maybe the models that we've run can help with this. Or possibly some sort of binomial distribution coupling MH and OH. I'll see what I can come up with for "a formula", but it may not look pretty.

Was wondering, did anyone ever come up with a scale string for Pawn? (enhnacement shaman of course). I have attempted it, and failed miserably...I have no mind for rnumbers (when did they start putting letters in there anyway?).

Kalman pretty much covered it. Just think about it logicially. If you multiply two things together and increase one, how much of a total increase there is will depend on the value of the other.

fyi, the newest version of the sim does much better with flurry, now within 2% of empirical testing results, above or below.

here's the file, rename to testSham.vbs to run it.

http://www.discofiend.com/wow/ShamanCalc/testSham2.txt

I'm curious - has anyone modeled the dps of WF Totem MH and WF Weapon 5 Offhand?

I've gathered that the totem does not incorporate the 3 second cooldown (right?) so would that be even close to what wf5/wf5 will do in 2.1? I mean, since most melee want the totem anyway...

WF totem is inferior to WF-weapon. I doubt the cooldown will change that.

At last, a slow off-hand weapon with great stats and not even unique!
Need Al'ar!

Sentinel

That axe has a colossal damage range, but at least there's something nice and slow out there.

Sentinel

I don't know if it's been modeled, but it has been tested. WFtotem/WF5 comes out roughly equivalent to WF5/rockbiter. Not competitive DPS at all.

If i do an interface, it'll probably be done through a web site. I still have to figure that part out.

what i'm *guessing* is the case, is that you have word-wrap turned on in notepad, which jumbles all the lines of the code. Download a fresh copy, turn off word wrap from the format menu, then try your change after saving. Hopefully that little change will take care of it.

I downloaded Disquette's simulation and set the TimeLimit variable to 3,600,000. This should simulate 1 hour of combat data. I then ran the sim 10 times without changing any of the variables. The DPS average over 10 attempts was 403.7 with individual values of:
398, 405, 406, 413, 402, 401, 407, 400, 400, 405. I truncated all the decimal values.

I'm no math expert but by setting the TimeLimit to such a high number, I would have expected the variance between the results to be a lot tighter. In other words, with no changes in the variables for +hit, crit, AP, weapon, etc, shouldn't all 10 of the results be within +/- a few decimal points of 403?

The APE Project

The APE Project is meant to address the question of "which item should i use?" in a generalized sense, with the objective being maximum DPS, and the input variables being only AP, Hit, and Crit.

Assumptions:
You're fighting a boss level mob:
+ 2.1 rules - you glance 24% of the time for a 24% reduction of damage on white hits
+ you have a 15% dodge/block/parry to deal with because you can't always be behind a mob
+ you're using a planar edge mainhand, and a 59dps 2.6sp offhand
+ 2.1 rules - you're doubling up Windfury5 on each hand, and both invoke the 3 sec shared cooldown
+ mob has 20% damage reduction due to armor
+ you have full enhance talents, but are not adjusting for totems (they're implicit in the AP and crit parameters)

Methodology:
Run the DPS simulator (which does indeed account for white misses - i botched that in earlier versions) and output the results for the following pseudo-code:
For AP from 800 to 1600 (200 AP increments)
- For hitBonus chance from 10% to 22% (6% hitBonus increments)
- - For crit chance from 14% to 30% (4% crit increments)
- - - Run the sim for 2 hours of combat time (7200 sec) and output DPS and a bunch of other data

Results:
Across differing levels of gear, a simple rule of thumb is:
34AP = 2% hit = 1% crit

Some of the actual results can be seen in the screenshot here:


And the output table the script produced (with a little excel manipulation) saved as an .htm file can be seen here:
http://www.discofiend.com/wow/Shaman..._AP_Equivs.htm

This squares pretty nicely with what I can theorize, based on the following facts:

- Once you're at the yellow hit mitigation (5%+), crit is more important than hit because it affects both white and yellow damage, and increases the uptime of both flurry and UR.

