Mitigation Comparisons – 4 tanks

[ziggyunderslashone]

Not only do you have two lines saying "warr wo/SB" but I can barely tell the difference in color between the two

Ah, bad typo on my part there, really going for clarity. Indeed it is the higher line thats the active one!

[Garath.Gorefiend]

I also tried increasing numbers of mobs as a graph, but its comparative result depends entirely on where you place their damage...

Ziggy

Simulators show quite clearly what happens when you have the damage at any point and increase the number of mobs

Basically warriors stay flat line at their mitigation point, regardless of the number of mobs, generally, the amount they block (blks/unavoided attack) will be less than one on trivial mobs as critical blocks are meaningless and will slowly grow to whatever their blks/unavoided attack cap is. That can be determined based on the formula I provided in an earlier post. (assuming they are spammin SB when it is up)


Paladin SB mechanics keep them at 1/blk per attack and will stay at 1 as long as the number of incoming attacks can be absorbed by the HS charges. After that, redoubt kicks in but it exponentially degrades as the number of attacks at any given point increases.

Code: Select all

The paladin blks per unavoided attack:

mobs   Redoubt      No Redoubt   Benefit
1         1             1           0
2         1             1           0
3         .99          .97        .02
4         .96          .90        .06
5         .94          .81        .13
6         .91          .76        .15
7         .89          .72        .17
8         .86          .67        .18
9         .84          .65        .19
10        .81          .64        .17
11        .81          .61        .20
12        .80          .60        .20
13        .79          .59        .20


The paladin mechanic seems to level off around .20 benefit for redoubt...will slowly decline and level off as mobs continue to decrease...

This was done with 21% block rating and 53% pure avoidance... didnt bother to see what changing the inputs would do.

Warrior constant can be calculated as:

BR=% Chance to Block
HC= % Chance to be hit

Now, 75% your in normal block mode:
NB= % normal Block Chance = BR/HC*.75*.7
CB= % Critical Block Chance = BR/HC*.75*.3*2 (double normal block)

25% of the time your in shield block mode
SB= % Shield Block Chance = .25*.7 *(1+1/1.3) (normal shield block)
CSB= % Critical Shield Block Chance = .25*.3*(2*(1+1/1.3)

Blk/atk=NB+CB+SB+CSB

Where atk is an unavoided attack.

Naturally if BR/HC >1 then it needs to be treated as 1. Note that this mechanic does not care how many attacks are occurring or what the attack speed of the mob is. It is chargeless.

Redoubt does provide some mitigation, it just isnt as effective as the warrior mechanic.

[Rhî]

What would happen, if Holy Shield would have no charges (just its runtime) and Redoubt, instead of 10% chance for 30% additional block (5 charges), would give a 10% chance for critical blocks?

[Garath.Gorefiend]

What would happen, if Holy Shield would have no charges (just its runtime) and Redoubt, instead of 10% chance for 30% additional block (5 charges), would give a 10% chance for critical blocks?

Did that calculation once...
Removing charges off HS means we stay at a 1.0 blks / unavoided attack

Giving us 10% chance to critically block increases the effective blocks per attack by 7.5%.

Giving us 20% chance to critically block increases the effective blocks per attack by 15%.

Etc.

Increasing the effective blocks per attack is the same as an increase in BV of the same percent.

[ziggyunderslashone]

Simulators show quite clearly what happens when you have the damage at any point and increase the number of mobs

Ya, I didn't mean to imply it wasn't modelable or relevant, the issue was the graph favoured paladins at one end of the damage scale (mainly as you say due to SB critical block capping) and warriors at the other. Without supplying a full range of graphs, the relative position of the classes would depend entirely on which picture I chose to present, the worst kind of statistics.

[Deathwing]

Great spreadsheet Ziggy.

Actually, reverse that, now I can't make up my mind which tank class to play.

More seriously, I'm curious why you didn't include block value in paladin's raw survival time? If they block every hit, shouldn't that be their worst case scenario?

Also, if a scalar of 1 is Naxx 25...what scalar would you say is WoTLK's last teir of raiding?

[ziggyunderslashone]

More seriously, I'm curious why you didn't include block value in paladin's raw survival time? If they block every hit, shouldn't that be their worst case scenario?

Aye, really you could at 102.4% avoidance assume that "worst case strings" would be a series of blocks. I suppose thats an advantage of the class. However I was looking more for an "effective health" figure, rather than an actual real life scenario. I suppose incapacitated would actually fufill there.

Also, if a scalar of 1 is Naxx 25...what scalar would you say is WoTLK's last teir of raiding?

Good question. Looking at previous tier levels, you gained on average 14% in vanilla and 10% in tbc per tier. I doubt we'll see that follow the trend to 8% and would expect t9 to be 234ish ilevel. I would expect arthas loot to be on the 1.3-1.4 scale.

[Deathwing]

Forgot couple more things.

Cell F11 is multiplying the old strength for DK's instead of the new.

What's the 1.05 multiplier is each class final armor tally? Is that the new meta gem? If so, I thought that is 1.02? If not, you should add it.

You set the before mitigation hit to 40k. Can anyone comment on how close to most boss encounters this actually is? I've been doing some extra calculations, namely each class' chance to get gibbed(how many hits from 100% to 0%?), knowing how realistic 40k per hit is important.

[ziggyunderslashone]

What's the 1.05 multiplier is each class final armor tally? Is that the new meta gem? If so, I thought that is 1.02? If not, you should add it.

Dang, must have that confused with the block one. Will update tomorrow.

You set the before mitigation hit to 40k. Can anyone comment on how close to most boss encounters this actually is? I've been doing some extra calculations, namely each class' chance to get gibbed(how many hits from 100% to 0%?), knowing how realistic 40k per hit is important.

I picked that as I've been using teron as my my baseline for months. He hits for about my health pool before mitigation, so I followed the pattern.

[steadypal]

point of damage shield for warriors? i've read that it is their version of ret aura, why did they get something so powerful? how does that compare to ret aura dmg wise? and did blizzard even think about warriors having damage shield AND ret aura in a raid setting?

[ziggyunderslashone]

point of damage shield for warriors? i've read that it is their version of ret aura, why did they get something so powerful? how does that compare to ret aura dmg wise? and did blizzard even think about warriors having damage shield AND ret aura in a raid setting?

Bit off topic, but it's more their version of holy shield than ret aura.

Corrected spreadsheet, updated the OP: http://www.savefile.com/files/1827360

[majiben]

point of damage shield for warriors? i've read that it is their version of ret aura, why did they get something so powerful? how does that compare to ret aura dmg wise? and did blizzard even think about warriors having damage shield AND ret aura in a raid setting?

Additionally since when did you need ret aura to effectively aoe tank? It's some minor threat gain that while nice is not ground breaking. Also it is looking like our aoe threat will be 50-100% higher than a warrior's so this would not tip the scales in their favor yet.

[ziggyunderslashone]

Tanking status: Current Max level gear (ilevel ~213)

[Racolus]

Tanking status: Current Max level gear (ilevel ~213)

O.o? Why the hell a bear is better in avoidance?

[majiben]

O.o? Why the hell a bear is better in avoidance?

Most likely becuase it takes more hits to kill them so while it may take 4 hits to kill a paladin/warrior/DK it would take 8 to kill a bear so even if they avoid 2/3rds the attacks we do it will on average take more time to get them down.