◊ chalon

So there are a few items I've gotten now that I've had to manually add into the sheet:
Leggings of the Honored - Item - World of Warcraft
Sinister Revenge - Item - World of Warcraft
Bandit's Insignia - Item - World of Warcraft

Additionally, the "Perfect" variants of all the uncommon gems don't seem to be there, eg:
Perfect Rigid Sun Crystal - Item - World of Warcraft

Also the "Berserking" enchant seems to be NYI, but I'd imagine a large part of it is due to the fact that proc rate information for it is unavailable on Wowhead. I'll be getting it enchanted on my daggers soon (hopefully in the next couple days), so I'll run some tests on it to see how often it actually procs.

EDIT: I did a bit of testing on Bandit's Insignia. The proc rate is listed as 15% on Wowhead. Hard to say how accurate that is, but from some combat logging it appears to have a 45s internal cooldown.

 

OengusSC

I was seeing some odd behavior similar to this too. You may want to make sure you weren't doing something like changing a Glinting to a Delicate in a yellow socket (therefore not getting the socket bonus). After messing around for a while with different gear sets and builds I came to the conclusion most of the "odd behavior" wasn't the sheet, but me :P I'm at the exact same hit rating to begin with on the sheet. Ultimately I found that Yellow - Glinting, Red - Delicate and the 2 Shifting's for the Meta produced the best results.

However, it only provided another 10 DPS over the current Yellow - Rigid, Red - Glinting/Delicate setup. With only a month left after the patch it's probably not worth the investment of 10-12 new gems in my opinion. Unless of course you have a stack of BoJ's that you don't have anything better to do with!


The Cursed Vision example is wierd though. I tried duplicating it, but I saw Rigid -> Glinting as an upgrade in DPS and then Glinting -> Delicate as a downgrade, as expected.

 

Lokar

Why is that expected? The socket bonus is 6 stam. So 5agi 5hit 6 stam > 10 hit 6 stam > 10 agi? Seems...wrong?

 

Amerilina

I posted a few pages back a combat log of about a hour of hitting a target dummy with Berserking on my main hand. Just from looking at it I believe it to be about the same as the mongoose proc rate, but I am not an expert.

Edit: Link to previous post.

Berserking Post

 

• Aldriana

I took a look at that parse and tried to work out the numbers, but it looked to me like you were dual-wielding for the test which made it hard to get exact numbers. If you could post another long with only the one weapon equipped, that would be helpful in terms of deducing the proc rate.

 

OengusSC

Ah, good point. I didn't even bother to pay that much attention to what the socket bonus was. Based upon the results I just presumed it was DPS based and didn't even bother to look. It does seem rather odd that a split stat gem is greater then either of the stats taken as a whole. The numbers are very close for me though.

Could this be rounding issues with some of the multiplier buffs and talents? I deselected Kings, Trueshot Aura, and the talent Savage combat and swapping the same gems resulted in Rigid Lionseye > Delicate Crimson Spinel > Glinting Pyrestone.

If these results seems unexpected and are wished to be duplicated:

My Armory:
The World of Warcraft Armory

Default buffs are being used on the sheet to the best of my knowledge. Cursed Vision has a Rigid Lionseye gemmed currently.

Existing gear Sheet DPS - 3433.89 DPS
Cursed Vision gem changed to Glinting Pyrestone - 3434.65 DPS
Cursed Vision gem changed to Delicate CS - 3433.59

Results are the same when performing the same gem swaps in other slots as well as long as they don't effect socket bonuses that have an impact on DPS.

 

Amerilina

OK, Any thing else I should do or keep in mind that will make things easy. Any other information you want?

Side note I have Icebreaker weapon, but as of the most recent patch it still does nothing so when that gets fixed I can post info on that also.

 

• Vulajin

Best way to test a proc rate is to use a single weapon performing only white attacks, with no haste gear equipped and no haste procs (be they meta gem, trinket, or otherwise), not performing Slice and Dice.

 

• Aldriana

So, let me just run down the list of proc rate testing best practices:

1) Equip only a single weapon.
2) Remove any items with haste procs you may be wearing (notably DST). Unless the item in question is the one you're testing, obviously.
3) Autoattack only. Do not use special attacks of any sort.
4) When you post the log, if you could also provide the base speed of the weapon being used and the amount of haste rating you had equipped, that would be helpful.

 

Amerilina

Here is the comabt log.

1.4 speed dagger
No Haste Gear

About 1 hour 45 mins of auto-attack on a level 80 target dummy.

Berserking.zip

If there is anything else I can do let me know.

 

Urotsukidoji

Good Morning, everybody.

