1) 33.28.0 has been proven to scale worse than full fire, and to top that off does not benefit from elemental precision. There currently is some debate whether AB/Sc rotations are viable but I have yet to see anyone proposing serious viability from such a setup. The major problem with it would be spellhit: Half your spells needing more than 200hit, and half needing only 64. This causes extremely bad synergy from gear: Hit rate benefits half the rotation massively, and spelldmg benefits the other half massively. A sloppy build for sure. As for 33.28.0 fireball spam, as I said, it's simply not up to 10.48.3's standards.

2) Unless you're running Mind Mastery, int is by and large an irrelevant stat; Even with gnome 41 arcane with spellfire, it -still- is debateable whether int gives enough of a benefit. Point of proof: top-tier mages -still- take 12dmg gems over any intbuff. Stacking crit in gear is accepted as the Worst thing you can do in terms of flat damage, and crit gained from int is incredibly bad. While not constant, for the purposes of practicality you can think of int as roughly +70int = 1%crit. So 3 int is about same as 1 crit rate, and 1 crit rate is about the same as 0.65 spell damage. This is why int is ignored.

3) Vontre's sheet is open to everyone. Park your gear's stats in and do the maths. Don't expect people to turn up, as though it's their job, take your requirements and run your numbers for you. EJ is not a renta-math service.

Edit: Never EVER sign your post, we are not blind and your name is coveniently glued next to your post. This is not a letter to your gas company.

It was proven given certain amount of spell dmg way back before 2.1 raiding, however now we have the coefficient change undone and getting 203 hit rating really isn't hard given BT/Hyjal gear. If it has indeed been proven with the coefficient nerf undone I would like to see a link to said tests and conclusions.

No, this is not true. Int for arcane specced mages in 2.3 will be even better than it is now, I remember seeing Kavan's math on direct impact of int on his performance and it is indeed not as bad as you make it out to be. Top Tier mages aren't by default smarter than everyone else, it's generally agreed on that 12 spell dmg is the best choise, I don't think it's an exact science though as alot of things factor in (meta gem requirements, socket bonuses, hit rating and probably something else that I've forgot) Stacking 12 spell dmg gems could easily be attributed to the fact that MSD requirements made it more or less impossible to use yellow gems.

Are you talking about the old scorch spam or actually spamming fireball with 33/28?

10-48-3 (10xfire, 1xScorch) : 1862dps

33/28 (10xfire, 1xScorch, AP): 1707(.91667) + 2415(.08333) = 1766dps

33/28 scorch spam : noth worth mentioning since the AP scorch spam is only 1640dps.

note - 10x fireball is possible with the amount of haste I have on the spreadsheet.

33/28 is considered for a 3 minute time period. It is also using Pom/pyro during the AP.

That's garbage. Stacking spell damage gems is done becuase that gives the highest dps return, regardless of metagem requirements or spec. The MSD requirement just gave further incentive to ignore socket bonuses and go straight spell damage.

While not worthless, Intellect has been a poor stat for mages. Arcane spec makes better use of it making it merely a sub-optimal stat. Other then moving evocate to be based of Intellect, what 2.3 change makes intellect better? Intellect has a horrible return for dps, the only real benefit of it is longevity. Since your starting mana pool is not overly important to your longevity, with Shadow priests, JoW, and Mana pots all providing better returns on longevity (spirit does to, and even moreso for arcane mages come 2.3).

Stacking intellect does not seem to be a worthwhile effort for any mage.

Why take my text out of context. If you need the hit the Veiled Pyrestone will be > Runed Crimson Spinel in damage returns, no?

My point wasn't that stacking 12's was SOLELY based on it's damage returns, it was that looking at so called Top Tier mages and deriving that 12 spell damage is always the best could easily be because of the MSD requirements. There are quite a few of socket boni that would be worth taking if it wasn't for meta requirements.

This is a question I've had for a while, actually...as it would take pretty subtantial modeling to figure out. At what point does the Scorch ramp-up not become worth it? Is it always worth it, or are there time-value points where straight Fireball/Fireblast is better? (Or perhaps less than a full Scorch stack... or Scorch/Fireblast, etc. Number of Mages/stacking time may be a factor here too?)

Given all the numbers to consider here, I'm having a difficult time napkin mathing this one.

