It seems to me that when Foxx stated that an increase from Dark Conviction effects damage additively and Desecration is multiplicative, his words were someone lost. My thought process here might be too simple, but if correct should show what Foxx meant.

Let's assume every attack can crit, so that Total Damage Output (before resists, misses, mitigation factors, etc) will be:

TDO=(1.05)(All Offensive Abilities)

With All Offensive Abilities being just that, the summation of all the attacks we can do. The 1.05 is the innate 5% chance to crit (which if I recall correctly we do have). If we increase the chance to crit by 5%, TDO would look like:

TDO=(1.05+0.05)(AOA)

Now if we instead increase all damage done by 5% instead of an additional 5% crit, TDO would look like:

TDO=(1.05)(1.05)(AOA)

If true, then the difference between 5% more damage and 5% more crit is the difference between:

(1.05)(1.05) ?>? (1.05+0.05)

(1.05)(1.05)=1.1025
(1.05+0.05)=1.1000

So 5% more damage beats out 5% more crit by 0.0025 (or 0.25%).

This is (what I think) exactly what Foxx was talking about. The 5% from Desecration modifies, in a sense, the bonus given by your crit chance; an increase in crit doesn't. Yes, if you have a 0% chance to crit, then 5% crit is the same as 5% damage; however, this is a pointless argument. There is no practical reason to consider this case in my mind.

GC addressed it. Pesil (in 3.1) will no longer spread diseases cast by others.

It is quite powerful now, especially on Sarth 3D with a spriest or two around.

Hi Ksearo,

I must first point out that we are in agreement that Desecration is going to be the choice comes 3.1. My finding suggests that if your crit chance is higher than 23%, Desecration is better than Dark Conviction. It is safe to assume that almost all Naxx geared DK are sitting on >25% crit chance, so Desecration is always perfered.

However, I disgree on the equation you have put forward in your thread.

When you consider damage done, you should consider the Expected Value. i.e. Expected Damage Done = (Chance of a Normal Strike)*(normal damage dealt by ability) + (Chance of a Critical Strike)*(normal damage dealt by ability)*(Critical Damage multiplier).

Your formula TDO=(1.05)(All Offensive Abilities) gives no such consideration. In fact, it is wrong on many levels. The rest of your equation is based upon this false premise, hence your comparison is wrong. Please do not take this personally, we are all working towards a common goal here

Even at 0% crit, a 5% crit increase is only a 5% damage increase if:

- Everything can crit AND
- Every crit does exactly 200% damage

Exactly. Since this isn't the case, it only gives further weight to 5% damage > 5% crit.

Even if we use your formula, we see the same results. If Expected Damage Done is:

Expected Damage Done = (Chance of a Normal Strike)*(normal damage dealt by ability) + (Chance of a Critical Strike)*(normal damage dealt by ability)*(Critical Damage multiplier)

then when we increase the chance to crit, we increase X:

Expected Damage Done = (Chance of a Normal Strike)*(normal damage dealt by ability) + (Chance of a Critical Strike + X)*(normal damage dealt by ability)*(Critical Damage multiplier)

If we increase damage done by a given percentage, we increase Expected Damage Done by said given percentage, even if you take all mitigating factors into account.

(1.05)Expected Damage Done=(1.05)[(Chance of a Normal Strike)*(normal damage dealt by ability) + (Chance of a Critical Strike)*(normal damage dealt by ability)*(Critical Damage multiplier)].

It's the Distributive Property: X(A+B)=AX+BX.

Maybe my thought process is too simple (and wrong), but it would seem to me that a flat 5% increase in damage will increase the damage that any single attack will deal by 5%, even if the given attack hits, crits, glances, partly resists, etc. An increase in crit only effects attacks that can crit, and doesn't benifit the critical hits that would normally be dealt.

Your formula is identical to his which can be seen by simple factoring, and thus yields the same results.

