DPS Spreadsheets
 

Nothing for 3.2 available.

My Frost Spreadsheet (messy but functional):

The numbers you get should be within 10% of your actual DPS based on the raiding I've done over the past week and, yes, with the 4th rotation (and a 1ppm / 15s proc/duration) Fallen Crusader seriously has a minimum uptime of ~48-49%; it's pretty awesome.

http://spreadsheets.google.com/ccc?k...9K0D_VfLLdkhSg

Although I have stopped posting updates, I do continue to update. My work can be found here:

tzenes Profile, tzenes Details - FileFront.com

Version 5A is planned for a release on the 22nd

I'm not frost so I didnt check that one, but I did check the other two. You both made some amazing work and your effort is def showing and is appriciated.

@dr_AllCOM3: I couldnt figure out how to switch from 2x 1h to 2h.
@tzenes: I think the file is corrupt, or theres something wrong at least, as while opening the file excel says it has issues and asks if I want to try and restore some datal. After that the spreadsheet loads but it seems messed up. Does it suppose to have any sort of "pretty" design (colors, dropdowns etc)?

If memory serves, tzenes only actually supports open office. It's a free download, if you don't mind installing a new program on your computer.

Both of these spreadsheets are very well done, though Dr's is a bit hard to figure out. I'm workin on it though. Would either Dr/Tzenes have any issue with me possibly taking some math for the Rawr module (credit would be given where it's due, of course)? Trying to get Rawr as accurate as possible.

Sure, go right ahead. If you have any questiosn, or if there is anything I can do to help, just let me know

Great Spreadsheet, Trinket question
 

I notice that you have trinket slots, I can see the problem incorporating the ad hoc abilities trinkets can bestow in just a single cell, but was wondering if you had worked around that? Or do you just incorporate the constant values into the normal stats?

I don't currently support 2007 (for reasons I won't go into).

If this is an issue with 2003, please send me a screen shot and let me know what version you are using and I will try to fix it.

The answer to your question is: Yes, it has some "pretty" designs (although not many) and drop downs for items.

The drop downs are broken on some versions and work on others (this has to do with a bug I keep fixing, and then forget to include the fix in the next version).

If you are asking me (and I'm not sure if you are), the answer is: a little from column A, a little from column B. Most trinkets can be easily modeled as a static number (on use mostly)... others, not so much

Ignore this, it was a stupid question.

Interesting question: how closly are the predicted values to the actual DPS output that people are seeing in raids? I'm seeing output about 10-15% lower than the super-optimal scenarios predicted by my sheet and I'm feeling pretty good about that. How is everyone else doing?

Please note where you notice any major failings as this will help to illustrate where the sheets are performing well, and where they need to be adjusted.

I use your spreadsheet allcom, great work!

I get very low value for hit rating in your spreadsheet allcom, but it could ve becasue I had 413 hit rating.

I could not see any date for last update on the download page, are you still developing it? Just want to make sure I use the most up-to-date theorycrafting possible.

I've been trying to make a quick spreadsheet of mine own; very basic, very unpretty and specific to my spec.
The World of Warcraft Armory
Spec is there, and is relevant because...
Everything I read says certain spells have certain coefficients, but when I use the coefficients in my spreadsheets, my math results are way off from my in-game results.

An example, and one easy to test for anyone, is Frost Fever; It has a coefficient of 5.5% of your attack power.
So if I raise my attack power by 1000, I should get 55 more damage a tic.
When I raise my attack power by 670 through my trinket, I should get I get 36-37 more damage a tick; I actually get 68 more damage.

Now, I do have impurity, which is 125% AP coefficients on spells. So assuming that affects frost fever (don't see why it wouldn't), 0.055*1.25=0.06875.
Again, raise my attack power by 670, I should get 46 more damage a tic. I'm getting an extra 22 damage a tic.
Trying on blood plague (since I have fewer points, I assume, that would affect that) and 670 attack power increases my blood plague damage by 53, or 46*1.15 from being in blood stance.

I wanted to check to see if Crit affects our dot damage like it does for warlocks and spriests now, removing 84 crit rating caused no change at all.

So either wowhead is wrong (which I doubt), or I have something else affecting my spells that I don't understand (which is more likely).

