Yes, I get the mathematical purity standpoint. (Though I still think that's a strange ideology. A number of the other stats don't contribute in a purely predictable fashion, either - crit becomes less valuable the more you get (or so I'm told), arp becomes more (I know for sure). Those weights are also guesses, based on the idea you are wearing top end gear.)
So again I must ask for the stat weight for the alternative, average weapon damage.
With that Wowhead becomes useful again - I don't need to compare weapons across ilvls.
Amroo did some quick napkin math for how the average damage difference between Spire and Might affects dps (however, SS is 50% now I believe, plus that is multiplied by the 75% as shadow damage with diseases up, so it's probably more than a 100 damage difference per strike for SS), but stat weights are in relation to AP, not dps.
Here's how we both win - determine a stat weight for how average weapon damage contributes to yellow dps.
Okay, I've attached an image of the math required to determine the stat-weighting for weapon damage. I'll attempt to explain what I'm doing as best I can.
First off, some variables:
d is the total damage done by an attack
P is the percent of weapon damage that applies to an attack (i.e. SS has 40% weapon damage)
D is the average damage of your weapon
AP is attack power
X is the bonus damage an attack gets (i.e. SS gets 318 bonus damage)
c is crit chance
m is crit multiplier (generally 1, but can be changed with talents/metagems)
Line (1) is the generalized equation for how much damage any one instant strike will do. You'll notice that I have left off disease multipliers, armor reduction and many other debuffs that affect the total damage done by an attack. As it turns out, these are irrelevant since they are all applied after a strikes "base" damage has been calculated.
Line (2) is an equation for the damage of an attack after we have increased the weapon damage by some value delta_D. The increased damage is here labeled as delta_d.
Line (3) is an equation for the damage of an attack after we have increased our attack power by some value delta_AP. The increased damage is again labeled as delta_d.
Lines (4) and (5) give us the value for delta_d as a function of either delta_D or delta_AP.
Line (6): Here we equate the two equations in order to determine the value of delta_D in terms of delta_AP.
Line (7) gives delta_D as a function of delta_AP. That is to say, we can enter a value for delta_AP and get the equivalent increase in delta_D.
Line (8) is the result of setting delta_D equal to one; we see that you can solve the equation for a value of 4.24 AP.
Now what does this mean? This means that for any given weapon strike, an increase in average weapon damage by 1 is the same thing as increasing your attack power by 4.24. 4.24 is, therefore, the stat weighting for weapon damage. Our story does not end there, though. This value of 4.24 assumes that all of your damage comes from instant strikes, which we know not to be true. We must therefore take the 4.24 and multiply it by the percentage of our total damage that is the result of instant attacks. This percentage will differ from player to player and from fight to fight and is therefore not set in stone; you'll have to determine thee value that is unique to you.
edit:
Originally Posted by Nathanael
Yes, I get the mathematical purity standpoint. (Though I still think that's a strange ideology. A number of the other stats don't contribute in a purely predictable fashion, either - crit becomes less valuable the more you get (or so I'm told), arp becomes more (I know for sure). Those weights are also guesses, based on the idea you are wearing top end gear.)
All stats actually see this set of "diminishing returns", it's just not as apparent with some as it is others. For instance, since we normalize everything to attack power, you don't notice that the value of AP decreases the more you have of it. This change is instead reflected in the increase in the value of other stats like crit or haste. ArPen is unique in that it actually sees increasing returns the more you stack, but that's a result of how Armor reductions are calculated, not stat weights. Furthermore, there is nothing at all unpredictable about these diminishing returns, they can all be easily calculated using similar methods to the one I just demonstrated here.
I think that's not even the end of the tragedy. In fact, we would have to multiply it by the percentage of our total damage that is the result of the non-fixed damage of our instant attacks. Like, when the old base SS is "An unholy strike that deals 40% of weapon damage as Shadow damage plus 259" we would have to take the 259 times its multipliers out, since this stays fixed.
If you had looked through the math that I posted, you would have seen that this value is irrelevant. It is accounted for and cancelled out early on in the calculations. If you have reason to believe otherwise, I would love to see your math.
