This site makes extensive use of JavaScript.
Please enable JavaScript in your browser.
Classic Theme
Thottbot Theme
24 stam vs 8str 12stam?
Post Reply
Return to board index
Post by
Porcell
So with that gear, you would need 56 more dodge rating to gain 1% dodge. 56 Dodge Rating would be 1.42 before DR, 1% after DR, so you are getting ~70% return from your dodge rating and 0.01786% avoidance per dodge rating point.
Now if you added 56 Defense Rating. That's 0.44 to each Miss, Dodge, Parry, and Block before diminishing returns. So before DR it's 1.32 avoidance +0.44 Block. After DR is' ~0.31 miss, 0.31 dodge, and 0.33 parry (+0.44 block). After DR that's 0.95% avoidance, you are getting ~70% return on dodge ratings, 75% returns on your parry ratings, and overall 0.01696% avoidance per point of defense rating. Overall that makes dodge rating ~5% better than defense rating for avoidance for you.
Also note that if you added 56 Parry Rating, that would be 1.13% parry before diminishing returns, 0.84% parry after DR. You get ~74% return from your parry rating, but it's still only 0.015% avoidance per rating point. Dodge rating is still 19% better than parry rating, even with your extreme gearing choices.
Post by
Sakkura
For the defense rating, I say approximately because there's no way to know the ACTUAL miss chance after DR since it is not listed on the character sheet.
...
So it doesn't "end up better" just because you are spreading around diminishing returns.
...
However, if you look at the math, defense doesn't "get better" because you are spreading around avoidance and somehow reducing your diminishing returns. It gets worse once you figure in DR because you are getting parry, which is bad for diminishing returns.
There is a way to know your actual miss chance after DR, and it's been tested. You just use the known block cap along with the known avoidance to find it. Takes a really long time though, not very practical. Instead, you just use the data that has already been calculated for it. Whitetooth has it all explained over at EJ.
As for defense getting worse due to DR.... Check
this
image, courtesy of whitetooth.
Post by
Porcell
There is a way to know your actual miss chance after DR, and it's been tested. You just use the known block cap along with the known avoidance to find it. Takes a really long time though, not very practical. Instead, you just use the data that has already been calculated for it. Whitetooth has it all explained over at EJ.
There's a post on Tankspot that has Expected Miss verses Diminished Miss. At 700 defense rating you'd expect 5.69 miss but you actually get 4.34% miss, which is about 76%. However from the looks of the charts miss is capped WAY lower than parry is (16 is the miss cap, as opposed to 47 for parry and 88 for dodge). Just for comparison, going from 700 dodge rating up to 1000 dodge rating yields a little more than 5% avoidance, but going from 700 defense rating up to 1000 defense rating only gives you a little more than 1% miss.
As for defense getting worse due to DR.... Check
this
image, courtesy of whitetooth.
Not exactly sure what you are trying to show with the TTL graph. It shows Dodge being a good step ahead of Defense, with defense slowly converging and finally surpassing dodge (in so far as TTL is concerned) at the obscene number of 1750 additional defense rating. Sending that graph out that far is kind of pointless when are already talking about having a baseline of 700 defense rating, 300 dodge rating, and 200 parry rating.
Post by
marklartank
Tanks should be geming Stam first then Expertise + Stam (Guardian's Puprle gem for Red) or Defense + Stam (Enduring Green gem for Yellow) to meet meta gem requirements. Most helms have a +12 Stam socket bonus so naturally you should use the Guardian's or Enduring in these slots.
these are actually the only gems i use currently.
Post by
Sakkura
There is a way to know your actual miss chance after DR, and it's been tested. You just use the known block cap along with the known avoidance to find it. Takes a really long time though, not very practical. Instead, you just use the data that has already been calculated for it. Whitetooth has it all explained over at EJ.
There's a post on Tankspot that has Expected Miss verses Diminished Miss. At 700 defense rating you'd expect 5.69 miss but you actually get 4.34% miss, which is about 76%. However from the looks of the charts miss is capped WAY lower than parry is (16 is the miss cap, as opposed to 47 for parry and 88 for dodge). Just for comparison, going from 700 dodge rating up to 1000 dodge rating yields a little more than 5% avoidance, but going from 700 defense rating up to 1000 defense rating only gives you a little more than 1% miss.
As for defense getting worse due to DR.... Check
this
image, courtesy of whitetooth.
Not exactly sure what you are trying to show with the TTL graph. It shows Dodge being a good step ahead of Defense, with defense slowly converging and finally surpassing dodge (in so far as TTL is concerned) at the obscene number of 1750 additional defense rating. Sending that graph out that far is kind of pointless when are already talking about having a baseline of 700 defense rating, 300 dodge rating, and 200 parry rating.
AFAIK that's before block is factored in (because its value cannot be modeled the same universal way that avoidance and EH can). So the graph shows that the avoidance part of defense is about equal to dodge (even surpassing it eventually), with defense the clear overall winner because of ancillary benefits (miss is always active, parry generates a bit of extra threat, block chance is... better than nothing).
Also, look at the stats; that character has a mediocre dodge chance, but a high parry chance. So with a more normal distribution of stats, defense would take less of a hit on parry and thus be even better.
As for the EJ thread about the diminishing returns of miss, that's the very thread I was referencing.
Post by
18907
This post was from a user who has deleted their account.
Post by
Hyperspacerebel
ZOMG I'm in someone's sig! =D
Post Reply
You are not logged in. Please
log in
to post a reply or
register
if you don't already have an account.