"u" is not a word. I didn't give an infraction because the the rest of your post is fine, but please refrain from using "u" anymore.

Great that you discover this only 1 month after someone else made the same remark.

Anyways, Ikajo, I specifically used lvl 73 mobs because I figured AP would most likely increase with mob level. Not sure how meaningful the numbers from the lvl 73 scorpion test are, but I would expect most melee raid bosses to behave similar to lurker. Even more so because both tests had very similar results.

Edit: Please disregard this post, didn't look close enough at the results.

We use CoW a lot these days because of it's high damage reduction.

On prince I apply CoW in phase 2 due to the high damage he puts out. This allow the tank to use his HP shout. I then replace it in phase 3 to increase the DPS. The maiden is a good example where your healers can be locked out of combat for a long time. Reducing maidens physical damage can safe your tank.

Talk to tanks and healers if they need any help with healing due to the bosses high damage. If that is the case then putting up a CoW so the tank HP shout might a good idea.

Demo Shout has nothing to do with Commanding or Battle Shout. Only those 2 are actual buffs and a warrior can only buff either of those 2 for their group.

That is true, but I was thinking more of using his rage on aggro gaining abilities rather than the shout. In addition that a CoW is easier to keep up than a demo shout. All in all phase two goes a lot smoother for us when I do it

The tooltip on the debuff is wrong, it decreases AP. When they changed the curse from damage on each attack to AP reduction somehow that slipped through the cracks.

I see no indication that the CoW debuff description is anything else but a typo. Both thottbot and wowhead list the effect of the debuff as "Apply Aura: Mod Attack Power" (opposed to "Apply Aura: Mod Damage Done"). In addition, i hope a damage-reducing debuff wouldn't overwrite an AP debuff with a better effect.

@Dots: Could you verify that you really saw a 20% reduction in damage from Imp. Demo Shout on the level 73 scorpid (in Blade's Edge i assume - was using the same mob for CoW tests)?
I was getting the following data on that try:

Base Damage: 345-489 (level 73) --------------- 417
Reduced Damage: 285-429 (CoW -350) -------- 357 [-14.39%]

Actually I only tried regular DS on the scorpion because i didn't have the talent. I tested Imp DS on lurker, but the hunter just gave me rounded numbers since it was a learning attempt. It turned out to be 20ish% there.

I dragged a warrior with Imp DS to the scorpion today and you are right about the cap there at least.
Regular DS: 289-433
Imp DS: 285-429

I will try to get some exact numbers from lurker next reset, unless someone else beats me to it. Should be easier now that we got it down.

It really just depends who all you have for Warlocks and why you need to use it.

Without checking, I believe Imp. Demo Shout is more effective than a non-Imp. CoW spell. Improved it's still only marginally better. As long as your tank is diligent about keeping their AP shouts up, I would say go for a curse which boosts DPS in almost every situation.

..that said, CoW is really fantastic in PVP to toss on a Rogue who isn't your first assist target. A nice little wedge of -320 AP can sometimes mean the difference between life and death if a healer gets CC'd.

My 2C

To be precise, Curse of Weakness is -350 AP and for two talent points can be improved to -420 AP (350, 385, 420)
Demoralizing Shout is -300 AP, and for five talent points can be improved to -420 AP (300, 324, 348, 372, 396, 420)

Interesting. The next logical thing to check would be: is this a global 15% cap (ie lowered from 30%), or a consequence of the scaling of atk to level 73?

I guess I really should try and find another warrior account to test with.

I still expect a cap on AP reduction .. from the new data one could interpolate a max AP reduction of ~319ap for 15.1% - my old data for the regular level 73 scorpid (with a much more narrow damage range) yielded a cap of ~324ap.

Still interesting that you got a higher percental reduction on Lurker than i got on all my tries so far, but as dojke already mentioned one might attribute that to atk scaling.

If the cap exists though, only one point in Imp. Demo Shout is useful for pve then .. sad enough

THat all depends, the improved means you can use CoR.

.. given you have a lock to spare (we tend to lack reliable warlocks at times)

But granted, i forgot about that .. assume a possible cap around 325ap and a max AP reduction from one spell of 420 - CoR Rank 4 would just fit in there. I doubt i get the chance to get a hunter, a lock and a warrior with Imp. Demo Shout in one place for testing any time soon .. if anyone could try that, I'd be quite interested in the findings

I would be really surprised if there is cap on the ap reduction itself rather than an indirect cap due to the mob hitting 0 ap. As for experimenting... one good way is using enslaved elite demons. With the right addon, it is possible to observe the stats directly, just as if the demon was an ordinary pet.

I'm pretty sure there are bosses (and elites for that matter) where you do get the full benefit from imp. DS or imp. CoW; the main factor is probably which basic template the mob is based upon. The amount of AP an enslaved demon gets from strength and agility depends on the class of the NPC, for instance.

Don't you have to account for the attack power bonus the pet gets from the warlock's spell damage bonus, though?

Lurker Below Screencaps



First Frame: No Debuff
Second Frame: Demo Shout Only (Non-Improved)
Third Frame: Demo Shout (Non-Improved) + Screech
Fourth Frame: Screech Only

Demo Only (-300 AP):

Min: 2429 / 11948
Max: 2677 / 16895

Demo + Screech (-510 AP):

Min: 2551 / 11948
Max: 2799 / 16895

Screech Only (-210 AP):

Min: 1880 / 11948
Max: 2127 / 16895

Using the data Leica's post I did some basic analysis.

