MrGreen88

I was wondering, how is the BiS looking for Mutilate?

 

• Feist-Mok

Out of curiosity since I don't have all the new gear added to my copy of the spreadsheet just yet what final expertise numbers are you coming out to in those setups Aldrianna?

I'm just curious how much, if any, impact racials would have on pulling either setup (axes for orcs, or the mace setup for Humans/dwarves) ahead.

Edit: just went ahead and looked it all up myself. The mace setup has 154 expertise, which puts it substantially over cap, even without factoring in racials - Aldrianna have you looked in evaluating that DPS number into dropping points from Weapon Expertise to fill out CQC for the offhand?

Edit again: The axe setup is the same w/r/t Expertise. So what we're looking at is, with 2/2 WeX, is 29 Expertise. 1/2 puts us at a more managable 24, for those without racials.

Orcs running the axe setup can drop to 0/2, but don't see any DPS gain from doing so - they do get to go 2/2 Thrown Spec at no DPS loss though.

Dwarves running the mace setup can drop to 0/2 and sit at 24 Expertise, with 3/5 CQC.

Humans running maces can drop to 1/2 to sit at 27, or 0/2 and 22 with 2 or 3/5 CQC.

 

• Aldriana

It's worth noting that points in CQC are worth very little dps to the mace setup - more than nothing, which is why you take it, but I would give up anything significant to get it. So, for instance, humans dropping to 1/2 to get to 2/5 of course makes sense, but dropping the second point to get to 3/5 is almost certainly not.

To put this in perspective - for an NE, who is 23 expertise over cap and thus getting less than half the benefit of the last point in WEx, it's still a 12 DPS drop to go to 1/2 WEx and 2/5 CQC. Points in CQC for mace/dagger are only worth about 10 DPS, which just isn't worth sacrificing much for.

 

Rahdik

Just as an aside, Talonstrike is also a level 24 one-handed sword >.<

Do the same setups you mention work for the non-heroic 25 man pieces, or do the extra sockets on most of the gear make or break the setup?

Also on the topic of survivability for those that would benefit from moving points from Weapon Expertise and the points in CQC being a small difference, is it worth it to place the last point into Endurance, especially for the setups that lose some of the extra stamina and dodge from the lack of AGI?

 

Kytrarewn

Running through Mavanas's DPS spreadsheet, it looks as though mutilate is about 10% behind Combat in 3.2 "looks like BIS right now?" gear.

That's 4pc heroic + offset heroic legs, two best rings, Comet's Trail + Dark Matter, etc.

It seems reasonable that Mutilate will get a bigger benefit from the 2pc T9 than Combat will, so that should help bridge the gap (55 energy spam, and 60-75% of all finishers will be 35e rather than 25e), and I'm not sure that I hit every single socket's color correctly, which would lower the DPS a bit since I assumed all socket bonuses, but across 300 iterations of a 300 second fight, the estimated DPS hovered around 8975, whereas Aldriana's guesses at Combat put it at about 9800ish.

Still, it's a bit disappointing to see it fall behind like that.

ΜΟΛΩΝ ΛΑΒΕ


Jesus don't want me in a sunbeam
Sunbeams are always made on me
Don't expect me to cry, for all the reasons I'm gonna die
Don't ever ask your kick of me.

 

Mavanas

I used Aldriana's exact gear and talent setup this time. There were a couple discrepancies with what I plugged in before. Notably, when trying maces, I did not put one extra floating combo point into CQC, which makes some difference, but not a whole lot. In general everything lines up more or less except for the Mjolnir/Tailoring setup.

Here are the results using the simsheet:
Aldriana's base cases, using the horde version of the axe, which is faster:
Maces - 9585
Axes, full armor pen - 9629
Axes with MR, soft armor pen cap - 9482

As a minor note, I do find berserking to be slightly better than accuracy for Axes (9636 dps, not a statistically significant change from 9629).

Also Aldriana did not mention which food he was using in his analysis. I used armor penetration food throughout, which put the mace spec above the cap, so I replaced one of the armor pen gems with a deft ametrine and activated the socket bonus for the boots.