- AP is weighted so heavily because it scales great with both hit and crit

- crit would eventually fall off as flurry and crit near their 100% uptime asymptote, but until crit gets to about 30%, that effect isn't seen so clearly

Unresolved Issues:
- This doesn't take into account set bonuses or other weapon procs
- whereas you can see in the columns of the Excel file that scaling doesn't affect numbers *too* much, it would eventually have a downward effect on both hit and crit, whereas AP would scale indefinitely
- I'm sure there are a bunch of other factors

So, until further discussion points out something that I've missed, for Max DPS I'll be using the formula:

34AP = 1% crit = 2% hit

Put another way, for Pawn Type input (I've never used it, so I'm guessing here)

CR = 1.54
HR = 1.02
Agi = 1.36
AP = 1
Str = 2

Quick Check, base on 100 str...
100 str * 2 value = 200 points
200 /1AP = 200 AP - good
200 / 1.54cr = 130 Crit Rating = 5.9% crit. 5.9% * 34 ~ 200 - good
200 / 1.02hr = 196 Hit Rating = 12.4% hit. * 16 ~ 200 - good
200 / 1.36agi = 147 * 88.5% cr/agi = 130 CR = 5.9%crit * 34 ~ 200 - good

Out of curiosity, why is crit rating worth more than agility? Agility scales with Kings and it takes less agility to get 1% crit than it does CR. Also, rating has the same cost as stat points in the item value formula. Thus, wouldn't agility be worth more than CR?

Great work. But are your assumptions correct? I haven't seen any rigorous tests for 2.1 glancing or windfury yet.

Also would be nice to get the relative values of haste and weapon skill too.

Which means 1 crit rating = 1.5 hit rating = 1.5 attack power, roughly. i.e. hit rating is even less effective and attack power is even more effective than my somewhat controversial rankings had them at? Does this mean we can put an end to the "you need to stack tons of hit rating!" mantra, or am I doing something wrong?

alright, a few responses...

If this is the case, then i have my agi/CR backwards. I thought it was 22.1 cr = 1% crit, and 25 agi. I tried finding the thread on this, but didnt, so i went off someone else's statement in the forum. Since this was an afterthought at the end, it doesnt change the body of the post which worked with numbers that I could get definitively. Also, kings doesnt factor into this. If you assume you have kings, the thing to do would be to multiply the value for the pawn string by 1.1

i dont know, regarding 2.1 glancing. Also, unless they change anything, WF is done correctly. Well, actually, i guess if in 2.1 glancing blows cant proc windfury, that'd change the model as well. I'm going from the reports we've had so far, and 24% chance of 24% reduction.

Haste is tricky (chance on proc? permanent enchant? etc). I'm not sure how to model those correctly. Weapon skill shouldn't be too bad once we find out exactly what it does.

That certainly is what it seems like to me, but that's one reason I'm looking for anyone to poke holes in my math or simulation script, because it definitely goes contrary to what we've heard. The thing is, I'm not sure how HITHITHIT started, and maybe there's good math behind it that i'm missing.

BTW, the reason that I was saying AP scales so well, is that I was looking at the 3 variables with respect to how much they cost to an item's budget, according to wowwiki. Wowwiki item budget formulas say that:

1 hit rating = 1 crit rating = 2 AP as far as costs go.

from this test data, we see that for dps increases:

2AP = 1.29CR = 1.96HR.

In other words, assuming a baseline of 2AP giving x benefit, then it costs 1.29 times as much in item budget to see the same result from crit rating, and two times as much in item budget to see the same dps increase from crit.

Hence AP scales great with item budgets for shammy, and my earlier statement.