I discovered some things as I went along, so please disregard if it looks like the analysis jumps around. I tried to rewrite it after I was finished, but I wanted to keep the data I found. The more important section is in italics.

Averaging out the % reduction in damage:

-210ap -- 13.9% reduction
-300ap -- 17.7% reduction
-510ap -- 18.55% reduction

It appears, clearly, that there is a cap of some kind or the relation of %
damage reduction per point of -1ap is non-linear.

Assuming linearity between 0ap to -210ap and -210ap to -300ap and -300ap to -510ap:

0.066% damage reduction per -1ap between 0 to -210
0.042% damage reduction per -1ap between -210 to -300
0.004% damage reduction per -1ap between -300 to -510

The % damage reduction per point of -1ap is not perfectly linear. The error
between the first two values is 56%. That error is too high to completely
ignore. I suspect this has something to do with the large damage range of the
boss or of the unaffected base damage.

Assuming 0.05% damage reduction per -1ap and a cap at 18.5%, the Lurker Below
has 372 ap. This is more than previous range (as far as I am aware) of 300-350ap
for most mobs, even with generous assumptions.

The theory is that a boss mob's damage is: (AP_dmg + Wep_dmg)xModifier%. If the
range in damage comes purely from Wep_dmg, than the equation can be changed to:
(Range_dmg + (Modifier% * (AP_dmg - AP))). If you consider the variable to be
(AP_dmg - AP) rather than AP, the equation is a simple y = mx + b.

Assuming that at -510 ap the x component is equal to zero and it cannot be
negative. That means that the value b is equal to 9397 to 14096. We also have 4 points for the line: (0 , 9397-14096) (ap-300 , 9519-14218) (ap-210 , 10068-14768) (ap , 11948-16895). From those points, the first thing you can see is the damage range is in the b component because the difference in the range does not change (much) except when the boss had zero debuffs.

Range:
0ap -- 4947
-210ap -- 4700
-300ap -- 4699
-510ap -- 4699

This either has something to do with the boss' stats changing or any amount of ap reduction takes away a third component of dmg that we are unaware of.

Finishing the linear model, b is equal to 9397-14096. This means that mx is equal to:
mx=0 , x=0
mx=122 , x = ap-300
mx=672 , x = ap-210
mx=2552 , x = ap (2552 is without the extra 'range' that was previously mentioned)

The new equations are:
ap = 122/m + 300
ap = 672/m + 210
ap = 2552/m + 0

Taking the first two equations and solving for m:
0 = 550/m - 90 <=> 90 = 550/m <=> m = 550/90 = 6 and 1/9th or 6.11

Ap is therefore equal to:
ap = 122/6.11 + 300 ~= 320

Solving for the last equation will show if it is linear or not:
ap = 2552/6.11 = 417 != 320

I hypothesize that there is some third damage component that is negated as soon as any -ap spell is put on The Lurker Below (or there as a change to the Lurker between screens). For further analysis I would need a fifth screen shot of some other =ap variable, preferably 1/5 and 2/5 imp demo shout or a lower rank of demo shout would be preferable.

If the ap is equal to 320, than this third component would be an approximate 100-350 dmg range.

I wrote this post not in an attempt to disprove established -ap boss theory, but
to see how many points into imp demo shout I should go. I am not completely
convinced that the old theory works, though I like it. The only thing that I am
sure of is that 5/5 imp demo cant hurt.

Please post your thoughts on the attempted analysis.

I think that this particular riddle would be solved best by using a different proof.

What we need is this:

A) A defrocked warlock
B) A healer
C) A demon

1) Warlock (not wearing any clothes) enslaves the demon.
2) Warlock records the demon's AP and base damage
3) Warlock frees the demon and let's the demon hit him a few times, to roughly validate the tooltip base damage
4) Warlock casts rank 1 CoW on the now hostile demon
5) Warlock re-enslaves demon
6) Warlock records the demon's new AP and reduced base damage
7) Warlock repeats this process with progressively higher ranks of CoW until the demon reaches what should be a negative AP value (from the original AP)
8) Record results

You could also pair the lurker information with a different boss, but looking at Lurker in a vacuum won't tell you if you are reaching an AP cap, or if there is a % cap. Unless I am mistaken (which would not be surprising) you need a different mob with different AP values as a point of reference (because you can't see if Lurker's AP is hitting zero, or some % threshold has been breached).

My instinct is that each mob has an inherent amount of AP that you can reduce-and that it varies. I'll test this out on my warlock alt over the weekend.

As a minor aside the Gnomish Shrink Ray also has this effect when it succeeds in shrinking the target. It's a -250 AP debuff for 20 seconds.

I haven't tried it on many raid bosses in TBC but in the olden days pre-TBC it could shrink any boss in the game. Just an observation from shrinking things to entertain myself!

I´m sorry sepulture, but i have some bad news.

I wanted to do as you said, but unfortunately there is no pet-info page for enslaved pets. (Although I could have svorn i´ve seen one before.....)

anyways, got some more raw data for the "what does -ap give me" study.

Target: lvl 71 aether-ray

257-356 cor 5 (+135)
251-350 cor 4 (+90)
247-346 cor 3 (+65)
244-344 cor 2 (+45)
241-340 cor 1 (+20)
238-337 base dmg
235-334 cow 1 (-21)
232-331 cow 2 (-41)
229-328 cow 3 (-64)
226-325 cow 4 (-82)
220-320 cow 5 (-123)
214-314 cow 6 (-163)
201-300 cow 7 (-257)
196-296 cow 8 (-350)

I wonder, could you use a pet in a duel to check the AP? a voidwalker has around 258 AP at lvl 70.