Goedel

Re which 2 pieces of t5 to choose: I found in a lot of time playing with Rawr that the helm + shoulders, not gloves + shoulders, were the best choice to keep the bonus with full Sunwell gear. The differences among all the gear slots' t5 vs. Sunwell margins are small enough that this answer may vary depending on what range of items you assume are available and what raid setup you assume. (Interestingly, helm + shoulders worked out to be the low-tier keepers in many cases, including 2t4 for feral druid dps.) In any case, you won't be at a significant disadvantage with any 2 pieces. While going through the process of gearing up, of course, having more options makes it much easier to swap. I would definitely grab the gloves if you can.

As for the helm from Void Reaver, Zinaida's suggestion holds true for both that and the superior Hood of Hexing. For arcane, they have a lot of wasted item budget on hit rating you won't be able to use, while lacking a valuable meta socket.

Re 10 int gems and WWS: My guild has been farming Hyjal/BT for a few months while remaining slow getting through Sunwell, with a lot of roster turnover, to give a general sense of the gear level here. It seems I haven't been in for the start of Hyjal in a while, and just Azgalor + Archimonde doesn't show much, but here are a few recent reports with what I hope is a representative sample: Hyjal + Illidan, Kalecgos & start of BT, BT Gurtogg - Illidan. I actually only have 10 int gems in my t5 helm + shoulders and [Cuffs of Devastation], with other options in those slots for elemental specs. Pieces that would be used for both specs I've gemmed with 5 haste + 6 dmg. So this may not be as pure an example as you're seeking, but it is one example (there being many more out there) of the spec performing well on a variety of fights, long and short. It's not too hard to reach the point where [Brilliant Lionseye] and [Reckless Pyrestone] are evenly balanced and after socketing one Rawr suggests the other as the optimum next gem.

edit: Whoops, this is posted as from a druid. The mage in question in the reports is Bohr.

 

madgazza

On flask selection, i generally go for Blinding light when i have a good support group, and supreme power when i know i may be frostbolting a bit.
With minimum group support, ie on Brutallus where i may not even get an spriest, and will probably end up using Blessed Weapon Coating etc to keep my mana a bit better, is Flask of Distilled Wisdom from the olde world actually the best? OF course it gives almost 1K mana, crit, dmg, and regen. Anyone done the maths?

 

Light4

I am considering now to stay with T5 chest and shoulders for the following reasons:

1) I have no T5 helmet
2) Legs are bad for the obvious and aforementioned reasons (vs. Legs of Channeled Elements)
3) We have at least 1 guy on our Realm who can craft the BOE Sunwell gloves which are way better than T5
4) We are still far from downing Illidan, ergo no T6

 

Wizeowel

I do tank at Council, but actually I shouldn't because apart from Conc aura and IV there is no pushback protection for me. The one week a fire mage tanked he did so much more damage than me, and didn't have to gimp his gear much to get the extra +spellhit.

I think it's very difficult to argue the case for stacking intellect just using spoken language logic. You list all the things you will miss out on if you didn't stack intellect, but they are kind of given in this thread. The difficulty in making an analysis which works as a general rule is that every situation is different. You may have lots of spirit on your gear and be assured of imp divine spirit, and since you mention innervate, yes those will all push the comparative value of intellect higher than for another player who won't get these. You could easily show me Brutallus raid parameters in Rawr where a bit of extra int is better than a bit of extra haste, just as I could easily show you raid parameters where the reverse is true. It still remains true that the longer a fight goes on, the less comparative worth there is for intellect to the other stats. Just load up Rawr to check this. Look at gems comparative values for a raid buffed fight of 240 seconds compared to 360 seconds compared to 480 seconds. You'll see quite a wide variation on int gems while other gems remain fairly equal to each other.

 

Light4

http://elitistjerks.com/797416-post2089.html

here you go

Of course, this _is_ equip and race dependent. Try Rawr and compare the dps-value of the corresponding buffs.

 

Saizul

The "problem" with haste is it is a redundant stat.