So something is amiss with Mjolnir Runestone, so I will state here my settings that affect modeling of the trinket proc:
45 second cooldown
15% proc chance
10 second duration
665 armor penetration proc value
During the proc, armor left is around 315, which translates into about 2.02576% physical damage reduction.
Trinket uptime is around 23%.
Fight duration is 300 seconds.

If you see any problems with the settings let me know.
For combat modeling, I am using 18/51/2 setup which I find superior to 15/51/5 by about 20 dps, however switching to 15/51/5 does not change the conclusions. I am modeling all synergies, including Embroidery with rest, and also cross-synergies between all trinkets. However I doubt synergies can explain the gap I am seeing here, they are of the order of 10-20 dps, and I won't even start to estimate the exact values because it would take too many iterations.

 

• Aldriana

I used Feast throughout. If you want to use optimal food, it would be Agi for the Mjolnir setup and ArPen for the others. Mace Spec still has a decent amount of room under the cap, so you shouldn't be capping just from food - even using ArPen food you're only at 92% penetration MH (and 77% OH). Unless, of course, I've entered some gear incorrectly, which I can't totally discount as a possibility.

Regardless, it sounds like there's some pretty major modeling differences between the two sheets, if I'm getting the two axe options within 10 DPS and you're showing them 150 apart. So we'll have to figure out what's causing that at some point - alas, I don't have excel, so I can't really pick apart your sheet to figure out where the differences lie.

 

Mavanas

For mutilate, Kytrarewn, the best dps I've seen in the simsheet is around 9200 (compared to 9600, I am getting for combat). That's using your gear setup, CT and Death's Choice for trinkets, and JC/Tailoring. Gemming for agi/crit in yellow slots, prismatic in the helm and agility in the rest. So it's not as bad as your numbers with CT and DM trinkets, but it's true that mutilate seems to lag behind both combat and HaT now. Looking at the gear, it's very hit heavy, which does not help mutilate asmuch as combat, and the fact that there are many armor penetration options, does not help mutilate either.

As for armor pen capping with maces, I am getting MH armor pen at 98% without armor penetration food. 502 armor pen from gems (three 34's and rest 20's). 516 from other gear, and 185 approximately from the mace spec. But it could be data entry errors or just a brain lock after hours of working with the spreadsheet.

 

• Aldriana

I've been reviewing my gear entry, and there was indeed at least one mistake - I'll have updated numbers available shortly. With correct numbers in place, I'm getting 505 ArPen from gear and 502 from gems for 1007 total, which is 1047 with ArPen food; with Mace Spec (~185), that comes to almost exactly cap (as in, within 1 ArPen). I'll have updated gear sets and damage numbers later this evening, but I suspect the errors will slightly favor the ArPen cap setup.

On a related note, if someone would like to doublecheck the gear entry on my beta version of the sheet, let me know and I'll send you a copy - it seems likely that if I've made one mistake, there are probably others.

Edit: Okay, updated numbers. Gear sets have changed from the last time I posted, but I'm not going to repost them until I get a chance to review the item entry in my sheet - take these with a grain of salt until such checking is performed. Regardless, these are the numbers I'm coming up with:

Mace/ArPen
AP Food	9829
ArPen Food	9876

Axes/ArPen
AP Food	9783.5
ArPen Food	9821.9

Axes/Mjolnir
AP Food	9684.2
ArPen Food	9687.2

 

Mavanas

Since your sets changed, I'll wait till you release them to compare output. It's good to see the order between axes full armor pen and axes with MR is about same as in simsheet because I was worried MR modeling was the cause of the discrepancy.

Also if some volunteers want to look through the new gear stats in the simsheet, I would appreciate it too. All new items are marked with 3.2 and heroic versions are distinguished as well. The socket bonuses are at the bottom, even though some of the socket color indicators are still missing. The latest file can be found here.