I guess I'm confused then. I'm basing things off of my in-game character sheet, which shows that for 231 agility I am getting 10.91% crit, then 1% crit = ~21 agility, not 25. This certainly would make agility more valuable than crit rating. Where do we get the 25 agility = 1% crit thing from? 'Cause messing around with items, it does seem that for each 21 agility (approximately, Blizz math seems to be a little screwy some times), I am getting an additional 1% crit. Character sheet numbers, of course. On the other hand, if I remove all of my +agility gear, I have a base of 61 agility, which the character sheet claims gives me 4.11% crit, which comes out to 14.8 agility per 1% crit!

I suppose I shall do some additional calculations, then. Base of 61 agility = 4.11% crit. If I throw on my enhancement gear, I get an additional 173 agility and 6.92% crit, and that definitely does come out to 25. Well, scratch that, then. CR > Agility! Damn character sheet throwing all my number crunching off!

yeah, character sheets are notorious for this ;-) I'm glad to have independent confirmation of it through your calculations though :-)

He was correct, 22.1 critrating = 1% crit = 25 agility. You get some crit just for leveling up, it doesn't start from 0. There was a recent blue post (on the test forum, I think) about this. The main takeaway from these simulations is that crit rating is undervalued for shamans on items, so we should load up on crit to make the most effective use of our item budgets. Of course he's just looking at DPS, agility gives dodge and a bit of armor too, and of course it's affected by BoK. Point for point, I'd rather have agility.

Do you have evidence that WF in 2.1 works like you assume, 20% chance with 3s shared cooldown? I've seen lots of tests that say otherwise, although none of them are definitive. People dualwielding 2.6s with WF5/WF5 tend to show an 18% procrate, which would be impossible under that assumption. This is kind of a big deal. Same thing with glancing, except almost nobody has tested it at all. Is it 24/24, or 25/35, or what? Nobody really knows.

I would assume haste as a permanent enchant so it can be plugged into the spreadsheet. You can make an educated guess as to the effective amount of permanent haste from a PPM or % proc with cooldown.

You have a small base chance to crit without any agility, which is what throws off the calculations. As you found, you can easily observe the amount of crit % granted by agility by unequipping an item with agility on it.

I've been playing around with Lootzor to see the "best" items, given these AP equivalencies. I used the values given above, then arbitrarily chose values of 1 STR = 4 STA = 8 INT = 8 MP5 to provide at least a little bit of balance on the gear, beyond just pure STR/AGI/CRIT/HIT. (I also limited the selection to mail armor, but whether I should have or not is probably debatable.) The results are kind of interesting... Cyclone isn't always as bad as I thought, just because it does do a nice job of stacking STR and has red sockets to throw in nice +8 STR gems.

Thinking about it a bit more, the "best items" change whether or not you have Kings, so if you want to try that out, change STR to 2.2 and AGI to 1.496. The change isn't always dramatic, but it does make a difference.

I'm really hoping that WF in 2.1 acts like dual WF5/WF5 does now - 36% chance on hit after the 3 second rule. With 2.6/2.5 weapons, i got about an 18% "overall" proc rate on live that way. I'm very comfortable at the moment saying that the model works that way:

- time 0.000 Windfury Proc
- time 0.001 to 2.999 no windfury procs possible
- time 3.000 onwards - each hit, regardless of hand, has a 36% chance to proc a windfury

This I've tested time and time again on live, probably over 5K white hits, and it is consistent within a % or two. It also has the side effect of looking like an 18% proc overall with two 2.5+ speed weapons, and 13 or 14% proc on two 2.0- speed weapons.

Yeah, even before doing this, i was thinking how i'd kill for rogue tier 2.5 and 3 gear.

She's got legs, and she knows how to use them:

http://www.wowhead.com/?item=22477
+31 Strength
+32 Agility
+25 Stamina
Durability 90 / 90
Requires Level 60
Equip: Increases your critical strike rating by 14.
Equip: Increases your hit rating by 10.

drop the agility slightly and add in AP for item-budget-compliant awesome shammy legs.