interested in best survivability you should actually consider it. Everything else does not yield relevant results. It's just not important if the values are 14 or 10 or 19 IF you don't take HtL and never use SB. Anyone not doing this either does not have any problems with survivability or does not know the class.

If you want to give such easy to remember values - which I appreciate - you should at least make some approach to include them. It does not have to be accurate since you just want to give some easy to remember numbers. But if you want to have those numbers taken serious you have to include such important things like HtL and SB.