I've been subscribed to and reading this thread for weeks now. I'm still stumped on the Shred vs Rip idol as well. I swear I typically see more dps with the rip idol. Does not having a mangle bot greatly increase the value of the rip idol? I almost never have a mangle bot so obviously I have to sacrifice shreds to keep it up. I have 594 ArP before Grim Toll proc (no runestone yet).

@nightcrowler: I don't understand your "Dispersion" tables. Starting with the 1st, I assume that the 1st column (the % one) is used to indicate the % of samples which fall below the given DPS vaule. But then I don't understand how you obtained different values for the various stats. Is this obtained by modifying them by a fixed amount? How much?

Excuse me for dumb question.
I read whole thread and tested simulation but one thing is still is not clear for me.
Should my kitty be hit capped, or it is not necessary and i can simply go ar-pen in gemming?

Hitcap is not necessary for cats. Expertise cap isn't either.
Even below the Hit Cap, ArPen and Agi are better stats.

But, of course, you should have some hitrating. If you are running around with 8% miss, then some "stable" dps cycle is just not possible. But in practice, there's so much hit floating around on the items, this shouldn't be an issue.

I've got my own question:


Nightcrowler showed in his last update, that mongoose is the best weapon enchant for cats. I'm just skeptical here. Has anyone actually tried this? I thought 400AP > 120agi +2 % Haste.
Toskk and RAWR both show mongoose below berserking (while it's VERY close on toskk's dps simu)

You didn't read the whole thread as it has been mentioned that it's not necessary to be hit capped.

I'm sure there is a number somewhere that you should be at at minimum. At one point I was at 208 hit rating and when I ran sims swapping a +16 Agility/ArP gem for a +16 hit gem was a dps loss.

@Syntax53: About RIP Idol vs. Shred Idol: Actually the difference between RIP and Shred idol is so small (around 30 dps with arpen gear and a manglebot) that you can't see wich one is better by a simple combat log analysis. Without a manglebot my bet is for RIP idol.

@Helistar: The dispersion table as I wrote on the graph was done with a -143 stats difference from the base line (-286 for ap and feral ap)

@Mox512: If you look at both T7 and T8 BiS gear the main problem is not overcap hit, there is really too much expertise and hit gear out there. But if you aren't at hit/exp cap simply don't worry gemming for them. Hit/exp are better than haste/ap and barelly better than crit but arpen is still more valuable if you have enough of it (and also agility is better).

@Murna: I also was skeptical about mongoose. With T7 gear berserking was very slightly better, with t8 gear mongoose wins again. The difference is really minimal and probably due to PPM implementation (I used 1.2 for Berserking and 1.0 for Mongoose, what use other theorycrafter?) also the haste implementation could be different. Mongoose doesn't give 2% haste but 2% speed and 2% speed is far better than 2% haste, wich one is better depends a lot of raid buffs (Wind fury, Heroism) and your gear expecially because the difference is pretty small, as for other simulation I don't know if rawr takes speed increase into account correctly and for toskk's using my value into his simulation the difference in favour of berserking is 8dps.

Hi all,
Thx for this post it's very helpfull.

I have just one question about this :
With the trinket = 612 ArPen we have 1178 ArPen on proc, so it is the real cap to reach in raid ?

Yes, that was what i was saying. I came up with this number by calculation (however, since there are many factors involved this is only an estimate and may vary depending on the rest of your gear). Sunder and FF have nothing to do with it (although i assume they are always up but it wouldn't change much).

Also if you forego setboni in favor of more ArPen, the break even point will be higher (since you have to compare to normal gear w/ the setboni).

Mjolnir Runestone proc adds 655 ArPen, the tooltip is incorrect (Which was also explained a few posts earlier).

Why don't Sunder and FF make much difference? I thought the ultimate goal would be to hard cap ArP, effectively reducing target's armor by 100%. Don't FF and Sunder contribute towards that?

From WoW Forums:

Boss base armor is 10,643. Sunder and FFF are multiplicative, and together are 24%. They reduce boss armor to 8088.

Then ArP comes in. And some funkiness.

The amount of armor that penetration affects on a boss is given by (armor + 15232.5) / 3, which in this case with sunder + FF is 7773. Your static 30% ArP results in 2332 armor ignore, leaving the boss with 5756 armor. When Grim Toll procs, it goes up to 5985 ignore, leaving 2103.

Damage reduction formula in this particular situation is armor / (armor + 15232.5). So the boss has 27.42% mitigation against your physical damage, and 12.13% when Grim Toll procs.

Armor penetration stacks with itself - the more you have, the better it is.
Armor penetration rating now caps at 100%, or 1231 rating. Because it stacks multiplicatively with sunder and FF, and because of the cap on the amount of armor that ArP can affect, the lowest you can possibly get a boss's armor now is 315, or 2% mitigation.

I was wondering, can you repost the BiS list for non-hard mode Ulduar? My guild's pretty far from downing some hard mode stuff, so I want to know what to keep an eye out for when it comes to easy mode stuff.