Line (1) is the generalized equation for how much damage any one instant strike will do. You'll notice that I have left off disease multipliers, armor reduction and many other debuffs that affect the total damage done by an attack. As it turns out, these are irrelevant since they are all applied after a strikes "base" damage has been calculated.
You clearly know your formula, but can you explain why my assumption - that the fact that they are applied after doesn't mean you can ignore them - is wrong. After all, isn't that (1+cm) applied after base damage too?
For example, using the normalized damage numbers from Amroo - Spire does 1770 physical damage, Might does 1832.5, for a difference of 52.5. But then the shadow damage is 1327.5 and 1367.625, respectively for an additional 40.125 difference. Then toss in only EP and the shadow damage becomes 1500.075 and 1545.4625, making that additional difference 45.3875 instead. I assume that other modifiers affect it as well, just as crit does (which double dips).
This makes me think the first line isn't completely accurate and needs to add in all the multipliers. But the rest of that formula would have been beyond my rusty math skills to provide, so thank you for that.
You clearly know your formula, but can you explain why my assumption - that the fact that they are applied after doesn't mean you can ignore them - is wrong. After all, isn't that (1+cm) applied after base damage too?
For example, using the normalized damage numbers from Amroo - Spire does 1770 physical damage, Might does 1832.5, for a difference of 52.5. But then the shadow damage is 1327.5 and 1367.625, respectively for an additional 40.125 difference. Then toss in only EP and the shadow damage becomes 1500.075 and 1545.4625, making that additional difference 45.3875 instead. I assume that other modifiers affect it as well, just as crit does (which double dips).
This makes me think the first line isn't completely accurate and needs to add in all the multipliers. But the rest of that formula would have been beyond my rusty math skills to provide, so thank you for that.
What he did (and what I overlooked at first) is simply ask, how much AP would I need to get the strike damage gain of 1 point average damage. Once that is done, you only need to multiply by the percentage strike damage does of your total damage, since that indirectly accounts for the fact that AP also does other nice things (like increasing spell damage).
As you correctly assumed, the additional shadow portion of SS increases the value of average damage. But this is taken care of by the fact this increases the percentage strike damage does of your total damage. So if you would multiply the 4.24 with 0.3 (since 30% of your damage might be strike based excluding the shadow portion) you now multiply it with 0.4 (since with the shadow portion 40% of your damage is strike based) - this leads directly to an increase of the EP-value of weapon average damage.
Originally Posted by Frozn
You can be sure that I will never post something anymore. Your arrogance and snobism makes me feel sick, enjoy your idiot infractions. Your community just lost one of the best moonkin of the alliance (gearscore).
You clearly know your formula, but can you explain why my assumption - that the fact that they are applied after doesn't mean you can ignore them - is wrong. After all, isn't that (1+cm) applied after base damage too?
For example, using the normalized damage numbers from Amroo - Spire does 1770 physical damage, Might does 1832.5, for a difference of 52.5. But then the shadow damage is 1327.5 and 1367.625, respectively for an additional 40.125 difference. Then toss in only EP and the shadow damage becomes 1500.075 and 1545.4625, making that additional difference 45.3875 instead. I assume that other modifiers affect it as well, just as crit does (which double dips).
This makes me think the first line isn't completely accurate and needs to add in all the multipliers. But the rest of that formula would have been beyond my rusty math skills to provide, so thank you for that.
You're right in assuming that the factor of (1+cm) is applied after base damage; you'll notice that as such, it does not affect the final result of the stat weighting. I included only so people would not think I was "cheating" when I calculated this value.
For the sake of argument, I've attached a modified set of equations. Here, the variable 'K' represents all other modifiers that could affect damage; EP, Armor, random fight specific buffs, anything at all. You'll notice that the final value for the stat weighting is still the same, as this value is canceled out in line (6).
Ah, now I get it. Thanks to the both of you - I haven't done much real math since I switched away from engineering 8 years ago.
Has anybody posted any recent parses from the most recent PTR, to give an idea of how much of our damage percentage is strikes now in an average fight? Obviously, live is completely unusable.