The entire idea behind Arcane specs these days is to come as close as possible to maintaining 100% AB spam. Usually what this means is you are trying to spam a spell that has some of the highest DPS in the game, but is unsustainable over the long term. If, as arcane, you are often finding yourself in the position where you cannot burn mana fast enough...that is - you are AB spamming the entire fight, start to finish, including through AP / IV and bloodlust, and still ending up with leftover mana, then in that situation haste would be a clear DPS increase. However, I would guess this situation isn't all that common...usually what will happen is haste is just speeding up the time until you run low on mana and have to switch to filler spells instead on a lower DPS rotation. It isn't that haste is a bad stat, but it is sort of stepping on the toes of what AB already does for itself. Certainly if you could have 0 haste or 100 without trading other stats, the haste would increase your DPS. But it isn't nearly as powerful as straight +dmg, and I'm not sure if it's even better than crit if you aren't sustaining AB spam.

 

Zinaida

Yea, I neglected to mention that - helm + shoulders is certainly a viable option and can be better than gloves + shoulders. As you say, it simply depends on what you have available to you.

Blessed weapon coating/mana oil are never a good idea since using the normal oil and a higher proportion of frostbolt gives a better tradeoff.

 

Kataali

Thanks to all who have responded. I think I will try to pick up the T5 gloves to have that flexibility, and thanks for mentioning the T4 hat, since I happen to have that one piece of T4 and it's a great piece for Arcane compared to VR's cowl right now.

 

Zaylin

Just how crazy would a full Arcane build be? Something like this: Talent Calculator - World of Warcraft
No fire or frost filler spells, Mage Armor to help the mana. AB (x3), AM, AB (x2), AM, AB (x2), AM, ect.

 

Tizzlewump

To me, the biggest difficulty of arcane isn't that it is particularly hard to play (no class or spec really is), it's just that there is nothing obvious about the way it works. It is rife with weird technical tricks like how the debuff ramps up, how applying arcane power *increases* the dpm of your ramped AB, how 2t5 looks like it isn't a dpm gain for AB but it really is, and how your "spell rotation" is more reminiscent of a shadow priest's priority list than a hunter's shot rotation. The various ratings -- crit, hit, and haste -- all have nonobvious tradeoffs too. I wish I could remember which of the theorycrafters here opined, "It's not that int is a good stat for arcane, it's just that all the rating stats are so bad for it."

Saizul is right that haste is bad because it's redundant, but crit is also bad because you crit at 175% as arcane (think "reduces the critical strike rating on all your gear by 25%"). Spellpower is bad because you take a huge mana penalty for arcane's equivalent of Improved Whateverbolt, and int and spirit are bad because they're stats. But arcane is a math-intensive build and you have to be willing to use rationalization after-the-fact and not before it. I literally did not believe my eyes the first five times (!) Rawr told me [Cloak of Arcane Alacrity] > [Cloak of Subjugated Power]. Check upthread for a nice story of why that might be so, but understand that it was numerical analysis that gives the conclusion, and only fun storytelling to observe that "casting N+1 filler spells in a fight yields more aggregate damage than casting N of them." Clipping the gcd of fully ramped AB under stacked haste cooldowns is the biggest problem with haste, but that's something that is more related to IV and heroism than your passive haste rating (grab a guild shaman and check it out for yourself) -- but once again, don't take my word for it, figure out the math or find someone who did.

Improving your AB uptime will generally improve your dps, but focusing on "100% AB uptime" is a bit of a bugbear. I prefer to think of it is "maximizing AB usage while minimizing FrB usage." This is because AB is both your best dps (fully ramped) and your best dpm (any other debuff stack), and FrB only helps you to manage your debuff stack. In practice so few interesting fights really involve pure AB spam that it is misleading to say "any stat that doesn't help you sustain full AB spam is a waste."

Zaylin might be thinking of a 50/0/11 variant that I don't understand at all but Kavan apparently was talking about in an unrelated post. Honestly, any time I see Scorch and AM appear in his posts my eyes glaze over so I don't know what (or how) this build accomplishes what it does but I am certain that it does not involve chaining three unramped ABs or using fillers for any purpose other than unramping AB.

