Enhance Shaman: The Collected Works of Theorycraft, Vol I - Page 297

panny · 11/01/07, 12:44 AM

Originally Posted by LazyJoe

Well, it looks like your calculations about RED are totally wrong because your crit percentage is calculated like most people do, ie not the way it really works inside an attack table.

I'm cureently having ~30% crit unbuffed, let's say my average crit rating in raid situation is 40% (pretty easy to reach with kings, MotW, the feral aura and double mongoose enchant.... if i drop GoA i have more than 50% when the two mongooose proc, yes i'm a crit lover

)

So my attack table on a boss looks like this :

13,3% miss
5,6% dodge
25% glancing blows
16,1% regular hit
40% crit

To simplify things, if I assume a regular white hit to be 100 damage, we have :

16,1% * 100 = 16,1 dps from hit damage
25% * 0.7 * 100 = 17,5 dps from glancing blows
40% * 2 * 100 = 80 dps from crits

so 113,6 dps, with a very significant part coming from crits

now if i have add the RED effect, crits will do

40% * 2 * 1.03 * 100 = 82,4 dps

This leads to a total of 116 dps with RED.

116/113,4 = 1,021 so the RED is a 2,1 % dps increase in my situation

Sigh, you're right about that, I forgot to account for misses and dodges. The actual equation should be:

FinalDPS = (CritRate * OriginalDPS * 2 * REDmultiplier) + [1 - (CritRate + DodgeRate + MissRate + ParryRate) ] * OriginalDPS

And once you have your OriginalDPS, instead of just multiplying by CritRate, you use the equation:

DPSwithCrits = (CritRate * OriginalDPS * 2) + [1 - (CritRate + DodgeRate + MissRate + ParryRate) ] * OriginalDPS

Then:

FinalDps - DPSwithCrits = RED contribution

The less crit you have, relative to hit and expertise, the less DPS contribution from RED. My earlier calculations show the value of RED if you never get a miss, dodge, or parry, making the values calculated for 1000 physical DPS with 30% crit rate (7 or 13.5 DPS) the minimum value of a RED.

Skiace · 11/01/07, 1:01 AM

Originally Posted by Malan

Skiace - you need to run the sim more than once. Shaman DPS is streaky and prone to silly luck driven RNG inflation/deflation of numbers. Run it 5-6 times. See if any of the AEP are outliers. Discard the outliers, average the ones that are pretty close in value, and use those.

As an example I did exactly this today. (While at work... supposedly while doing "work") I ran the sim 5 times at 1000 hours and got some pretty crazy outliers for a few values. Then I ran it 5 more times at 5000 hours each and got 4 sets that were very close to the 1000 hours numbers, so I took those 4 and averaged them. My very first run of the sim was a huge outlier though, it had some really wonky values.

Ok, bad. Don't do that. Yo added a Q/A that addresses this issue - if you put kings, the sim expects your crit and AP values to already include the extra 10%, the sim has no way to do that for you. Your values may look odd because of this.

I'd also add that you really need to keep in mind that the "baseline" AEP stats were generated from a model that had around 2500 AP, 30% Crit, some value of Hit that I can't remember. Your stats aren't exactly there, so you're going to differ.

As I mentioned in the first post, the "oddball" values came from an average of 5 tests at 10,000 each. The screenshot is just to show what I was inputing.

I know the kings issue is bad, so that could be the issue (though it would surprise me).

As for the baseline stats compared to mine, I have a around 2700ap/32% crit raid buffed; I was under the impression that the sim values are to be unbuffed (except kings).

edit: Running another 5 tests at 4000 without kings. If I get more typical numbers I'll just forget about it.

edit2: Results below

Mengus · 11/01/07, 2:07 AM

Originally Posted by Jerem

I never really looked at Expertise like this.
(I never really looked at it AT ALL, to be honest).

So, if I understand correctly, any one of your swings will result in one of the followings :
- Normal Hit
- Miss (number can be decreased by +hit)
- Dodge (number can be decreased by +expertise)
- Parry (number can be decreased/cancelled by positionning)
- Crit (number can be increased by +crit/+agi)

Right ?

You forgot:

Block
Glancing Blow

And it is in a specific order, namely:

Miss
Dodge
Parry
Glancing Blow (only players and pets versus mobs)
Block
Critical
ordinary hit

This of course is for white damage, as the aforementioned 2 roll system may or may not exist for yellow damage.

Patonus · 11/01/07, 2:24 AM

Ok, I have this weird conflict. Using the weapon AEP scoring system...

Netherbane gives Total OH AEP = 447.05

Yet my newly acquired Rising Tide gives Total OH AEP = 444.14

So...does this mean that I should just clearly continue using Netherbane over Rising Tide?

Mengus · 11/01/07, 2:42 AM

Originally Posted by Myul

6. Farm as much [Badge of Justice] (200 possible) for patch 2.3 and the new loot.

Why just 200? Based on what I see, it's 315 for me for ENH upgrades, not to mention close to the same amount in Resto and Elemental upgrades as well.

Kyrious · 11/01/07, 5:29 AM

200 is the current cap for badges per stack. I assume it's the global cap

daveqstone · 11/01/07, 5:50 AM

Originally Posted by Ghando

Might be worthwhile to dispel some common misconceptions and set in stone a couple things, such as:
--Windfury Weapon suffers neither from the OH 50% penalty for OH procs, nor the dual-wield 24% miss rate. This means that (assuming identical weapons), MH and OH procs are identical.

i DO understand that "you can not crit when you miss" is FALSE (atack table)

but i DO NOT understand why WF does not suffer the miss rate. You can get WF when you miss ?
I doubt that , when i was at 62lvl i start to skill 2h axe (blood furnace drop). I had about 170 2h axe rating so i had a LOT of misses , i didnt get any WF proc at the miss hit.