Ok, i've been looking at something, and it seems that blizz has been a bit lazy and used the same mechanics from the OOC talent for the T8 2pc. Set. I did some testing but i need a lot more numbers (i havn't been able to really go thoroughly since i did 5min just now and figured i'd ask help from the community) but it seems that because OOC is a PPM mechanic therefore haste lowers the proc chance (balanced out by more hits), you actually LOSE (theoretically, havn't finished testing this) clearcasting procs from the 2pc from having haste on your gear. I guess the question is, is this loss of procs (ideally a free shred) lesser or greater than the increase in white damage?

I'm normally not a numbers person and my computer can't handle having anything else other than my mod heavy WoW running, so i can't really show much for this. Just figured i'd toss this up to the people w/ more resources.

Omen of clarity isn't normalized to attack speed in the first place - it has a flat 3.5/60 = 5.83% chance to proc on each white attack.

According to Wowhead spell data, the 2-piece Nightsong set bonus has a 2% chance to proc.

It does not work that way.

OoC: 3.5/60 % chance to proc on autoattack. It's not a PPM mechanic anymore. Given that, the more Haste you have, the more procs occurs in a fixed time interval.

2T8: If you look up it on wowhead you can see that it does not use the same mechanic as OoC. The only similarity is that OoC and 2T8 both apply "Clearcasting" when they proc.

Actually Omen of Clarity is a fixed proc chance 5.83% on each white attack, if you stack haste you will do more white attack in a unit of time and so omen of clarity will proc more. Also I'm not certain that PPM wasn't changed in WOTLK (Like they changed ooc, it was a PPM before) looking at many reports about berserk/mongoose uptime it seems that they changed them to works in the same way as OOC does. If someone has some heavy haste set and want to test them he will be welcome, actually my maximum haste is 10% and it's too low to test 2T8 or weapon enchant (I've runned something like 2 hours test on dummies but the error is still too high).

What you should do is go on a dummy, load you max haste set (you need a good 20% haste at least if you don't want to stay 24 hours on a dummy) remove omen of clarity from your spec, put on 2t8 and a berserking enchant (we know it's 1.2PPM) and just start autoattacking. A good addon to record procs is "Proculas". Or you can do a wws report.

2T8 is a fixed 2% chance on every dot to proc. Dots spacing don't increase/decrease with haste or any other stats/talent so the only think changing it is your bleed uptimes.

@Cal, I loaded up Rawr, threw on every piece of ArP gear I could find, (keeping t8 helm/gloves for the ArP they offer and the bonus though I do not know at what % ArP the actual stat outweighs the bonuses) and was only able to attain 918 ARP, fully gemmed with Blacksmithing and JC. Someone else please feel free to check this, but at the moment I don't think the 85% mark would be attainable currently.

I do have a question, because I am generally terrible at math. What % haste would result in an overall increase of in OoC procs in a certain amount of time? Like Nightcrowler said OoC is capped at 5.83% on white attacks, however in another thread someone mentioned that 1% haste was equal to 1% more OoC procs. To me that does not seem correct as it would only lower the swing timer to .99.

I guess my question could be asked like this. If in a 6 minute fight you proc OoC 21 times, (.0583x360) what % haste would it take to increase those procs to say, 25?

@Nightcrowler. Maybe I don't understand your intention, but if you were testing what I mentioned above, why would you drop OoC out of your spec? If you were testing the amount of haste required to increase OoC procs in a given time it seems to me that having OoC in the spec would be important, as the 2pt8 is a given 2%.

Nightcrowler, someone has already asked this question in this thread but I haven't seen an answer yet. Why do you use [Effulgent Skyflare Diamond] meta gem in your tanking gear choice?

I'm not sure if he's in his regular gear set or not, but I keep a helm with that gem in it for specific fights. I used it on Sarth3d quite a bit. Haven't found much use for it yet in Ulduar, but I haven't really looked for one yet. Maybe I'll use it when I start doing more hardmodes.

Ooc is not "capped" it is a flat 5.83% chance per white attack, so when you do more attacks with haste you will get more procs in the same given time period.

If you reread what he is saying, he is no longer discussing Ooc proc because he already said it is a flat %. removing Ooc from the spec is to use haste to test if 2T8 and/or wep enchants are ppm or flat % same as Ooc.

I have both mjolonir runestone and grim toll, and and 11%passive ArP. When both trinkets proc, about 12% of ArP is lost due to the 100% cap. I'm wondering though if low passive ArP and using these 2 trinkets is not a viable alternative to higher passive ArP and DCG + Mjolonir runestone.


Second point of debate is this MMO-Champion BlueTracker - Feral dps supposed to be this high? 8-9k dps?

There's a pile of factors that have been repeated obsesivlly for which the simulation and the actual deal don't really match. In my opinion the most relevant proof that there should be no real worry about cat dps being to high is this WoW Meter Online - Combatlog Replay in Engrish . By watching other statistics as well you can clearly tell that the general cat dps for the current bosses, is let's say, mid-squad compared to the other dps classes, so hopefuly there will be no actual echo to the post that I first linked.

I'm sorry but you've misunderstood my question. I am not talking about nightcrowler's gear, I am talking about the tank leather gear rankings he has posted on the first page of this thread (the links to wowhead). All helms there contain the -2% spell damage meta gems.

I see that meta gem as situational since most boss damage is physical. But it is only my opinion - maybe nightcrowler has some data based on which that meta gem seems a better all-round choice?

Wowhead puts the gem in there, not night. Wowhead is probably only paying attention to the +32 stam and ignoring the second part of each of the gems.