 

Bliksem

Combat vs Mutilate in 3.2

Whilst I respect the numbers given above me, are they given con or sans Murder?

 

nonmagical

These are Combat figures. So no, no Murder.

 

Kytrarewn

The mutilate numbers I put in, I accidentally neglected Murder. With Murder, against murderable targets, we're hitting 9300 DPS mutilate vs. 9700DPS combat against all targets.

One thing to consider though is that this is merely at the "BIS" level. It's far easier to get two 196.7DPS daggers than it is to get the hack and slash combat equivalents (Northrend Beasts OH, Faction MH for daggers; Jaraxxus OH, Anub'Arak MH for axes).

So, while progressing through Colliseum (Which is where everyone is at the moment) and finishing up your last Ulduar hardmodes, Mutilate might well come out ahead. That's a whole different issue of "best before hardmodes, best before anub'arak is open, etc" that each person needs to work out for him or herself.

ΜΟΛΩΝ ΛΑΒΕ


Jesus don't want me in a sunbeam
Sunbeams are always made on me
Don't expect me to cry, for all the reasons I'm gonna die
Don't ever ask your kick of me.

 

Mavanas

Kytrarewn,
you are also comparing apples and oranges. Simsheet estimates of combat dps are a bit lower, somewhere around 9600 range, not 9700-9800 as in Aldriana's spreadsheet. Also in your estimates you used wrong trinkets. In BiS trinkets, mutilate without murder does around 9200 and with murder does just shy of 9600.

The biggest nail in mutilate coffin though in my mind is HaT spec, not combat. Because on single targets, where mutilate was traditionally used, especially if target was murderable, HaT seems to outperform now. Of course if you get high ilvl daggers first, it might change the balance for a bit.

 

Serol

So I see a lot of posts about axes/swords vs maces, and I'm wondering is CQC now worse than those other 2 specs?
And also it seems that even on murderable fights that mutilate loses to combat, so does that mean mutilate is out now or am I missing something?

 

• Maestroquark

CQC falling behind is a natural consequence of there being no known PvE Fist from the 258 ilvl items.

What are you waiting for, a certain shade of green?

 

• Feist-Mok

Lionhead Slasher has been fixed on live. It's 1.5, just like it's horde equivalent.

 

• Aldriana

Okay, new BIS proposal. As usual, all drops are heroic if available:

T9 Helm, Shoulders, Chest, and Legs
Collar of Ceaseless Torment
Vereesa's Dexterity
Bracers of Dark Determination
Gloves of the Silver Assassin
Belt of the Merciless Killer
Treads of the Icewalker
Ring of Callous Aggression
Planestalker Signet
Death's Verdict
Comet's Trail
Remorseless
Steel Bladebreaker
Talonstrike
BS/JC
1 Nightmare Tear in the helm, everything else straight ArPen.
All enchants the obvious thing; boots are Cat's Swiftness, not because it's best for sustained DPS but because I suspect we'll find it to be a good idea in general from what I've seen so far.

As a check, fully enchanted and gemmed, this comes out to:
20 Str
1419 Agi
1467 Sta
2645 AP
651 Crit Rating
417 Hit Rating
154 Expertise Rating
194 Haste Rating
1007 ArPen Rating

So, food is ArPen, flask as usual, target fully debuffed, 5/5 Mace Spec, 1/5 CQC, Haste Potion used once in a 5 min fight.

So, at 15/51/5, doing the usual Xs5r5e cycle, this comes out to do 9884.1 DPS. Which is pretty good. But here's where it gets wierd.

If we drop Glyph of Rupture for Glyph of Evis, and 2 points of Blood Spatter and 1 of Ruthlessness to get 3/3 Improved Eviscerate, and then drop Rupture from our cycles entirely and just do Xs5e5e, that number goes up to 9945.0. And if we replace that cycle with the optimal one (which I'll define in a moment), our estimate gets all the way up to 9979.8. (And if we switch to Icewalker on Boots, we crack the 10k mark - 10008.8. But that's sort of a separate issue).