Illundai · 11/01/07, 6:33 AM

This was answered somewhere on the last page or 2... Sigh.

Anyway, no Windfury can't proc off a miss. But due to the Windfury mechanics and procrate you'll never have Windfury on cooldown for longer than a few seconds. That's why Hit isn't THAT important.

Nisall · 11/01/07, 6:37 AM

He means the following (I think)

I know it won't happen due to talents, but lets say you have less than +9% hit. Since you aren't hit capped for special attacks it is possible the WF swings will miss eventhough the white swing that procced it hit. So he is saying is that WF is a special (yellow) attack and therefore only has a 9% hit cap instead of the 28% dualwield white miss chance.

LazyJoe · 11/01/07, 6:39 AM

First you never miss with stormstrike even with 0 +hit, which is a significant part of your total hits (around 16% with 2.6 weapons).

Next, as it has been said many times, you spend a lot of time inside the WF cooldown, and so most of your miss happen during the cooldown. Increasing your hit rate will have an effect to your windfury procs indeed, but it will be a minimal one, in fact you will mostly increase your number of hits inside the cooldown without increasing your number of windfury procs.

You may even say that increasing your hit decreases your WF proc rate but i personnally dislike this because it tends to confuse people when they see it, you don't reduce your number of windfury procs, but instead you increase your white hits (which leads to a smaller proc rate).

daveqstone · 11/01/07, 6:59 AM

Originally Posted by Illundai

This was answered somewhere on the last page or 2... Sigh.

Anyway, no Windfury can't proc off a miss. But due to the Windfury mechanics and procrate you'll never have Windfury on cooldown for longer than a few seconds. That's why Hit isn't THAT important.

OK , now its more clear.

Hit maybe isnt THAT important , but with just +9% from talents you have base 17% miss rate. 21% agains a boss. At every fifth atack you have no chance WF proc , with low or none haste ratings it is a lot IMO.

What i want to say that in the Theorycraft is written that hit rating changes only 45-50% of our damage (white attacks , +9% hit from talents). I say hit rating also lowers the time between WF procs.
With high haste rating and crit chance this influence is very low , but without haste it has some kind of importance, becose i want my WF as often as possible.

Pitbuller · 11/01/07, 7:01 AM

Originally Posted by Patonus

Ok, I have this weird conflict. Using the weapon AEP scoring system...

Netherbane gives Total OH AEP = 447.05

Yet my newly acquired Rising Tide gives Total OH AEP = 444.14

So...does this mean that I should just clearly continue using Netherbane over Rising Tide?

AEP difference is pretty small only 3. And this whole weapon AEP is pretty new threory. I personally use 8.48 mh Wdps AEP and 4.24 oh Wdps AEP. If you use those Rising tide > Netherbane. But my suggest is put your Rising tide to mainhand then choose Merciless gladiator pummeler or Netherbane to OH. Then just wait/hope that Syphon of the Nathrezim drop.

daveqstone · 11/01/07, 7:10 AM

Originally Posted by LazyJoe

You may even say that increasing your hit decreases your WF proc rate but i personnally dislike this because it tends to confuse people when they see it, you don't reduce your number of windfury procs, but instead you increase your white hits (which leads to a smaller proc rate).

First i was really confused , but thinking on it i know what you mean , but this CAN be right and CAN be wrong. No one can really say that with low hit rating you will miss during WF cooldown and hit during no WF cooldown.

Wait , that would be interessing adon. Can anyone do this ? Adon that will show how much % of attacks you miss while the WF cooldown is ON and how much % while is OFF

Disquette · 11/01/07, 7:38 AM

Originally Posted by daveqstone

OK , now its more clear.

Hit maybe isnt THAT important , but with just +9% from talents you have base 17% miss rate. 21% agains a boss. At every fifth atack you have no chance WF proc , with low or none haste ratings it is a lot IMO.

What i want to say that in the Theorycraft is written that hit rating changes only 45-50% of our damage (white attacks , +9% hit from talents). I say hit rating also lowers the time between WF procs.
With high haste rating and crit chance this influence is very low , but without haste it has some kind of importance, becose i want my WF as often as possible.

Perhaps it would help us understand your point if you wrote a small "combat simulator", maybe in visual basic, maybe in java, your choice really, to mimic combat, taking into account such things. Then you could show us the actually affect of hit rating vis a vis windfury on dps, and we could look at numbers and results instead of just feelings.

LazyJoe · 11/01/07, 7:41 AM

Originally Posted by daveqstone

OK , now its more clear.

Hit maybe isnt THAT important , but with just +9% from talents you have base 17% miss rate. 21% agains a boss. At every fifth atack you have no chance WF proc , with low or none haste ratings it is a lot IMO.

Just to clear things up :

Against a same level opponent you have 6% miss, against a lvl 73 or boss it is 9%.

So when dual wielding it is 25% and 28% respectively (this has been discussed and proved some time ago in another thread).

So with 9% from talents you have 19% miss against a boss.

Anyway if you use Yo's simulator it doesn't speculate on the value of hit, it just simulates a (long) fight, so the benefit of hits in WF procs is already taken into account when the AEP is calculated. And all the results are showing hit is the less valuable stat for us (for example, 1% crit has generally the same AEP value as 1,5% to 2% hit).