If you're not interested in actively managing your mana bar, running Magegraf/Rawr, and most of all reading a lot of old work on arcane builds that has already been done this is just really not the spec for you. I have no clue where the arcane pioneers came up with this stuff, but if you're not interested in at least reading the math that's been done, you're certainly not interested in recreating it yourself. You are not missing out on massive dps and fire is not an "easier build", you're just skipping a lot of numbers and technical stuff that some of us enjoy for its own sake.

 

Wizeowel

I know we're not supposed to put QFT posts, but that was really great, Tizzle.

About the whole haste thing. I think it's unfortunate that Alvira has put something so confusing in the OP of this thread. Instead of trying to look at how to make haste work for arcane, or promote some balanced view, he's said he personally prefers crit, "no matter what the theorycraft says. Why? Because haste just runs us out of mana that much faster." This is kinda the same thing Gediablo said in his posts yesterday. I think it's a shame that despite a whole thread devoted to arcane and us having one of the best theorycrafters working on one of the best modelling tools, that a lot of people still didn't get the point about how arcane really works. It seems people are assuming that you can or should expect to cast AB for 100% and ignore the very real fact that unless you can rely on getting 2 or more innervates there will eventually be a fight where you have to cast filler. It's like, perhaps at a stretch, rogues saying that since their white damage accounts for 70%, they might as well just spam SS to keep SnD up. (Yes, I know a lot of them do indeed do that.) Similar with arcane: anyone can spam AB and then spam FrB when low and it'll work fine, so there's not much motivation to look deeper except to prefer a resto shaman for mana tide and bother your druid tank for an innervate. You don't want to wear haste gear because then you can't Hero+IV+Skull like you can with fire, and rather than having to plan your cooldowns for optimal effect, just go for the easier option and ignore the haste gear for crit gear, or the haste gems for the int gems.

 

pugs355

I have just started using RAWR but am very unfamiliar with it. For me it puts the spell haste one in the lead but I am not sure of how to really use RAWR and all of its settings. I will have to try to find some kind of RAWR howto or something.

 

Alvira

Well, it was one of the discussion points I raised in the opening thread. And even Kavan said in this thread that if he had the choice, he would swap out the haste stats on his gear for other stats (he said it on page 2, post 45). The truth is that haste does run you out of mana that much faster, so unless you have lots of mana support to spare, its not as ideal a stat. Its great if you can still sustain near 100% AB spam with it though.

I don't really assume that everyone reading the thread has already combed through and absorbed the entire evolution of arcane since TBC started. I assume some are newbies to arcane raiding. Hence, my desire to start them off on the right footing. At this stage, rotations are not the right footing to start arcane raiding with. Trying to maximise arcane blast spammage is.

 

ceeth

Hi, first off, very good article and comments :)

Check my armory to get a guage on all of my current gear if you wish. My spec is 40/0/21.

My dilemma to all the arcane mages out there more experienced than I is this. I already have the new boots and ring from the new badge gear. I have almost saved up another 100 and I am having some troubles deciding which to get:[Shroud of the Lore`nial] or [Legwraps of Sweltering Flame]. Currently i am wearing [Legwraps of the Aldor] and [Vestments of the Aldor]. I see the tradeoffs with each decision. Hit and spirit increase with robes (allowing more flexibility in other slots and gem sockets) and a very nice damage increase with the legs, but my guild is currently in ssc so I may have a shot at the T5 legs soon. But there is the dkp issue also, which I am middle of the pack at best. I would rather get the pants and the increase in damage right now. Then again, the other side of me says wait for the T5 legs and just grab the chest right now.. however long it may be before I get T5.

What do you more experienced arcane mages think about this?

 

HellaBooya

Well first off, to be competitive as Arcane you really need two piece T5, but if you are content on staying arcane there is a few things that you could do to increase your dps. First off you are WAY over hit capped. The hit cap for Arc/Frost is 126 so if I were you I would replace your hit trinket with basically anything else that you have. The spirit is nice on the badge pieces that you linked, but personally I would get the crit chest/legs just because you don't need spell hit as Arc/Frost.

