Has anybody reported it as not double procing as a bug? It would be the first weapon enchant off the top of my head that I can think of that doesn't stack on itself.

Deleted, wasn't thinking clearly.

This is good to know, and thanks for running the numbers. I have a question: it seems that armor penetration is going to appear on more and more gear in 2.3 and beyond. I assume that as you gather more ArP stats on the rest of your gear, the relative value of Executioner compared to Mongoose will go up? As it seems, to my limited understanding of the mechanics, that stacking ArP leads to exponential (and not linear) gains in DPS.

If this is true, and since it seems that the proc doesn't stack, it looks like Mongeese is going to be dumped for mongoose/executioner...now which enchant goes on which hand? Will it matter?

Jairek you have another typo there probably:

Should be 1832.22 as far as I can see.
And let us not forget the 120 agility which Mongoose provides significantly buffs Rupture damage aswell, which Executioner obviously does not.

We all just seemed to accept that executioner doesn't stack.

But I started a thread on the PTR fourms to find out, keep it bumped if you can.

It just seems odd that this would be the first weapon enchant to just, not stack for whatever reason.

WoW Forums -> Hortus, Executioner enchant?

200 minutes worth of data, that's nice!

The following section is on how to process the data.

I've been thinking about how to best analyse the data.
Uptime, while it's what really matters, is not a nice number to do calculations with. #of procs doesn't count refreshes. There is hope!

Uptime = probability executioner buff is active at any given time, t = probability executioner proc'd in the interval t>x>t-15 (i.e. in the 15secs prior to t)

Either using Poisson distribution - Wikipedia, the free encyclopedia and asking "what is the probability that executioner doesn't proc in a 15sec window, given it procs during the fight at a constant rate of p procs per second". Or noting that for ever smaller intervals of time, 1/n, the chances that a proc occurs in that interval is p/n, and calculating the limit (1-15p/n)^n as n tends to infinity the following relations are revealed:

Uptime = The proc's uptime as a ratio between 0 and 1
p="average procs per second during the fight"

Uptime = 1 - exp(-15*p)
and
p=-log(1-Uptime)/15

p is a very easy number to work with since it scales linearly with haste%.


The following section is on processing the data.

Using your first set of data:
Your mainhand is 2.6 speed, your offhand is 1.5 speed, so
1.5/(2.6+1.5) = 36.6% of your hits are from your main hand.
We can do calculations:

MH speed: 2.6
MH hits: 2789
Test Duration: 6152 seconds
p=procs/sec during test (calculated from uptime): 0.0228045
Number of procs = p*Test Duration: 140.3
Proc chance on hit: 5.03%


Using the second set of data:

MH speed: 1.5
MH hits: 4923.5
Test Duration: 6172 seconds
p=procs/sec during test (calculated from uptime): 0.0318476
Number of procs = p*Test Duration: 196.6
Proc chance on hit: 3.99%


Conclusion:
5% proc chance off the 2.6 speed weapon
4% proc chance off the 1.5 speed weapon

100minute tests should get reliable data. I'm too lazy to start calculating confidence intervals.

The following is pure conjecture and calculations assuming the data is free from anomalies:
Perhaps it procs not using the PPM mechanic (Which says: PPM/60*Weapon Speed = proc chance%) and not using a flat %on-hit proc mechanic. Perhaps blizz have come up with a new way to work out it's proc% chance that doesn't favour slow weapons. I know for a rogue assuming the PPM mechanic sinister striking with a 2.6speed mongoosed weapon gives more procs than backstabbing with a 1.8speed dagger.

Linearly interpolating (or using one of the formulae below) gives a 1.8 speed dagger a 4.27% proc chance.

Assuming no potency/haste:
For 2.6speed sword, white+SSing gives roughly 38 attacks per min, and with 5% proc chance that's 1.9 procs/min

For a 1.8speed dagger, white+BSing gives roughly 43.3 attacks per min, and with 4.27% proc chance that's 1.85 procs/min

The same calculation with 1.3 haste (SnD) and +20% energy regeneration (potency):
Sword: 2.40 procs/min
Dagger: 2.36 procs/min


Assuming the proc chance - weapon speed relation is linear:
proc chance ~= 2.57+0.945*speed
(Rounding to nicer numbers gives: proc chance ~= 2.5+speed)

If this is true it is a step closer to making daggers~swords, instead of daggers<swords.


