Mysticum

In my Rawr, Massacre is really superior to Mongoose. Have you really loaded your gear into it correctly?

Note also that the armor penalty for Berserking will be lowered in 3.0.8.

 

Monedula

I got Rawr 2.1.6 myself and for me massacre appears higher (50% more about) then mongoose. So massacre is the way to go. You must have real weird gear for that to be worth more.

Looking at berserking it seems really high. Discussion on it has been that it indeed is better, but maybe only 25% then massacre. To have 100% better then massacre seems a bit overrated. Maybe the 45 second cooldown has not been implemented for it in rawr.
To check I switched to bear stats. Mongoose there is also not correct I think. It is 2-ce as good as the 35 agility enchant. You yourself said overall mongoose is about 40 agility spread out over a fight. I myself think it is really close to the 35 agility enchant. Rawrs says it is 2-ce as good (i.e. 70 agility). I don't think so.
I think Rawr has problems counting the value of procs. I do believe them on strength, stamina, agility and all, but the procs just seem wrong.
One more thing Rawr is wrong on is expertise. It has only offensive value (as tank). It totally forgets that expertise will decrease the parry chance of a boss which will not let him get haste buff that much. Some raw calculations were done and 1 expertise is about 1/3 of the mitigation value 1 agility has (different for each boss though).

And Aldhissla, yes I meant overall score.

 

• Polynices

Astrylian has addressed the expertise in the bear model issue before (I was the one pestering him about it). He said he needed a detailed model of how it affects mitigation before he can plug it in. There was also some debate about how easily it is modeled since it doesn't apply to some bosses: The Druid Wiki Rawr discussion Is the Bear model valuing expertise correctly?

 

kalbear

I think modeling it situationally (with a checkbox) would be fine. While it's very hard to make perfect, you can make some very good basic assumptions about what it can and can't do with just a few variables.
At best, parry haste is going to increase the bosses's overall attacks by 15% (base parry chance) * 30% (average hasted attack speed) = 4.5%. Each % of expertise reduction reduces that by .3%. That's the absolute best case for the extra damage from parry haste, no matter what. The variables that can modify this are:
base parry chance
is parry-hasteable
damage per swing
swing timer
length of the fight

From those and the information from the character sheet we can determine how many more swings we'll get from a boss and how much they'll do. From that, we can determine how much damage is mitigated in a given period, which can be simply output as a value. If you let the user input how much damage as a percentage is the base melee damage, you can then determine the overall value of expertise as a mitigation tool for that fight.

I think that's about the best you're going to be able to do. Have it optional, have it user-entered, and have it be situational based on the fight. But that should at least give good estimates as to how valuable it is relative to the rest of the stats.

 

a civilian

In Rawr, both berserking and mongoose seem to scale inordinately well with haste. Checking windfury and improved windfury buffs results in berserking improving by a factor of 1.26, mongoose improving by a factor of 1.31, and massacre improving by a factor of 1.09. Perhaps the internal cooldown is not accurately modeled for either effect?

 

kalbear

If there is an internal cooldown for mongoose it's particularly small. I've had it refresh itself repeatedly when swiping multiple mobs.

 

RareBeast

From memory, the proc chances of Beserking & Mongoose are based on your unmodified attack speed. So in kitty, 1 proc per minute will mean 1 in 60 attacks will proc (1sec attack speed). 20% haste from Windfury brings your AP speed down to .8 secs meaning you do 75 white attacks instead of 60 which on white attacks alone will take you up to 1.25 procs/minute on average.

So haste has quite a large affect on mongoose & beserking because it can greatly increase the proc rate due to the raw increase in number of attacks.

 

Arî

To start with - both enchants do not have an internal cooldown. But Rawr's handling of procs is still not very useful, in my opinion. [Romulo's Poison Vial]'s proc was calculated as "xx Weapon Damage", for example. Furthermore, the [Dragonspine Trophy] was granting static haste in Rawr's calculations instead of procing.

I gathered some data for berserking in the last weeks. Today, I tried berserking with windfury with the same circumstances at the dummys. It was the worst try ever. I know, 40 minutes aren't long enough to get very reliable data but the worst try out of ~10 ... It should have been at least one of the best if haste would have such a great influence

I think, berserking has a common ppm-mechanic, which means: The chance is about x,x procs per minute, calculated by the current attack speed. So - as ridiculous as it seems - berserking (more general: ppm procs) becomes worse with more haste (at least "haste" from buffs ... I don't know it for "real" haste rating) for ferals.