 

Gralin

One thing I don't see mentioned in this thread, possibly because it's a long-answered question, is whether it might be advantageous to wand instead of using filler spells, in order to speed mana regen and get back to spamming AB. Is this a horrible idea?

 

Wizeowel

Again, you are stating this as if it's somehow a bad thing which doesn't need any further explanation.

Say you have 0 haste and cast 170 arcane blasts and 30 frostbolts during Brutallus. Then next week you change an int gem for a haste/spelldamage gem. Now you do 169 arcane blasts because you had less int, but had time to do 32 frostbolts and the average damage of all those spells went up due to the extra 3.25 spelldamage. Do you see what I'm getting at here? There is a tradeoff, and the tradeoff is not only to do with how quickly you use up your mana, but you have to look at the whole picture.

The magic with arcane is that everything affects everything, and for optimal damage you have to work it all out. I don't care what my damage is at Illidan or Kalecgos. It only matters for Brutallus and so I'm gearing for that. If my passive haste means I can fit 3 regen cycles in while the meteor slash is on the other soaking group, then I'm gaining damage. And I expect that unless you are standing with the melee it would apply to everyone else as well. Or unless you're not doing cycles of course...

 

Zinaida

This section of the first post is in my opinion, very misleading. You say that crit is always better than haste regardless of theorycrafting results due to haste "running us out of mana faster" which is a very narrow minded view of the spec. Haste increases dps at the cost of higher mana consumption, and the net benefit is still considerable despite it reducing AB spam time.

As Kavan said, if I could trade all my haste for other stats(damage or int) then I would do so, as it is not an ideal stat. However saying that it is strictly inferior to crit is just plain wrong - acording to Rawr this is only the case once you have a very large amount of haste, i.e. heavy Sunwell gear.

 

ceeth

Thanks Hellabooya,

Someone gave me some wrong info regarding the hit cap for my spec.. I thought it was 164. Which I now know is for a fire spec with 3/3 EP. I am thinking about getting the IOSC now to replace my hit trinket/blue crit trinket replacement I have for that hit trinket.

 

Alvira

Hence I qualified it by saying that was strictly my opinion. It was a discussion point. And hence, up for debate. I am actually quite happy to see there is more discussion this point now. Once everyone can reach a more general consensus of how haste rating should be treated on gear, I will put it in. I do think there are a lot more issues to consider while adding haste, so its definitley subpar. Perhaps the more controversial issue is whether haste should be ranked below crit when looking at gear. For me, I would not swop out all my crit for haste if I could, but I would swop all my haste rating for crit (if forced to choose strictly between theses two stats).

So, by all means, let's discuss this, since it is a point for discussion.

I have edited that part to make it clear it is preliminary and valuing crit over haste is strictly just my personal opinion only. Right now it currently reads like this on the opening thread.

Once we have enough discussion on this, I will edit it further to put up something more conclusive.

 

Light4

It also depends on the procs you use most. If you have a lot of "proc-on-hit" like Spellstrike-2 haste might be more an advantage than crit, because you can increase proc uptime. If you have more "proc-on-crit" then swapping out crit for haste is probably not the right way. I used Rawr for my gear (T5-4 and TLC) and came out with haste ~ crit with TLC having the most impact on the item value. Once I rid myself of all "proc-on-crit"-stuff haste comes ahead of crit.

 

Zmaj

The frostbolt in your rotation is significant portion of your arcane dps. It's called filler because arcane blast is far more damage, but its far from useless. If you find yourself not being able maintain your mana pool, consider re-gemming your gear or using a different rotation.

 

Wizeowel

I'll need to say it again, the first two are false assumptions; the third misleading.

Haste causes you to run out of mana faster because you are casting faster. It doesn't in itself mean that you can cast less arcane blasts, you could still be doing the same total damage, just quicker. The number of ABs cast has more to do with the total length of the fight in terms of I5SR regen, consumables, VT and JoW and not at all with how quickly you are casting them.