(Chance of a Normal Strike)*(normal damage dealt by ability) + (Chance of a Critical Strike)*(normal damage dealt by ability)*(Critical Damage multiplier)

(normal damage dealt by ability)*((Chance of a Normal Strike) + (Chance of a Critical Strike)*(Critical Damage multiplier))

(Chance of a Normal Strike) + (Chance of a Critical Strike)*(Critical Damage multiplier) can very easily be simplified to = Chance of a Critical Strike+1 assuming a crit multiplier of 2 leaving you with Expected Damage Done = (Chance of a Critical Strike + 1)*(normal damage dealt by ability) which is identical to Ksearo's formula. The only potential problem is that a crit multiplier of 2 is unrealistic, however for the sake of theorycrafting here it is fine-- it doesn't change the fact that crit is additive and dmg% is multiplicative.

Your math in #1956 is identical to that which ksearo used, he just approached it in a different way to get the same results, with the one exception being the crit multiplier. In his case he under values that crit multiplier at 2, and you vastly overvalue it at 2.3. The point is tho that it is completely irrelevant to the point ksearo is trying to make which is 5% dmg will always scale better than 5% crit at any realistic gear level due to the fact that it is multiplicative and not additive.

expired321,

No, it's not overvalued. See Vicious Strike. It's a Tier 1 talent, you cannot go to Tier 2 without it. Sure, you can argue that it only affects PS and SS, but then if you are an Unholy DK and SS is not your top damaging ability, I don't know what rotation you are using


Sure, if the CDM is 2, you can simplify E(TD) to TD*(1+crit chance), but CDM!=2, at least not for our top damaging ability (i.e. SS) so it changes things considerably

I will elaborate it with some real numbers. Hope this helps.


We agreed that we should look at the expected value of a Total Damage done by any given ability, so

let TD = Total Damage done by any given ability without Desecration or Dark Conviction talented
let's assume your crit chance is 15% and your critical damage multiplier is 2.3 for argument sake.

so E(TD) = 0.85*TD + 0.15*2.3*TD = 1.195TD

with Dark Conviction talented, your crit chance increased to 20%

E(TDDC)=0.8*TD + 0.2*2.3*TD = 1.26TD


with Desecration talented, your damage is increased by 1.05

E(TDD) = 1.05*E(TD)=1.05*1.195TD=1.25475TD

As the number clearly demonstrate, Dark Conviction win over Desecration in this case.


Again, I think we are in violent agreement here. Desecration is going to be better than Dark Conviction come 3.1 for reasonably geared DK

Ok, here is how I see it.


You are correct, at some point Dark Conviction is better then Desecration because of Vicious Strikes (or any talent that increases the critical multiplier attached to any given ability). However, I feel you over shot the crit chance required.

Let's deal with E(TD) VS TD*(CDM -1).

E(TD) VS TD*(CDM -1) and E(TD)=(1-c)*TD + c*TD*CDM


E(TD) ?= TD*(CDM -1)

(1-c)*TD + c*TD*CDM ?= TD*(CDM -1)

TD[(1-c) + c*CDM] ?= TD*(CDM -1)

(1-c) + c*CDM ?= (CDM -1)

You stated:
This I question. How did you get this value? A CDM value of 2.3 would only apply to Plague Strike, Scourge Strike, and Death Strike.

Runic Focus gives us double damage for our spell critical hits, but other then that I don't see any other talents that modify the CDM of any of our abilities. Now, you are correct that Vicious Strike will increase the overall value of our CDM above 2 if we average it out, but I don't see how you can get a value of 2.3; that would be the value if it increased the damage done by ALL critical hits by 30%.

For now, let us assume a CDM value of 2:

(1-c) + c*CDM ?= (CDM -1)

(1-c) + c*2 ?= (2 -1)

(1-c) + 2c ?= 1

2c ?= 1-(1-c)

2c ?= 1-1+c

2c ?= c

For this example, this equation can only exist is if c=0, or your chance to crit is 0%. This is the case that most people where referring to. This case assumes that no talent increases the CDM of all/any signal ability.

Now, we know for a fact that the CDM is greater then 2, but we don't know its exact value. My most recent log in Recount for KT puts my SS at 28.7% of damage done and 0.6% for Plague Strike. So, let's assume that SS damage and PS damage is 30%. If my thinking is correct, the averaged CDM value would be

230(0.3) + 200(0.7) = 209

So let's round (in favor of Dark Conviction) to 2.1 for our CDM. We then have:

(1-c) + c*CDM = (CDM -1)

(1-c) + 2.1c = (2.1 -1)

(1-c) + 2.1c = 1.1

2.1c = 1.1 - 1 + c

2.1c = 0.1 + c

2.1c - c = 0.1
1.1c = 0.1

c = 01./1.1

c=0.09091

So, with a CDM value of 2.1, Dark Conviction is only better then Desecration if your crit chance is below 9.091%.