I will try on it the next days/weeks but I do not thinkt that the data will change a lot

 

Monedula

Titanguard, the 75 stamina on weapon enchant will not be implemented in 3.0.8. They are looking at other options but they will not be there in 3.0.8. It is not that hard on druids since mongoose might be better (depending on your gear).
For druids the 75 stamina would have been a good option if you needed stamina. If you have comfortably enough stamina mongoose still seems like the best option.

[edit]

The berserking enchant has it's -25% armor reduced to -5% armor in patch 3.0.8. This minor debuff, considering you are not getting melee hit, is not that bad anymore, making berserking a better choice over 110 ap. The warning I placed next to the berserking enchant has been removed in the first post.

 

RareBeast

That is particularly the case when our Kitty gear has recieved such a big armor buff with the patch. Even with the -5% armor, we still have much more Armor than we did pre-patch.

This has removed any hesistation I had in getting the enchant (other than the cost!)

 

kalbear

Looking at rawr's code for berserk and mongoose, it appears that it calculates very incorrectly based on the PPM mechanic and how it works. It greatly overestimates the uptime of the proc compared to real life testing.

It's quite possible that berserking works on the current attack speed and normalizes based on that; I know that for some proc mechanics that's exactly what they did. It would make sense. In which case, haste will actually reduce your chance of a proc by a fair amount.

 

david0925

Straying a bit off from the Berserking vs Massacre discussion, for tanking I've still decide to use 35 agility over mongoose. While avoidance itself has uncertain elements, I think as a tank it is better to have staple elements over procs because taking more steady damage is important. Only exception I can think of is the old Commendation of Kael'thas, while the trinket is a proc, it procs at a time when you actually will need it the most.

Back on the topic of Berserking vs Massacre. While haste will reduce the change to proc, the uptime over a extended period of time should not be affected because it is a PPM mechanic. If berserking is better with zero haste, it should still be better with 2000 haste, unless there are something I am missing. Sure, you are getting a lower proc chance due to faster swing speed, but you are swinging faster to offset that.

Maniq is my hero

 

mesullivan

You're forgetting about special attacks.

If the proc chance works off current attack speed, then white hits will proc at the same rate no matter what your haste is. Most of your yellow hits are limited by the hard GCD or energy/rage, not by your swing timer, so haste doesn't speed them up. But the reduced proc chance based on your current attack speed still applies, so haste will in fact reduce the overall proc rate, by causing specials to proc less often.

 

david0925

Yes, spending too much time with OoC has made me overlook that. Still, that means that Berserking will have a reduced uptime assuming you have the same number of special attacks. This is partially offset by the fact that you will be doing more specials with haste through OoC. Will it make up for the loss? I haven't really looked at it yet, but unless Berserking and Massacre were awfully close to begin with, I don't think things will change very drastically.

Maniq is my hero

 

Monedula

Calculated berserking vs massace, mongoose vs 35 agi are about the same. I think I know the reason why they are valued more is that you are able to eb smart. Do a rip when they proc. The rip does way more damage. That goes for any dots. use +spelldamage trinkets for affli warlocks are probably worth more then to destruction warlocks.

 

david0925

In terms of DPS, yes, but only in a sense. It is very difficult to hold off a rip for trinkets to proc because it'll mess up a good rotation (not the best word, but bear with it) if Rip has already fall off. On the other hand, if a trinket procs and you're at 1 combo point with 10 energy, you're not going to get a good rip off it no matter what.

In terms of tanking, Mongoose may provide a good mitigation for a shorter amount of time, you don't know when it'll go off and even when it does go off, there is nothing to take advantage of, except maybe announcing on vent that you have 10% more dodge.

I understand that both agility and mongoose have volatile elements because that's just the nature of avoidance. However +35agi has a less volatile component and it's what I think, is more important for a tank. While I understand that this has been said briefly on the first post, I wanted to expand on that.

Maniq is my hero

 

kalbear

They're valued more in Rawr only because of the incorrect calculations. there's no optimizing based on when you could rip or anything like that.

In terms of 35 agility vs. mongoose - I think that it's somewhat unreasonable to suggest that the 35 agility is more reliable given that it's avoidance. More agility over time is going to be more valuable and let you dodge more than less agility. Unless you have 70-80% avoidance, there's no such thing as 'more reliable' dodge. Furthermore, mongoose's secondary benefit of being up 100% of the time against multiple targets can't reasonably be beat.

 

thalys

This is a really good point I've had to explain to some people I play with--the difference between 40% dodge and 42% dodge (randomly chosen numbers, please don't obsess over them) is not in any way, shape, or form a noticeably more "spiky" incoming damage profile. If anything, it's *less* spiky, in practical terms, until, as you say, your avoidance ends up in the 3/4 range--at which point your healers had still better be healing as if 2 hits in a row were going to land, because that's still going to happen around one in every twenty times the boss swings at you. Anyone who pooh-poohs Mongoose on the grounds of damage spikiness is not thinking clearly about how incoming attacks and damage actually aggregate.

