PDA

View Full Version : Do bosses have a 5% chance to miss base?



Pylae
06-08-2009, 08:53 AM
I'm being told they do but in 3+ years of reading forums I've never seen any reference to this. It'd make a difference to me because the more avoidance I have the more reliable it gets, (DR aside).

According to the armory (and this is against an 80 with no DR, right?):
25.77% dodge
18.41% parry
2% to miss a Night Elf
3% to miss frigid dreadplate (frost DK talent)
6% to miss from defense
=55.18% if bosses do not have a 5% chance to miss, 60.18% if they do

Dylae

Kazeyonoma
06-08-2009, 09:09 AM
you add 5% base miss chance to your avoidance because yes bosses do have an innate 5% miss chance. It'd be easy enough to parse if you could find a simple boss and a few healers and just have him beat on you for an hour.

Fetzie
06-09-2009, 05:36 AM
Isn't it 5% miss if it is a same level target hitting you? I thought everything was 0.2% lower per level difference so a boss would only "base miss" 4.4% of the time.

Satrina
06-09-2009, 07:53 AM
The 5% miss chance is for an equal level opponent trying to hit you. For every level above you, their chance to miss is decreased by 0.2% so yes, a boss actually has a 4.4% chance to miss you. At the same time, your chance to dodge/parry/block is reduced by 0.2% per level the opponent is above you. It's generally easier to say you need 100% + 0.6% miss + 0.6% parry + 0.6% dodge + 0.6% block = 102.4% than to work with reduced numbers directly, so we've been using that convention since the start. People get less confused that way.

Pylae
06-09-2009, 08:35 AM
Ok 60.18% from my character sheet - 1.8% for -.6% miss, dodge, parry (no block) means I'm getting smacked 59.38% of the time?

Kazeyonoma
06-09-2009, 09:26 AM
there we go, yes.

Satorri
06-09-2009, 02:04 PM
Also, small nitpick, Nelfs don't have an additional 2% to be missed, it's a 2% additional chance to dodge.

Miss chance is determined by the difference between the attacker's weapon skill (only determined by level) and the defenders defense skill (players' is improved with defense rating, mobs' is only based on level).

Satrina
06-09-2009, 02:23 PM
The night elf racial used to be dodge, but is now miss.

GravityDK
06-09-2009, 03:42 PM
Ok 60.18% from my character sheet - 1.8% for -.6% miss, dodge, parry (no block) means I'm getting smacked 59.38% of the time?

Right.
Addons like TankPoints or TankTotals do these calculations for you, of course, but I think it's key to understand them first!

And nelfs are miss, too.


According to the armory (and this is against an 80 with no DR, right?):
...
Dylae

The character sheet has DR factored in. The mouse-over tooltips don't.

Pylae
06-10-2009, 08:38 AM
LOL...dodge + parry + miss means I'm NOT getting smacked 59.38% of the time!

Ok so now we can do some more useful math.

My chance to be cracked in the face is 100-59.38% = 40.62%

My chance to be hit twice and three times in a row is 16.5% and 6.7%



If I didn't have frigid dreadplate my chance to be hit would be 43.62%

My chance to be hit twice and three times in a row would be 19% and 8.3%.

(19-16.5) / 19 = frigid dreadplate reduces my chance to be hit twice in a row by 13%...seems pretty good to me! (And 3 times in a row by 20%!)

Satorri
06-10-2009, 03:14 PM
The night elf racial used to be dodge, but is now miss.

Wow, how did I miss that?! I'm usually on top of this stuff!

Satorri
06-10-2009, 03:19 PM
Tee hee Pylae, your use of %'s on %'s makes it a little misleading. =P

Your 3% additional chance to be missed makes you 2.5% less likely to be hit twice in a row. That is a 13% reduction in the % chance.

Though I agree, 1% avoidance per talent point is delightful (and the standard budget, see Anticipation and Deflection) and Frigid Dreadplate is good stuff.

Pylae
06-11-2009, 12:26 PM
I need a job in statistics/lying to management.

Satorri
06-11-2009, 04:48 PM
Word. Or was that sarcasm, have on already?

They use this on the public too. Ever heard about the studies that say X makes you live longer, or die sooner? A lot of those studies are actually even statistically horrible. And many more are just a bad use of statistics. Just because you can ask 10,000 people how many hours of sleep they get per night and whether or not they chew gum, does not mean even if there appears to be a statistical correlation that one actually affects the other.

Martie
06-11-2009, 11:50 PM
My chance to be hit twice and three times in a row is 16.5% and 6.7%



If I didn't have frigid dreadplate my chance to be hit would be 43.62%

My chance to be hit twice and three times in a row would be 19% and 8.3%.

(19-16.5) / 19 = frigid dreadplate reduces my chance to be hit twice in a row by 13%...seems pretty good to me! (And 3 times in a row by 20%!)

These numbers are dangerous to use.
First of all, the chance to get hit twice in a row isn't that important - the chance to get hit immediatly after you got hit is important. Since that first hit already happened, the chance for that second hit is your normal chance to be hit.
Secondly, people think that numbers like 6% are a low chance for something to happen. This is -not- the case. 6% is more often then once every twenty hits. Since boss fights tend to take longer then the 40-50 seconds it takes a boss to do 20 hits, you will see a lot of 3-hit strings, so 6% is a pretty damn big chance. (Even something like 0.1% chance is something you need to take into account, you are getting hit that often during a raid.)
Lastly, I can't think of any real uses those numbers have except to confuse people - and confused people make more mistakes, which leads to more tank deaths.

Martie
06-11-2009, 11:52 PM
Word. Or was that sarcasm, have on already?

They use this on the public too. Ever heard about the studies that say X makes you live longer, or die sooner? A lot of those studies are actually even statistically horrible. And many more are just a bad use of statistics. Just because you can ask 10,000 people how many hours of sleep they get per night and whether or not they chew gum, does not mean even if there appears to be a statistical correlation that one actually affects the other.
54% of all statistics are made up anyway.
Flying Spaghetti Monster - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/Flying_Spaghetti_Monster#Pirates_and_global_warmin g)

jere
06-12-2009, 03:50 AM
These numbers are dangerous to use.
First of all, the chance to get hit twice in a row isn't that important - the chance to get hit immediatly after you got hit is important. Since that first hit already happened, the chance for that second hit is your normal chance to be hit.

Umm, you might want to restate that a bit differently. It doesn't seem to make any sense, and it seems pretty wrong. What it looks like you are stating is that P[A|B] is more important than P[A]P[B], which isn't correct. A is independent of B, so P[A|B] always will equal P[A], which really tells you nothing. Knowing the chance that you will receive 2 hits in a row at any given point, is much much more important than simply knowing the chance to get hit at any given point. At any given point, you don't know if you have been hit or not. Perhaps I am just misunderstanding you, but what you say seems wrong. You might want to expound on it.