Edit: Ok, I went back about 10 pages and found a parse that, while older, put strike damage at somewhere near 36.4%.
4.24 x .364 = 1.54 as the AEP for 1 average weapon damage.
Now, to put this into Wowhead, I have unfortunately have to translate it into a weight for weapon speed. To be generous, I'll calculate it at the 245 ilvl dps rate of 1 speed = 255.7 average damage (I say generous because speed becomes weighted more the higher the dps of the weapon when you calculate it this way). So 1.54 x 255.7 = 393.78 AEP for 1 weapon speed.
Interestingly enough, this indicates that weapon speed (well, rather, average damage) is a rather large factor. DBB and Mor'Kosh are both ranked higher than Ferocity, and the well-itemized 3.5 Sharpened Obisidian Edged Blade is actually higher than 3.2 Spire and 3.4 Citadel Enforcer (which drops in ICC). DBB's average damage is higher than Ferocity's, after all (920.52 vs. 909.16). Might is the clear winner even though the other two QD rewards have better stat itemization.
Now if I low ball it, and only do the contribution of dps by SS alone in that parse, you get 4.24 x .266 = 1.12784 x 255.7 = 288.39.
With that stat weighting, things stay much the same - only difference is that SOB's higher average damage is no longer enough to put it over Spire or CE.
*Note: I only ever use the spell hit stat weight when comparing individual items, because it's pretty hard not to be melee hit-capped these days, and it looks like the offset ICC items are going to be loaded with hit as well.
I think there should be a stat weight for dps (determined by it's change in white dps and other things related to it like necrosis), average damage (determined by it's change to strike damage and bcb damage and anything else affected), and weapon speed (it's change to BCB damage due to the number of bcb attacks, as well as it's affect on the number of other flat percent procs).
As far as how to use the average damage weighting on wowhead, there is an option for melee minimum and maximum damage. You can just take the weighting for average damage, divide it by two and put that value in the minimum and maximum melee damage spots.
I think there should be a stat weight for dps (determined by it's change in white dps and other things related to it like necrosis), average damage (determined by it's change to strike damage and bcb damage and anything else affected), and weapon speed (it's change to BCB damage due to the number of bcb attacks, as well as it's affect on the number of other flat percent procs).
The stat weighting for weapon DPS is simply 14*(percent of damage that is white), since it takes 14 AP to get the equivalent on 1 white dps. Additionally, one could come up with a stat weighting to reflect how a weapon's speed will affect BCB procs, but the change in overall damage one will see in going from a fast to a slow weapon is so small it can be safely neglected in these calculations.
Nathanael: massacre has the right idea when he says you can just enter in the stat weight for damage into max/min (after you divide by 2). It will produce the exact same result as if there were a weighting to average damage.
Since I am still downloading the PTR and will have to wait till tomorrow night before I can play. I was wanting to ask a question about the 3.3 rotation.
I noticed the rotation goes back to the "old" rotation but it has |ps| first then |it|. I was always of the impression |it| was used first because it could be cast as you were running in to a boss/pack of mobs. This in turn allowed you to apply the next disease upon arriving in position.
Has this line of thinking changed and I missed it someplace?
If you are starting at range, yes, it's better to IT as you run in. If you are starting in melee range, it's better to PS. The raw dps difference between the two scenarios is extremely slim, especially next patch where we have all of these free gcds (making the IT at range, although still better, not quite as valuable as it is now).
There's a reason it's called "Generalized Rotation". There will be occasions - such as being at range, but also numerous others - where you slightly deviate from it. Simply use your best judgment.
Searched this thread and the forum for it but could not find anything exact on my question:
Scourge Strike 3.3
Main effect: physical, mainhand, direct_dmg = 50% wpndmg + 400
Secondary effect:
Can't miss, independant crit (I just assume all crit talents work)
direct_dmg = main_direct_dmg * ( .25 * diseases_up ) * ???
I know that EP/E&M/CoE works, but that is only .75 * 1.13 = 0.8475, what else buffs the shadow portion?
From my DK on PTRI know that with EP up: Secondary dmg > Main dmg, so everything added up has to be > 1.0
PS: Needed for the DK module in SimulationCraft
Last edited by Starfox : 11/16/09 at 1:55 AM.
