So, it appears to me, based on this, that Armor Penetration rating applies after Sunders. When you reverse engineer the armor values for the sundered boars, one finds that the sundered boars have about 3600 armor, and with 25.39% ArPen they mitigate as though they have about 2650 armor; If Sunders apply first, you'd expect to see 3600 * (1-.2539) = 2685 armor; if they apply after, it would be 6200*(1-.2539) -2600 = 2025 armor. Since the former is a much better fit, it appears that ArPen applies after Sunders.

A typical raid boss at 70 has 7700 or 6200 armor; after sunders and FF, this is reduces to 4500 or 3000; thus, 1% ArPen will give either 45 or 30 reduction. Since it takes about 7.4 ArPen Rating to get 1% reduction at 70, you thus get either 6 or 4 armor reduction per point of ArPen rating. Since ArPenRating has replaced ArPen at about a 1:7 ratio, it appears to me that this is, for PvE purposes, a moderate nerf to the power of ArPen rating.

How this works out at 80, of course, depends on how much armor bosses have, which I haven't seen any numbers for yet.

It's zero. That was my first hope of testing armor penetration effects, but it proved useless.

It's not zero, it takes my 4045 AP bloodthirst down to 1128 from 1911 unmitigated. I'm trying to figure it out right now.

Experiment:
I took off all of my armor penetration gear, and barefisted a bloodthirst onto the dummy. I did 559 damage with 2127 attack power.

2127*.45 = 957.15 unmitigated.

559/957.15 = .584025 applied.

so therefore

.584025 = 15232.5 / (armor + 15232.5)

armor = ~10850.

A second test, I reapplied my armor penetration gear, yielding 3.38%, also bare fisted with 3104 AP. I did 824 damage.

3104*.45 = 1396.8 unmitigated.

824/1396.8 = .589920

.589920 = 15232.5 / (armor + 15232.5)

armor = 10589

10589 and 10850 are definately related by 3.38 percent.

It seems to have about 40% armor reduction, about 10280 armor, using a very unscientific testing method. (Note that the Training Dummies in Orgrimmar definitely have zero armor. It seems the ones here in Dalaran are different.)

(edit) More test details:

Wielding a single [Dirk] with exactly 1400 AP and no armor penetration rating, my damage range is 161-163. My experienced damage range against Heroic Training Dummy with no armor debuffs active is 94-95. My experienced damage range against Heroic Training Dummy with 5-CP Expose Armor active (3925 armor reduction) is 110-112.

Let's assume that the expected damage is always the mean of the minimum and maximum of the range (not necessarily accurate, but close enough for this). Thus, my expected base damage is 162, my expected damage with full armor on the target is 94.5, and my expected damage with Expose Armor up is 111. Thus, the base reduction equals 1 - 94.5 / 162 = 41.67%, and the reduction after EA equals 1 - 111 / 162 = 31.48%.

To figure out the actual armor count we use the formula A = R * F / (1 - R), where R is the damage reduction percentage and F is the "armor factor," equal to 467.5 * Attacker Level - 22167.5, or 15232.5 for a level 80 attacker (me).

For the base reduction of 41.67%, we calculate A = 10880.36. From this value we subtract 3925 and calculate the resultant damage reduction R = A / (A + F) = 31.35%. This is just slightly off the experimentally calculated damage reduction.

Now for the reduction after EA of 31.48%, we calculate A = 6998.72. From this value we add 3925 and calculate the resultant damage reduction R = 41.76%. Again, only slightly off.

It looks like our experimentally determined values are too close together to represent reality. However, we can see that the armor values are reasonably close to one another. Most likely the actual armor amount is somewhere between 10880 and 10924.

(edit 2) 10900 seems like a very good approximation for it. The damage reduction provided by this amount of armor seems to be slightly less than what "high armor" bosses had in BC (7685 armor, or 42.13% reduction, compared to 41.71% reduction here).

(edit 3) Together with Landsoul, took a lot of different approaches and mostly ended up corroborating the estimate of 10900. This will likely be a good value to work with in theorycrafting, going forward.

I believe this statement is false. The amount of armor removed (post non-rating based debuffs) is still proportional to the amount of armor penetration rating. Therefore, nothing has changed from pre-3.0 in this regard: armor penetration rating becomes better as it is stacked.

1000 armor, 93.84% damage done
900 armor, 94.42% damage done, .58% increase
800 armor, 95.01% damage done, .59% increase
700 armor, 95.60% damage done, .59% increase
600.. 96.21% ......... .61% increase
500.. 96.82% ... .61% increase
400 97.44% .62% increase

well I guess you are right, but the increasing gain is small. It's good to provoke contradictory statements.

still, the possibility to reach 0 armor on a boss is pretty much gone, it would require 740 Armor penetration rating.
7 ignore armor = 1 ArP rating => 740*7= 5180 armor penetration required pre 3.0 to reach the cap.

This is actually quite a nerf, on a 6800 armor boss you required 1950* ArP + executioner to reach the cap, now the same amount(counting that executioner is also changed to rating) would reduce armor by about 37% + 16% = 53%.
Or counting fully debuffed 6800 armor boss(which would only be sunder armor + CoR/FF now?) would have 3400 armor left before ArP comes into play, effectively removing 3400*0.53=1802 Armor removed.

So basically it's a (1950+840=2790, 2790-1802=988) a loss of 988 ArP if you are capped now before the patch, or about 35% less armor removed.

*(1950 ArP required if using Sunder armor, curse of recklessness and faeri fire, 6800-2600(SA)-800(CoR)-610(FF)-840(executioner)=1950)

Edit: my mistake, you are correct and the math bears that out symbolically.



Where D_0 is pre-armor DPS, q is the armor penetration % (as a decimal), w is the initial armor, and w_0 is the armor constant at whatever level.

If we imagine a small change in q (that is, delta q), then we should, after no small amount of manipulation, arrive at...



Which, as q increases, the delta D increases. So you are correct, again, my mistake.

Strange. This runs counter to previous algebra I've done on the armor formula and how a hypothetical rating might work. It could be that trying to set delta w = kw was the problem.

The effects of stacking Armor Penetration have not changed between live and beta; the mechanics by which it works, and the effects on low versus high armor bosses certainly have; but on any given boss, there's still a straight conversion of 1 point ArPen rating = -n points enemy armor. While the number is smaller than it was before, it's still a flat number and has no (known) diminishing returns that change ArPen's previous positive scaling with itself. So it's still true that the more ArPen you have, the better it gets.

right, I knew there was something wrong there, was sure that the cap was at 1350, but couldn't see why it didn't add up, thought it might have been with improved expose armor and didn't have time to check everything up.

However, my point stands clear.

Has the ArP Mace Spec model been tested for confirmation? Is the 15% applied before all armor debuffs, or after Sunders and what have you? Will this compete with Sword Spec with the Windfury changes causing a slight nerf?