 

Monedula

I know you are thinking of your first encounter with Patchwork 25 when you were a bit undergeared (at least I was and a 2nd hit could kill me)
It all depends on how hard the boss hits. If it hits for 1/4th of your life and you have 75% avoidance the healers can heal on pretty much on easy mode. The chance it hit you 5 times in a row is really, really low. And if it does there is enough time for some swiftmend or a panic button you as a bear use.

 

thalys

Eh, it was a general statement. If 2 hits landing back to back isn't that big a deal, then they can heal as if it's not a big deal. In most fights, that's exactly the case. But, even if it doesn't kill you, 2 Patchwerk hateful strikes back to back is more than enough that your healers can't heal reactively--which is my real point. The specifics of what "heal as if 2 attacks will land back to back" means depend on the fight. But, they'd better do it, whatever it is.

(Technically, when I made the post I was thinking of Sartharion, who has the annoying habit of hitting like a truck because he's angry that you killed his cousins, and then breathing on you, or vice versa--which gives you an effective health pool of something less than 10k to play with. I haven't had a problem with Patchwerk-25 that wasn't directly attributable to bad or undergeared healing. But, either way, the general point remains--75% avoidance doesn't mean your healers can behave as if every fourth attack hits and the other 3 don't, at least not if they want to collect gear instead of paying for repairs.)

 

GleamKT

But what are you trading for that 2% extra dodge? If you are trading 500 or 1000 armor then your incoming damage will most likely be more spiky, and you could easily be taking more damage overall as well depending on your gear. Even if you do take slightly more damage overall the armor is going to make your incoming damage less spiky, which is a good thing for your healers. I much prefer armor + sta above avoidance and I really don't like procs for the avoidance I do get (Mongoose, Nobles Deck, Idol of Terror, etc).

I took a 35 Agi enchant for my staff over Mongoose but I'd prefer an armor enchant over both. (my gear is lvl 200-ish).

 

kalbear

In this case, you're trading a bit less than 1% dodge all the time for 3-4% dodge some of the time, along with a bit of haste.

In order for an armor enchant to give the same TTL that mongoose does, it would have to provide something on the order of 1500-2000 armor for most druids.

 

thalys

35 Agi remains strictly inferior to mongoose in incoming damage profile. I admit that I considered precisely the same thing because for a while I was having a similar problem understanding what's going on. Simulate it instead of trying to visualize it. It's *not* meaningfully spikier, because the difference between 42% dodge and 40% dodge is the difference between being hit 30 times in 50 attacks, and being hit 29 times in 50 attacks. As kalbear referenced, until you reach high levels of dodge, where you're talking about something like the difference between 9 hits in 50 and 10 hits in 50, there's no change in spikiness. 29 hits in 50 for 101% damage each is strictly superior to 30 hits in 50 for 100% damage each. (And in reality, we're not talking about anything like 1% mitigation on an enchantment, but I'll pretend for the sake of your idea of an armor enchant to weapon).

What's spiky is your dodge rate--but that's not what healers heal, they heal damage. Your healers will never notice any 'spikiness' from mongoose procs. They'll notice spikiness from a very high dodge rate coupled with a Poisson distribution of hits, misses, and dodges (give or take parry haste)--but mongoose won't give you a sufficiently high dodge rate for them to notice this unless you already had one.

The straw man of 'well armor would be better', while true for a sufficient amount of armor, is not relevant as long as we're talking about what's currently available. 35 agi does not compare favorably to mongoose even on boss fights, and on multimob pulls it's not even close.

 

Drashian

Question on a similar subject here. From my understanding, it's accepted that in the case of multi-target tanking with swipe, Mongoose is vastly superior because it has the ability to proc from every individual hit rather than once per swipe. This gives it an uptime that I've heard estimated approaching 100%, meaning that its overall value is almost the entire value of its proc, etc etc. With the addition of kitty swipe, does the same apply to Berserking? In any pack with a decent number of targets, if Berserking procs the same way as Mongoose, it should have a significantly higher uptime than on a single target. Even if the lower rate of use of kitty swipe slows down the proc rate somewhat (since it's not spammable on GCD after the initial burst), it seems like Berserking is hands down the premier overall DPS enchant.

Anyone gotten a good impression of its uptime on swipe packs? Thoughts?

 

endymonium

Hi,

there's a new Shoulder Enchant available: Greater Inscription of the Gladiator - Spell - World of Warcraft.

Seems to be the best shoulder enchant now, as long as you are not "stamina capped"