I´ve tested 2 builds in the PTR_0.41 sheet.
A Combat build 5/51/5 with Vanir MH and Tracker´s Blade OH and a Mutilate build 51/5/5 with Tracker´s Blade MH and Swift Blade of Uncertainty OH.
The Combat build gets a result of 2505,05 DPS, the Mutilate builds gets 2344,14 DPS.
I am a litte confused now. I alwasy thought a Mutilate build beats a Combat build for now.
I used the same equipment in both builds. Armory Link Except the weapons, of course.
Did i made a basic mistake? Or are these Daggers realy that bad for a Mutilate bulid?

 

guljiny

On the spreadsheet people have found that mutilate only beats combat if the mob is murderable. If its not, combat wins.

p.s. is there some nice list somewhere of bosses and their murderable status?

 

Urotsukidoji

Thank you.
I changed Target Level form Boss to 72 and Target Type from Demon to Humanoid.
Combat build raised to 2567,17 DPS and Mutilate build raised to 2555,61 DPS.
Mutilate catched up but Combat still is 11 DPS ahead.

 

◊ pewsey

Can you share what buffs you had when you did this, and on what level mob?

Pewsey has heard about tact and discretion, but tends to regard them much as children view vegetables.
Nemesis: "Pewsey is single-handedly turning around every guy in the BB that didn't want to have kids."
Viator: Because I had a baby so I'm better than non-breeders.

 

◊ chalon

I ran some tests tonight as well. They weren't as long as yours, but I also tried to test Mirror of Torment and Bandit's Insignia at the same time. I removed all my haste rating gear, but since Sinister Revenge (1.8s) has 24 haste on it inherently, that couldn't be removed of course. I then ran a second test with 0 haste rating and Webbed Death (1.4s) instead.

Sinister Revenge (w/ 24 haste rating)
684 attacks landed
Berserking - 20 procs, ~2.9% proc rate (1 PPM expected ~2.9%)
Reflection of Torment - 4 procs, ~0.5% proc rate
Bandit's Insignia - 22 procs, ~3.2% proc rate

Webbed Death (w/ 0 haste rating)
668 attacks landed
Berserking - 14 procs, ~2.1% proc rate (1 PPM expected ~2.3%)
Reflection of Torment - 8 procs, ~1.1% proc rate
Bandit's Insignia - 19 procs, ~2.8% proc rate

Based on these numbers it seems rather likely Berserking is 1 PPM, since the percentages for both tests did turn out to approximately what you'd expect from 1 PPM. It can't have an internal cooldown as it will refresh itself even with a single weapon. My sample size isn't nearly as big as what you've got from your long parse, so that would be the final confirmation.

I did some further testing with all my gear on and both weapons equipped, and I can confirm that each weapon maintains its own Berserking buff independently of the other. Thus it is possible to have two Berserking buffs up at the same time, for a grand total of +800 AP and -50% armor (confirmed this on my character sheet).

Now, I don't know what the deal is with Mirror of Torment. FWIW, on single target DPS fights looking at my StasisCL parses, I get a range of 16-20% uptime typically. I wouldn't be surprised if it had an internal cooldown of some sort.

As for Bandit's Insignia, it does seem to have that 45s internal cooldown, as that's precisely the smallest duration between procs that I see. But you're pretty much guaranteed to have it proc fairly soon after the 45s cooldown, so I would buy the 15% proc rate listed on Wowhead.

Logs are here:
http://gamecanary.com/wotlk/ProcsSinister.txt
http://gamecanary.com/wotlk/ProcsWebbed.txt

Based on current known 25-man content, the following bosses are Murderable:
Naxx - Grand Widow Faerlina, Maexxna (all other bosses are Undead)
Archavon (Wintergrasp Raid Boss)
Malygos
Sartharion

 

Urotsukidoji

I have these buffs:
Commanding Shout
Leader of the Pack
Rampage
Windfury Totem
Moonkin Aura
Elemental Oath
Power Word: Fortitude
Mark of the Wild
Blessing of Kings
Agility Food
Bloodlust
Haste Potion
Drums of Battle
and Flask of Relentless Assault

I have 3 major Glyphs:
Sinister Strike
Rupture
Slice and Dice

And as is said, first try was on Demon-Boss, second try was on Level72-Humanoid.
Maybe i can do a 3rd try with less buffs.

 

Indz

I took both Combat Logs about Berserking and I found out that the Buff has an Uptime of 25% over the full length of the log.
The buff gives 400ap and has an Uptime of 15sec so you have an average AP buff of ~100ap.
The Buff also can override his own buff like Mongoose.

A quick math also shows that 15sec of 60sec are 25%, so it could be 1 PPM. (but I don't know if it's so correct)

 

◊ chalon

I realize now that I misread Mirror of Truth. It says on *critical hit* not on melee hit. That would explain why I saw a pretty bad proc rate in my tests compared to the uptime I saw in the parses I was looking at.