Test done on a level 80 test dummy with 0 buffs on it and myself (including cinderglacier).

Any help on the math here?

Black Ice is 30%.
Blood Prescence is 15%.
46*1.15*1.3=68 (and a little extra).
That's why you're 22 points over expected.

In case any one is interested in a way to calculate Killing Machine, you can use this formula which is very close to ingame data:
Critmodifier is the average modifier through crit
KM is the proc chance of Killing Machine (0.0 to 0.5)
Meleecrit is the melee crit chance (0.0 to 1.0)
AbilityDelay is either the time between the current ability benefitting from KM and the last ability that benefitted from KM or an average delay between abilities consuming the Killing Machine debuff (a PS>IT>BS>BS>HB rotation would have an average delay of 10 seconds/2 abilities=5 seconds) (in seconds)
MHSpeed is the average main hand weapon speed (in seconds)
OHSpeed is the average off hand weapon speed (in seconds)
AbilityCrit is the crit chance of the ability without KM (0.0 to 1.0)
Mod is the crit modifier, this would be (1.0 untalented, 1.45 with GoG)

CritModifier=1+(1-((1-KM*Meleecrit)^(AbilityDelay*(1/MHSpeed+1/OHSpeed)))*(1-AbilityCrit))*Mod

I'm posting this because most spreadsheets seem to use estimates for Killing Machine.

Edit:

I've just looked through the spreadsheets here, the only one attempting to model KM correctly is Feorthas' spreadsheet but the very last step looks wrong to me. You simply add the uptime for KM mulitplied by the crit damage difference to the total damage.
So if you write the ability damage into one formula, it will look like this:
Damage = NonCritDamage+CritChance*CritDamageDifference+KMChance*CritDamageDifference
Damage = NonCritDamage+CritDamageDifference*(CritChance+KMChance)

So you imply that the crit chance granted by KM is additional to your usual crit chance which is obviously false, actually it'd need to look like this:

Damage = NonCritDamage+CritChance*CritDamageDifference+(1-CritChance)*KMChance*CritDamageDifference

Feel free to correct me if there's an error in my calculations, I'm half asleep so it's not unlikely at all.

As far as I'm concerned, Hidden, that's still an approximation. And not a very clear one... are you trying to add KM to the crit bonus damage? That would certainly give a number (as most spreadsheets tend to do), but I can't wrap my mind around how that could possibly be the right number. It would only apply to the crit bonus damage on white swings to proc it but it depends on the damage of yellow swings for how much you gain... and using an average delay would definitely produce highly skewed results when comparing rotation segments like HB IT IT DC DC to HB DC IT DC IT DC.

Unless I've made a major error somewhere, my spreadsheet should model Killing Machine as accurately as can be. I still have a few major kinks to work out before it's release ready though...

My model is the common sense/brute force/simulation-like approach: for each white swing, calculate the chance of proccing KM; then combine that with the previously calculated chance that KM had already procced (or 0% to start things off), and reset that to zero whenever a FS/HB/IT occurs. Then when calculating the damage of an FS/HB/IT, instead of its normal critrate use (chance of killing machine proc at this time) + (critrate * (1 - chance of killing machine proc at this time)). Of course to do this you have to have a rotation timemap or whatever you want to call it and generally processing stuff starts taking forever, but that's the price you pay for accuracy.


To clarify why I don't think assigning the KM bonus to crit damage rather than crit chance makes any sense at all:
Increasing the damage of your crits causes each percentage chance to crit to be worth more DPS.
Increasing the chance of your crits causes each point of increased damage to be worth more DPS.
Or viewed another way, you could say that crit chance and crit damage have relative diminishing returns (though it's actually increasing returns on the other). Or stated very simply, 100% crit doesn't help much if you crit for 5, and critting for 5 billion percent doesn't mean much if you never crit at all.

Killing machine in actual practice is an increase in chance to crit. By assigning its bonus as increased critical damage, you skew the scaling factors such that crit% looks more valuable than it is and increased crit damage (whether through GoG or AP or metagem or whatever) looks less valuable.

If you actually meant something else by 'average modifier through crit', I'm sorry, but that's what it sounds like, and since you did include the GoG mod in your calculations it seems like you're doing something involving crit damage which really has no place in KM calculation.