The lower the value of CDM, the lower your chance to crit needs to be for Desecration to be better then Dark Conviction. The converse is true as well; the higher the CDM value, the higher the chance to crit you will need for Desecration to be better then Dark Conviction. Since the CDM is determined by what percentage of our total damage Scourge Strikes and Plague Strikes are, we can then say that as Scourge Strike and Plague Strike go up in percent of total damage, the required critical strike chance for Desecration to be greater then Dark Conviction goes up. This can be shown by:

(1-c) + c*CDM = (CDM -1)

1 + (c*CDM)-c = (CDM -1)

1 + (CDM-1)c = (CDM -1)

1 = (CDM -1) - (CDM-1)c

1 = (1 - c)(CDM-1)

1/(CDM-1) = (1 - c)

c = 1 - 1/(CDM-1)

If it is a good approximation for CMD = 230(SPD) + 200(1-SPD), then we have:

c = 1 - 1/[230(SPD) + 200(1-SPD)-1]

c = 1 - 1/[30(SPD) + 199]

Where SPD is Scourge Strike and Plague Strike Damage.

Any the end of the day, Desecration > Dark Conviction.

Ok, did some testings on a 100% full health Heroic Target dummy, in full T7 Naxx gear:

SS did 30% of my dmg, no surprise there.
Melee did 18.5%
Death Coil did 12.6%
Wandering Plague did 8%, but I'm unaware of whether the 1 hp dummies next to the Heroic one count as only 1 point (like Necrosis) or if the overkill counts in Recount.

Skipping down, it shows Necrosis doing 3.1% and BCB doing 2.4% (spec I was using was 12/0/59 with points in BCB instead of IUP). Necrosis was doing exactly 20% of whatever the melee hit was, so no more Black Ice or Ebon Plague bonus for Necrosis.

It looks as though Necrosis is worth about .6-6.5% dps per point, as is BCB. This makes me think that it might be better to go back to 17/0/54 and skip Necrosis in favor of Dark Conviction, unfortunately you just about have to put 2 points into Necrosis to get to tier 5.

Excellent ksearo, you got the idea!

A CDM of 2.3 is based on Vicious Strike talent and admittedly applied too generously across the board in my initial post. But as I stated yet and again Desecration > Dark Conviction in current setting. We were in violent agreement

I tried to get some stat weights for 3.1 Unholy.

- I used http://elitistjerks.com/f72/t50274-k..._beta_0_7_3_a/ (great tool, take a look!).
- 12/0/59 build and glyphs used.
- Gear is pretty much what I have at the moment.
- Ghoul is assumed to have 100% uptime.
- I did EP calculation 10 times then throw away min. and max. value of each stat and averaged and rounded up final value, it is not perfectly perfect but it is good enough to point potential upgrades.

Stat	Value
AP	1.0
STR	3.2
AGI	0.9
Hit	2.2 (0.45)
Crit	1.3
Haste	1.1
ArP	1.0
Exp	1.2
Armor	0.027

Cpt. Obvious note: hit value goes down to 0.45 after melee cap (8%) and to 0 after spell cap. Expertise value goes to 0 after 6.25%.

Loot rank list is here - just keep in mind that it uses pre-cap hit value and quite random weapon dps value (I set it to 10 for the sake of it).

I think We are Forgetting something between the comparative of Desecration and Dark Conviction: "Targets in the area are slowed by 50%"... yes maybe is not the best thing for raiding, but It can help in some Cases (like the fight in Gluth and walking zombies), as I recall is not only your DPS what It counts... Its how can you help the raid, and if you can have an ability that can Increase Dps (in any way) + something else... Why don't take it?

In Another order of Ideas... have someone check the new build in MMO? Looks like now u have to select NoTD to get Perma Ghoul... Can anyone confirm this plz?

Tenemit - NoTD is indeed a requirement for taking Master of Ghouls on PTR, and it is planned to stay like this (a blue confirmed it).