Reason: typo
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
Just because the secondary affect does higher average/total damage than the main strike, that doesn't mean everything has to add up to higher than 1.0, due to how critting works. The fact that the secondary affect essentially double dips into crits is going to push it's average/total damage up regardless.
It is greater than 1.0, but your logic for why that's so is incorrect (not that your logic matters for your purposes, only your math).
Ebon Plague (/EM/CoE), Blood Presence, Bone Shield, Desolation, Two-Handed Weapon Spec, Ferocious Inspiration (/Paladin Equiv) all are double dipped, and thus affect the secondary effect.
0.75 x 1.13 x 1.15 x 1.02 x 1.05 x 1.04 x 1.03 = 1.118 modifier (relative to the damage of the physical strike).
Both abilities have the same crit rate (i.e, both are affected by Subversion and Viscous Strikes) and both have the same crit damage modifier (236.9% - they're each affected by both CSD and Viscous Strikes).
Nathanael: massacre has the right idea when he says you can just enter in the stat weight for damage into max/min (after you divide by 2). It will produce the exact same result as if there were a weighting to average damage.
You're right, I hadn't even considered that. And either I was doing something wrong with my math earlier or the dropped decimal places were too significant and threw things out of whack, because when you do that (1.54 / 2 = .77) you get different results yet again.
In this ranking, QD Ferocity and Citadel Enforcer both top the 245 ilvl axes, if only just. Spire is still beneath them but the Ony25 sword drops below it. At the same time, it does seem to confirm that across weapons of the same ilvl, average damage will still pretty much trump all else (again, excluding the statless wonders unless the procs actually turn out to be amazing, which so far seems unlikely).
Actually there's at least one other thing you have to consider when calculating the effect of weapon speed on dps, and that's the uptime on your Rune of the Fallen Crusader. It can proc off all strikes as well as autoattacks, and as it is set to 2 ppm, a slower weapon will proc more often.
ETA:
When using the GotG, using death runes for putting up the diseases is probably a good idea, as this means you you always use your UF for SS. If you have 2 piece T9 you could actually look to to do a 20 second rotation that separates putting the diseases up (and thus separates your BS) by between 7 and 14 seconds, as this would allow you to proc the set bonuses more often. In practice this is a little bit more difficult that you might think as you'd want to ensure the disease ticks are spaced by at least a second due to the IC for WP.
It's only a small boost to the uptime: 21% vs 17%, but could be worth playing with.
I encountered a problem with melee hits in unholy lately, I was wondering if anyone else has had this happen as well.
On certain fights (mostly twins, but now also patchwerk apparently) it seems as if i have a very low amount of melee hits in unholy specc. Last night we did an immortal naxx run and on Patchwerk i noticed i got very low dps. opun checking the recount and later the World of Logs - Patch, it showed i only got 1 melee hit in. All the other abilities were correct, including everything that requires melee range (SS, PS, BS).
Last week i had a similar thing happen to the other unholy dk in our guild. On twins he got only half the melee strikes i got (i was blood specc at the time). Again, the amount of SS and PS was normal.
Can anybody tell me what is happening here? Is this an (un)kown bug or am I just very stupid?
I encountered a problem with melee hits in unholy lately, I was wondering if anyone else has had this happen as well.
On certain fights (mostly twins, but now also patchwerk apparently) it seems as if i have a very low amount of melee hits in unholy specc. Last night we did an immortal naxx run and on Patchwerk i noticed i got very low dps. opun checking the recount and later the World of Logs - Patch, it showed i only got 1 melee hit in. All the other abilities were correct, including everything that requires melee range (SS, PS, BS).
Last week i had a similar thing happen to the other unholy dk in our guild. On twins he got only half the melee strikes i got (i was blood specc at the time). Again, the amount of SS and PS was normal.
Can anybody tell me what is happening here? Is this an (un)kown bug or am I just very stupid?
The only explanation I can think of is that you weren't facing the mob. There is an angle where you can still hit your target with styles, but not with auto attack (~85 degree or something), although you would get the warning "You are facing the wrong way!" permanently. I don't see why you would stand at this angle on either of the fights, but perhaps some strange latency issue caused it. Otherwise, this is a very strange problem indeed.