The following is not based on the data.
Mongoose/Executioner & white:
Thanks to Jairek:
armor mitigation with executioner = .11337
armor mitigation with mongoose = .17179

Thanks to the fixed bug with the tooltip&mongoose proc, looking at my char sheet (this factors in the 120 ap, and 2% haste, but not the crit):
Unbuffed white DPS on char sheet without mongoose: 370
Unbuffed white DPS on char sheet with mongoose: 393.2

Now I add in a 10% miss chance and 30% (or 33% crit chance with mongoose) to those numbers:
No-mongoose: 444dps
Mongoose: 483.6dps

Now I do some mitigation calculations:
No-mongoose, no executioner: 367.7dps
Mongoose, no executioner: 400.5dps
No-mongoose, executioner: 393.7dps
Mongoose, executioner: 428.8dps

Mongoose appears to be better than executioner for white by looking at this, some calculations need to be done assuming the rogue is raid buffed to be sure. All the values scale linearly with haste so haste can be ignored.

So I'm going a slightly different path on my warrior and want to know what people think is better, mongoose or executioner.

Right now I'm fury, 25/36 with sword spec, dual-wielding swords, and trying to get as much haste as possible to get maximum effectiveness of DST, mongoose, WF, and extra attacks.

Contrary to what I thought, this build hasn't been nearly as useless as I thought it would be...seems my damage ceiling is higher and floor is lower yadda yadda, ANYWAY...

Mongoose wil synergize with the haste build, and with the amount of swings/minute I have I can count on mongoose being up more, which is 4% crit
(and of course 2% haste), which helps with flurry and keeps the whole haste train rolling.

Armor ignore is a flat dmg increase.

So assume for a minute that even though my spec is as gimmick as it gets I'm still going for the most effective enchant, which would I go with?

No, the more the Armor of your target is reduced, the more dps you will gain.

Image:Armor.JPG - WoWWiki, the Warcraft wiki

and please, dont start a Time to Live discussion again. This is plainly about personal dps increase.

Um, I said "flat damage increase" not "flat damage curve".

It isn't either, the lower the armor of your target at the moment of an Executioner proc is, the more you gain. I wouldn't call this a flat damage increase. See the table in my previous post.

Katherine, just a small point but Executioner is -840 Armour not -860

Havent read everything yet. Anyhow...

I've noticed, that a proc won't be counted as a new proc while the executioner buff is still active and only becomes refreshed.

Refering to wws and procwatch the ppm might be wrong.


edit:
Ive run two tests against a servant in bl counting the procs by myself, because of the crappy cl

2x 300 seconds
overall: 400 hits(600sec) with OH (1.5 Speed)
8 procs on average in 300 seconds

This is only vaguely on-topic, but does anybody know if there's a mod that tells you the current effective armor of your target, at least for bosses? It would require a database of the initial armor values, then factor in Serrated Blades, CoR, Sunder or EA, Faerie Fire, gear passives, and gear procs. Would be nifty, if it exists.

Not sure if they list any mods but if so it might be in this thread.
[RAID] Boss armor values

Its a flat damage increase because at a given number of armor penetration it gives you a certain percentage of extra damage, period, no side effects other than I suppose more rage.

Anyway, any thoughts on mongoose vs executioner?

Executioner is undoubtedly better than Mongoose for DPS Warriors, whilst Mongoose is better for tanking.
The jury only stands for Rogues, although seemingly Mongoose is a bit ahead.

Havent seen a jury on shamans either. Would love to know

Anyone on the task of testing this on the live-servers? As far as i know, no one has the enchant on my server, so i will not be able to assist with any test samples yet.

Would love to see some conclusions on whether this is better than Mongoose for rogues.

I have access to the enchant (I have it on my Main-Hand Sword), and would be happy to do some testing if someone is willing to explain exactly what the optimal procedure for doing so would be.

Standard way of testing proc rates is to go to one of the Servants in Blasted Lands, unequip OH and any haste itemization you may have, and autoattack the servant for a while - generally takes about a half hour of hits to get a decent proc rate estimate. Ultimately this will need to be done for weapons of two different speed, but whatever you can do would be very helpful.

Would it be wrong to MH Goose and OH Executioner? I'm a warrior and just wondering. I currently am going twards hast and plan on getting a 2.6 one-hand arena weapon for my OH. Just wondering the thoughts thank you much

I'd be happy to do this, and I'll work on it when I get home. I don't think I'll be able to do it for two weapons though (I have a [Talon of Azshara] so someone with a faster weapon could probably do some).

One other thing: would it be necessary to use a macro to cancel the buff when I get it? And is using an addon such as Procwatch sufficient?