So, what is this optimal cycle? Briefly stated, the priority is to refresh SnD early as little as possible. Hence: we perform an SnD. We then build 5 CPs. If there are 5 seconds or less on SnD, we pool energy as high as we dare, and then refresh SnD. Otherwise, we drop an Evis and start building CPs again. If at any time SnD has less than 2 seconds left, we just refresh it with whatever CPs we have. That is: the priority is to refresh it as late as possible, more so than worrying about how big the SnD winds up being.

Note, of course, that this only works at extremely high gear levels. But assuming you do have said very high gear level, this appears to be optimal.

 

Nerio

If Ruptureless cycles become popular again, hopefully we'll see a change/buff to our 2 set since Blizzard seems to be trying to avoid that all together.

 

natox

What is really important to know is when exactly we have to switch to the unusual Xs5e cycle. I think it has something to do with the 4pc T8 loss and the ArPen hardcap.

 

Grunge

Well, whether or not a Xs/5e will be superior to Rupture based cycles hinges on the value of 2pc.

So unless 2pc is worth at minimum a ~120ep, we're stuck with Xs/5e.

Fans glory to the Gladiators,
Gods glory to the Heroes.

 

ekval

Hopefully two finisher cycles won't come as it will be damn boring to play Combat then again. But I guess with Hack and Slash you are still doing 3 finisher cycles due lack of Mace Spec.

 

mpiemont

On a tank and spank boss fight where it's easy to do the optimal cycle maintaining a Xs/5e cycle would be pretty straightforward. The question I have is if you drop your run speed enchant and the boss fears and/or stuns (edit: in a situation where tank is kiting boss around) would rupture still be beneficial to use (just before the fears) to maintain DPS while you are CC'd? Or would it still be better to spam that last eviscerate and not break the optimal cycle?


Edit:
One other thought, it may appear that energy capping during a bloodlust + AR situation is a concern. Would it be viable to use rupture there to prevent energy capping and not use rupture all other times to maximize DPS? Perhaps not if you're eviscerate spamming

 

PessimiStick

Either Rupture is better, or it isn't. Also, using a finisher that costs 10 less energy isn't going to help keep you from capping.

 

• Aldriana

The numbers I posted assume the 2% proc rate listed on wowhead for the 2/5 T9 set bonus. In order for Rupture to keep up in full BIS, the proc rate would need to be more like 6%. Which is possible, but I don't see that there's any reason at this stage to assume that wowhead is wrong.

The whole point of switching to evis-only cycles is that, due to being at the evis hard cap with ~50% crit (and the lack of the 4/5 T8 set bonus), Eviscerate just does more damage than Rupture - in fact, it does *enough* more damage to make up the 10 energy cost difference as well. Thus, while it's true that doing a rupture before running out means you do damage continuously while out, it's also true that if you'd done an Evis instead, you'd have *already* done more damage than that rupture ever will. Hence, it's still better to use the Evis.

I don't see how mixing rupture would make you less likely to energy cap - in fact, between the 2/5 T9 set bonus and the lower energy cost to start with, using rupture would appear to make you *more* likely to energy cap.

Also note that for this to be optimal you need to spec for it, at which point your ruptures are weaker than usual; if you're not specced for it, rupture will still be worth using. Thus, it's not a case of switching back and forth between them based on whichever is convenient at the moment - you use whichever one you're specced for, and that's the end of that.

As a final note, the proposed eviscerate cycle is actually the sort of thing we should be doing with rupture, anyway - as 15/51/5, it's generally better to use a small SnD towards the end of the previous one than to refresh early with a large one. It's just harder to describe (and execute) with a 3 finisher cycle, as you need to anticipate when both effects will be dropping and plan accordingly. The Xs5r5e cycle listed in the sheet is used because it's easy to model and isn't a bad approximation in terms of the number of each attack you launch, but in reality it's not anything you can actually execute in practice... nor should you even try, really. It's just that we don't have a good characterization of how to do the actually optimal thing, so as a reasonable approximation, we talk about rotations of that sort.