Originally Posted by Frozn
You can be sure that I will never post something anymore. Your arrogance and snobism makes me feel sick, enjoy your idiot infractions. Your community just lost one of the best moonkin of the alliance (gearscore).
I encountered a problem with melee hits in unholy lately, I was wondering if anyone else has had this happen as well.
On certain fights (mostly twins, but now also patchwerk apparently) it seems as if i have a very low amount of melee hits in unholy specc. Last night we did an immortal naxx run and on Patchwerk i noticed i got very low dps. opun checking the recount and later the World of Logs - Patch, it showed i only got 1 melee hit in. All the other abilities were correct, including everything that requires melee range (SS, PS, BS).
Last week i had a similar thing happen to the other unholy dk in our guild. On twins he got only half the melee strikes i got (i was blood specc at the time). Again, the amount of SS and PS was normal.
Can anybody tell me what is happening here? Is this an (un)kown bug or am I just very stupid?
I experienced this last week on our Heroic Twin Valkyr's kill. We use "the door strategy". I had a high number of blood, plague and scourge strikes in my parse, but absolutely no melee hits on the same target.
In my experience, there's always been a very tiny window where I can land special hits but not white hits due to server positional lag or something like that, I dunno if that could be it? But then you do get an out of range error, I tend to notice it after a few sec when that happens, so this sounds different.
Im 100% i was facing patch directly and 99% sure i wasnt to far away from him.
The strangest thing to me is that it has never happened to me while in blood specc, but a couple of times while in unholy specc. Could just be coincidence ofcourse...
Ill try to keep an eye out for this in the next few weeks of raiding.
A few moments of testing confirms your findings Consider, I removed all gear except a basic weapon (no runeforge) and a greatness card. I waited for a proc, applied both diseases and then refreshed once greatness was down. Every time without fail the diseases used my current AP when I refreshed. It would appear that the functionality of GoD has changed.
I cannot confirm this. When I refresh via Pestilence when using Glyph of Disease, my current attack power not is used; as expected, it uses the tick strength already on the target. Moreover, any disease applied to secondary targets is applied at the strength of the disease on my current target. This means, for instance, rolling a strong disease on Anub'arak will push the strong disease to the burrower adds. This is a very significant DPS boost, even with diseases being reset every burrow.
I tested both in normal raid gear (with proc'ing trinkets, set bonus, sigil, runeforge) all up, refreshing after all fall off, with damage remaining. Likewise, I tested naked and using only Icy Touch for Frost Fever and Horn of Winter to change strength (easily applied and removed). In every case, rolling diseases retained the same tick damage regardless of AP, and always renew at the strength they were applied (with adds getting the disease at the strength of the target, not based upon your current AP).
Were you testing on the PTR? I did my testing on live.
All the current GoD discussion revolves around changes seen on the PTR. GoD currently functions as always on live.
Ah, indeed. Verified on the PTR. Percent scaling such as Bone Shield, strength/ap changes, etc do not change the tick strength until re-application (either via GoD or IT/PS), but anything that extends/reapplies it does. I wonder if this is a DK-specific change or all DoT-extending mechanisms.
Ah, indeed. Verified on the PTR. Percent scaling such as Bone Shield, strength/ap changes, etc do not change the tick strength until re-application (either via GoD or IT/PS), but anything that extends/reapplies it does. I wonder if this is a DK-specific change or all DoT-extending mechanisms.
This applies to all DoTs in 3.3. One reason why is so "skilled" DoT classes don't get a large dps increase by maintaining a buffed up DoT (spriest was the main reason), now that some caster DoTs scale with haste.
Millions of words are written annually purporting to tell how to beat the races, whereas the best possible advice on the subject is found in the three monosyllables: 'Do not try.'
I wonder, with this dot application mechanic change, would it be viable to use the Glyph of Scourge Strike to have ap buffs applied more smoothly throughout the rotation? Or would using it after a buff has fallen off lower tick damage?
11/14/09, 10:30 PM
