Skip to content
Maintankadin

Determining miss rates, or how much over 102.4%?

Lavnihs
Fri Jan 02, 2009 2:55 am

Determining miss rates, or how much over 102.4%?

It is my understanding that the in game tooltips show your actual percentages after DR for everything but miss rate.

Obviously as we approach infinity on defense rating, the curve for miss becomes flatter and will flatten out completely. Thus, for varying amounts of defense a macro will not work to tell you exactly what your DR from miss is at all points, as it is not a linear function.

After restating the obvious above (or displaying a complete lack of understanding) I don't understand why it isn't possible to calculate the exact miss rate at 540 defense. If we know that point, then it is extremely simple to determine what sum of avoidance + block is required to achieve block capped.

I know this can't be done with a macro, and I will admit that I don't know how to model it, but isn't it possible to derive miss at that point on the curve?

I should be able to do the math, but should and can are 2 very different things.

I apologize if I'm just stating the obvious and being rhetorical here.

Miss is first on the attack table, or at least it was in my understanding, thus for the purpose of data collection through logs dodge/parry/block ratings would be irrelevant.

I could gather some logs with an old world boss if that would help. Granted sample size would play into it, but that would help give an idea at that point on the curve.

Using 540 as a baseline we know that above this amount of defense we will always have at least X amount of miss, and anyone trying to reach block capped below 540 defense has their priorities a little out of order. That being said, assuming that boss attacks are one roll event, you should push crits off as well being block capped. (This is being addressed in another thread, so I won't comment further)

Granted this is somewhat trivialized by the fact that not being block capped now isn't nearly the issue that being below 102.4% was in TBC, but it's still useful information.

If anyone would like to help I would be happy to donate my time for data collection, or if this has already been done I apologize for the spam post.

Thanks.
_________________
Lavnihs -BE Prot Pally, <Deceit> Anetheron

1/1 Hogger (pre-nerf)
Digren
Fri Jan 02, 2009 2:40 pm
I sort of see what you mean, but I don't think it can be down without the same sort of in-game testing already used.

To repeat back what you say to make sure I'm clear, the combat table at 540 defense consists of:
Miss + Dodge + Parry + Block + Hit

The percentages for dodge, parry, and block can be read directly from the character sheet, post DR. The miss and hit percentages are not available.

At this point it doesn't matter if hit is pushed off the combat table or not.

A) If hit is pushed off, then in-game logs will show all incoming attacks as misses, dodges, parries, or blocks. The percentages for dodge, parry, and block won't exactly match the reported values until the size of the log file is infinite. Until then, the miss percentage reported is an estimate.

B) If hit is not pushed off, then in-game logs will show all incoming attacks as misses, dodges, parries, blocks, or hits. The percentages for dodge, parry, and block won't exactly match the reported values until the size of the log file is infinite. Until then, both the miss and hit percentages reported are estimates.

Do you mean to acquire this number based on parsing of in-game logs? If so, then this has already been done to a large extent by Harlekin (see his DR spreadsheet thread in this forum). If not, then please explain further, because nothing you say makes sense as a method to acquire this number without log parsing. (Indeed, I don't think it's possible without log parsing given the data we're provided.)
_________________
Paladin of the Argent Dawn since 2004
Digren
Fri Jan 02, 2009 2:51 pm
Ok, well I thought about your post longer and I maybe understand what you mean.

Do you mean, when you are block capped, that the in-game percentage for block rate caps?

If it did, then 100 - dodge% - parry% - block% would equal miss%.

Alas, while I don't recall off hand with certainty, I am relatively certain that the block rate percentage shown doesn't cap. The game will quite happily display a block number that, when added to the dodge and parry percentages, is well over 100, if avoidance+block is pushed that high. Thus it's not possible to calculate miss using this method.
_________________
Paladin of the Argent Dawn since 2004
Lavnihs
Fri Jan 02, 2009 6:38 pm
i apologize if I was unclear, I'll try to explain.

We know we see block/parry/dodge post DR on the tooltip. We know that miss isn't correct as it is pre-DR. Thusly, you don't really know when you're block capped, you just take a shot in the dark and say "well I haven't seen a normal hit in a while, so I must be block capped."

I'm aware that if you are block capped, you will no longer have a normal hit, and if you aren't then you will take normal hits.

What I want to find out is a number for miss rate at 540 defense. If I KNOW what miss rate is at 540, then I know that I need a combined dodge/parry/block + Holy Shield of X% to be block capped.

I will look into Harlekin's spreadsheet and see what I can find, if not I'll just go run a bunch of logs and see what happens.
_________________
Lavnihs -BE Prot Pally, <Deceit> Anetheron

1/1 Hogger (pre-nerf)
Digren
Sat Jan 03, 2009 11:55 pm
If you test for 10,00,000 incoming attacks and see zero unblocked hits, then you could say with some amount of certainty that you are at or above the block cap. You'd want to retest with a slightly smaller amount of either parry or dodge rating to make sure that you then did see a few unblocked hits, to confirm that you'd been exactly at the cap. (If you were over the cap, then the math to calculate the miss percentage would be inaccurate and you'd derive the wrong number.)

I think it's actually easier to test somewhere below the block cap. If you have 540 defense but something that you know is less than the block cap, and, after 10,000,000 incoming attacks, you find that the miss percentage is, say, 10.23%, then you can just go ahead and add that point to the table as the miss percentage. There's no worry that you overshot and skewed the data, and you can gauge how accurate your pool of data is by how closely the other numbers (dodge, block, parry) match the percentages from the character pane.
_________________
Paladin of the Argent Dawn since 2004
Arizara
Mon Jan 05, 2009 12:48 am
but blizzard uses funny numbers and fuzzy math... look at it this way, if you're at 102.4% avoidance without miss DR, you're miles ahead of the other tanking classes anyway. we ARE the physical damage tanks regardless of us being at 101.9% or higher. holy shield shapes our niche, if that .002% chance of not blocking is a wipe, then you had other things against you already...

though i would be interested in seeing my actual miss % too, i've kind of just figured it wasn't a big enough number to matter.
_________________
Pancakes
Tue Jan 13, 2009 8:01 am
may be posted elsewhere, way too lazy to read everything at 2am.

afaik from stealing other peoples work and my own petty research the miss cap for the DR equation is ~13 (its actually higher than that but its nice to be safe)

DR equation = 1/x` = 1/c + k/x
(stolen from http://elitistjerks.com/f31/t29453-combat_ratings_level_80_a/)

where
x`is the diminished stat
c is the stats cap
k is a constant depending on your class (pallys is 0.956)
x is the stat pre DR

Basically this means 1 / ((1/13)+(0.956/(140*0.04)) = 4.038

so at 540 def you will get 4%ish, + a mobs 5% chance to miss you gives you 9% as opposed to the expected 10.6% without DR

every 5 def ontop of 540 will give you ~ 0.09% extra miss
_________________
Pancakes - Unholy - Thaurissan - US
majorwoo
Wed Jan 14, 2009 2:29 pm
This is a very good question. Assuming we are using the 102.4 then, how confident are we of ~9%?

I am not capable, (well, not willing) to do the math to verify your numbers so I shall instead watch this thread Smile
_________________

Despite all my rage, I am still just a Paladin Tank.
Pancakes
Wed Jan 14, 2009 2:54 pm
well the miss cap for the equation seems to be higher than 13 so im pretty well convinced that at 540 def miss will be at least 9%, probably closer to 9.1% but i prefer to be safe (not that its a big difference anyway).
_________________
Pancakes - Unholy - Thaurissan - US