With regards to haste messing up AB cycles, if you can't watch your gcd timer due to fight mechanics, passive haste still isn't a problem. It is activated abilities which affect haste which can throw your timing out: drums, skull, IV, heroism, etc. Quite honestly though, if you spam your button you should be okay. This is more of a learn-to-play issue than any kind of real drawback.

Lastly: yes you are right you don't want your passive haste + drums to put your AB under 1.0 seconds. Preferably you'd also rather that heroism gives the full effect while casting AB with drums. But seriously what kind of gear will do that? Even full Sunwell gear for arcane probably only has little more than 300 haste. Furthermore, there are many ways to make full use of IV, skull and heroism even if you have high passive haste. Whether you figure out a solution with AB or you end up hasting your regen cycles, it doesn't matter. You just need to be more flexible, rather than stacking all your cooldowns on 1 macro and hoping for the best. Rawr is excellent at providing suggestions for how to stack your cooldowns, so have a look at it.

 

Tizzlewump

Everyone intuitively believes that wanding is suboptimal, but since I just went on a rant about "run the numbers before you write a story" let's see if I can put my spreadsheet where my mouth is. First, a reminder about where the tradeoff coefficient comes from. ("The Kavan Tradeoff Coefficient"? Gnomes and algebra, a winning combination.)

I have some spreadsheet data for an almost arbitrarily chosen 1400 spell damage and 32% base crit (0 haste, it didn't start out as a spreadsheet). Let's grab 2/2 Wand Spec and a [Carved Witch Doctor's Stick] and take the 219.6 dps at face value. Ignore shadow priests' and warlocks' debuffs since they'd only yell at us about ISB charges. Thus

wand_dps = 219.6
wand_mps = 0
frb_dps = 1126
frb_mps = 113
ab_dps = 1786
ab_mps = 448

where "frb" refers to using frostbolt spam as a filler. This is pretty well-known to be suboptimal, but we can give wanding the benefit of the doubt since we all know who's going to win this fight. These numbers place FrB spam at 1.97 dps sacrificed for each 1 mps gained by dropping from AB spam, but puts wanding at a horrific 3.21 dps sacrificed for each 1 mps gained.

Looking back at this brief work, I realize just how badly I butchered the computations. My "mps" estimates are atrocious. The wost offender is personal spirit regen, which should modify FrB and AB mps by 60% of your regen while giving you full 100% regen while wanding. When comparing spell cycles you can neglect things like Misery and CoE+CoS but not when wanding enters the picture. Fortunately correcting these errors only makes wanding a worse tradeoff, so there is no need to work harder. JoW should appear there somewhere too but doesn't. Further critique of my math would be instructive, but I've hidden so much of it from you that I don't think anyone can do much better than a laundry list of stuff I "might have forgotten" -- PMs are nevertheless appreciated.

So now I have some numbers to compare. They don't really pass the garbage-in-garbage-out test, but this competition is lopsided enough that I'm still willing to stand by the result. Now I get to make up my story about why wanding is bad.

Story #1 (good): "Gralin, the reason wanding is bad is that the dps is so awful when compared to both ramped AB spam and bad filler spam that the extra mana you gain will never have a chance to catch you up to someone casting correct filler."
Story #2 (bad): "Gralin, the reason wanding is bad is that since you already have 60% passive regen you only get 40% of your spirit regen in return for not casting and instead shooting your wand."

These may sound similar and even be equivalent in the scheme of arcane dps, but it would be very irresponsible of me to offer Story #2 as an explanation. My awful calculations solidly support Story #1: I didn't really compute how bad the dps tradeoff is, just that it's really bad. Since I completely skipped the issue of spirit regen in my math, offering you Story #2 would be deceptive, even though it's a valid point, so I'd better keep that one to myself.

While this may not be the most rewarding analysis done in this thread, it's still a really nice example of how thorough our TC framework is.

 

Wizeowel

You forgot to calculate how much dps your shadowpriests and warlocks would lose by you wanding... most wands are shadow damage so they eat ISB debuffs.