Just wondering, is the main build going to be the 12/0/59 or 17/0/54 for blood subspec in 3.1? Also, how are the numbers for 0/10/61, is it still viable when compared to blood subspecs, and if so instead of going with BCB 2/3, putting points into utility talents, i.e CE and Frenzy? (which i believe the utility has been previously mentioned, but not much discussed)

I'm quite surprised how high armor penetration is valued. With it only affecting white damage + blood strike + plague strike + BCB (which together is what... 30% of our damage ?

In surfing through other threads I stumbled upon a great resource for ArP, and most likely will answer a few of your questions. While I don't think he has all of his math 100% right in the link I am providing, it is pretty good for those who wish to get a feel for ArP, and why it's valued highly.

Armor Penetration

Ok, just did a 5 minute test on PTR with this build: http://talent.mmo-champion.com/?deat...0&version=9742

I stripped out the Wandering Plague dmg done on the two other target dummies, as it was throwing off the results.

Raw data

SS: 259888
Melee: 158008
DC: 134279
WP: 100988 (36507)
FF: 57368
BP: 55516
BS: 45756
Nec: 30076
PS: 27516
IT: 25747
BCB: 23025
Bandit: 14584

Total: 936751 (872270)
Ghoul: 88881
Time: approx 5 min

The numbers in parenthesis are the modified numbers for Wandering Plague. By stripping out Bandit's Insignia (as it's not relevant) we get a total of 857686, and dividing each amount by this number we get a more accurate reflection of each abilities contribution to overall dmg.


Scourge Strike: 30.3%
Melee: 18.4%
Death Coil: 15.6%
Wandering Plague: 4.2%
Frost Fever: 6.7%
Blood Plague: 6.5%
Blood Strike: 5.3%
Necrosis: 3.5%
Plague Strike: 3.2%
Icy Touch: 3%
BCB: 2.6%


Rotation was PS-IT-BS-BS-SSx4. I didn't use UB or Gargoyle, and Glyphs were IT, SS, and Ghoul.

Based on these numbers, BCB seems better than Necrosis, point-for-point (testing was done on a full 10million health Heroic Testing Dummy, so the Necrosis bug on dummies wasn't an issue). I think I'll be dropping necrosis in favor of Dark Conviction points.

Just curious so I tried to figure out what value of ArmorRedux (%) must be present for 3 disease OB to be equal to 3 disease SS (assuming 10/10/51).

So, i had
(OB + OB bonus dmg)(armor mod Y) = (SS + SS bonus dmg)
gave me an equation of
1.32xy + 762.3y = .892x + 708.2
>Assuming OB glyph, 3 disease, EbP, Outbreak
so, i had 2 variables so i figured out what x (wpn dmg) was by assuming 3400 AP and BoH (758 base wpn dmg) which gave me an x = 1559
plugged in for x
2057y + 762.3y = 1391 + 708
so, simplified to
y = 2099/2820
y = .744
So, you find that to get the ArP% to get y = .744

So, without 0 ArP, bosses have a y = .6 (40% armor)
So you have to reduce armor to 25.6% in order for OB to equal SS.
Due to the way that armor has a curve, I don't know how much ArP you actually need to reduce their armor redux by 14.4%. This also is not comparing crits, crit chance, or Viscious Strikes, which would then require OB to have more ArP to achieve the same goal, but it is a ballpark figure.

ArP has been changed so that it reduces the targets armor by a percentage. At lvl 80, 15.39 armor pen rating will reduce the targets armor by 1%. So, to get a reduction of 14.4%, you would need 221.616 ArP rating.

Is that considering the armor penetration buff in the patch?

First off I want to thank you guys for all the help. I got back into WoW a little over a month ago and immedietly made a DK, because of the information on this site I was able to become one of if not the top dps in my guild, and we have had naxx on farm for a while.

With that said, I have a question about specs in 3.1. I've been following both this board and the Blood dps thread and I'm wondering what spec is doing the most damage as of 3.1. I love Unholy, but I've been thinking about switching to Blood for the increased survivability while I learn Ulduar fights.

Keep up the good work and thanks for the help. And if this question is in the wrong spot, please let me know and I'll move it.