However, on the topic of Berserking to be honest the enchant seems a bit underwhelming if it is in fact 1PPM. I threw it into the spreadsheet, and at my gear level (5k DPS buffed range), I'm seeing that Berserking is only slightly (<10 DPS per weapon) better than Mongoose, and Massacre slightly better than Berserking. Given that Mongoose has no drawback, whereas Berserking has a very serious and PvP-unviable -25% or -50% armor loss associated with it, I'd really expect only picking up Berserking if I had an easier time getting the mats together for it.

As for Massacre, I do tend to agree that it has to be intended as a 2h-only enchant, especially since as a 1h enchant it outperforms Berserking for far less mats.

 

Leto

I am curious on this as well, but from a more specific tweak in talent points. My quibble is whether to put 3 or 5 points in relentless at the sacrifice of opportunity.

The spreadsheet says 3/2 is more dps as opposed to 5/0, but I'm noticing that it also lists a 90% up time of snd with 3/2 as opposed to 97% with 5/0.

I think the problem is that the sheet isn't modeling the situation of < 100% up time similarly to a rotation I would actually be using, as I would prioritize HfB and snd up time over maintaining the set finisher cycle. My intuition is that adjusting my cycle to maintain 100% up time would result in mutilate accounting for a larger portion of my dmg than in the set cycle, so opportunity would be more valuable there.

On the flip side though, max rs would allow me to maintain the cycle more easily, and that the spreadsheet isn't accounting for the cycle change to maintain upkeep, it is just assuming that I would have snd down more often.

I'd imagine this would be impossible to model without some form of simulation feed into a state machine, so what experiences have you all had when comparing the talent point shift with in practice cycles?


I'll likely go 3rs/2op for more burst while soloing, but I am still curious what the optimal would be.

Rogue at heart.

 

◊ pewsey

I tested both 5/0 and 3/2 on the 70 target dummies because I was interested in the difference between theory and practice.

In theory 3/2 does more.

In practice 5/0 does more for me because with 3/2 I can't keep SnD up, and my rotations go all out of whack.

I found 5/0 much easier to keep up SnD (but I don't have the Glyph) so things may change there.

Pewsey has heard about tact and discretion, but tends to regard them much as children view vegetables.
Nemesis: "Pewsey is single-handedly turning around every guy in the BB that didn't want to have kids."
Viator: Because I had a baby so I'm better than non-breeders.

 

• Aldriana

So, with regards to Berserking:

In Amerilina's log, I count 122 procs in 4420 swings, which works out to a proc rate between 2.3 and 3.2%; in PPM terms, that works out to somewhere between .98 and 1.39 PPM. Hence, it's certainly plausible - I daresay even likely - that it has the same proc rate as Mongoose (1.2 PPM). If this is correct, the actual uptime for a Mutilate build will be in the neighborhood of 30-35% for both hands, while a Combat build will be more along the lines of 45% MH and 25% OH. Hence, from a strictly PvE perspective, it likely beats Massacre on both weapons for Mutilate, though only on the MH for combat.

Relative to Mongoose: if we posit that the proc rate is the same as Mongoose, it basically comes down to which is a better buff, 400 AP or 120 agi + 2% haste. At level 80 conversions, 2% haste will give benefit roughly comparable to 65 points of haste rating, which works out to something around 90-100 EP. Hence, the EP of the Mongoose buff is around 330, whereas Berserking is 400. Hence, Berserking would appear to be superior to Mongoose for the moment.

Thus, the conclusion is: assuming Berserking has the same proc characteristics as Mongoose (and Executioner, for that matter), from a purely PvE standpoint Mutilate rogues will want to use it on both hands. Combat rogues will want to use it MH for certain; whether they want it OH as well depends on the 1H availability of Massacre.

Regarding Mirror of Truth: I think we need some longer tests to nail this down, but off the top of my head it looks like it may have a 60 second internal cooldown, as there's a couple of procs that occur soon after 60 seconds (65, 69, and 78 seconds) but none below 60. However, the fact that it also sometimes goes 4 minutes without proccing may just be an indication that it has a really low proc rate. We'll need to get an hour or two of autoattack with it before we can totally nail it down, I suspect.

 

◊ chalon

I'll try to run a much longer test tonight with Webbed Death, and that should help nail down Berserking, Bandit's Insignia, and Mirror of Truth.

 

• Aldriana

When you're doing that, try to get your crit rate as high as you can - I suspect part of the reason for the variance in your existing tests is that your crits are somewhat infrequent, so the dry spells for the trinket are probably partly just dry spells in crits. So if you can get your crit rate as high as you can, that will reduce variance and hopefully yield better numbers.

 

◊ chalon

Yeah, I'll just run it with my full set of armor with some higher crit pieces swapped in. That means the test will be with a great deal of haste rating, but I figure that will be easier to account for than an abnormally low crit rate.

 

• Aldriana

Haste Rating is trivial to adjust for, assuming we know much there is. It's only haste procs that cause a problem.