With the changes coming to MSD and the long overdue unnerf to fireball and frostbolt how are people seeing the benefit if any for those T5 mages who have 2 piece arcane and have been tearing up the meters with am AB/AM/AM/AB spam.

My hunch would be that AB spam is still king on trash, but 10/48/3 and a 6FB:1scorch rotation would provide better returns than working in AB into your rotation. Has anyone done any testing here?

We can break down smaller point to smaller point and lose sight of the overall conversation, which provides no information to anyone who is reading this thread and is generally unproductive. Or we can look at the bigger picture.

The question was asked about the value of stacking Intellect, and Pint answered that the value of doing so is low. He used the example of top mages socketing for pure spell damage as an example of his point. Again, the main point being that gearing with a focus of getting as much Intellect as possible is a bad idea.

You refuted him with several points. Mainly being that, Int is getting better in 2.3 and that just becuase some people do it, it doesn't mean that its the best option and their could be other resons for it.

Now, it's true, that the metagem requirements of the MSD (the defacto best metagem pre 2.3) discourage using yellow gems. But thats not the only reason top mages gem for pure spell damage. Pure spell damage red gems are the gems that give the best dps returns. The only caveat to that is in the case of spell hit for mages below the cap. None of that changes the fact that Intellect is a suboptimal stat for mages with regards to their dps. So really, PintofBrew's point had the implicit statement - "Good mages don't gem for Intellect". Since the primary dps increasing affect of Intellect is to increase spell critical chance, and gemming for spell crit is something that is done at a last resort (note, your using spell hit gems rather then spell crit gems in your example), Pint's original point of 'stacking Intellect (and spell crit) is not a recipe for optimal dps' is still something that we should agree with.

Quibbling over semantics and ignoring the larger point are both confusing and detrimental to discussion.

In reference to your latest post, you again focus on a smaller point without paying notice to the larger picture. As such, your point is confusing rather then enlightening. The whole picture is something like this:

Spell hit is one of the most important stats, however, since high end gear has so much of it and their is a hard cap, you rarely see it socketted for.
Assuming hit is not needed and ignoring socket bonuses, pure damage gems priovide the best dps return. However, there are often cases where becuase of the socket bonus, a mage could get high dps benefit from socketting a orange gem.
The MSD gem requirement discouraged socketting yellow or orange gems, becuase each one would require a corresponding purple or blue gem which would lead to a dps loss often despite socket bonuses.

That's pretty much the whole picture, is it not? Everything there says that it is in the mages best interest to focus on spell damage in situations where spell hit is not needed, unless they can get significantly more spell crit rating then damage (not the case with gems). None of that refutes PintofBrews statement that stacking intellect and crit' is bad for dps, nor does it make the spell damage focused gem choices of high end mages a bad example of what is a good dps choice.

Vhad: While I'll concede that perhaps my evidence of gem choice is a poor example of relative quality of int over spelldmg, I will stand by my guns and maintain that the return from Int isn't good enough to vouch stacking int over spell damage, irrespective of spec.

I will accept that given spellfire set (netting a total of 32% int -> damage) and lowbie gear, perhaps indeed int is good enough. Perhaps if we're talking pre-kara and kara gear it's a decent choice, also because at that level not all raiders will have a shadow-priest to feed their arcane-induced mana habit.

But for SSC+ raiding? I have yet to see conclusive evidence pointing to any validation that int is useful. Indeed, once you lose the spellfire bonus it becomes clearly inferior as far as I can see.

Edit:

While i was typing this Pheroz answered in a much more coherent response than I did. See above for more coherence and better quality.

Apparently, very few people are aware that [Runed Crimson Spinel] wasn't meant to be 12 dmg. If you follow the ilvl formulas, they took the blue gems and upgraded them from there. As such, runed crimson spinel was intended to be 11.5 dmg, and since rounding is done on even numbers, it was rounded to 12. Also, dual color epic gems are typically rounded down (as far as caster gems are concerned).

This alone to me is enough to not consider using anything else than +12 whereas possible, given the dps return irregardless of spec.

Another small detail. I do not believe any socket bonus is worth taking on hyjal/BT gear. The sole exception is [Cowl of the Illidari High Lord]. The problem is that they seem to follow the general formula that they give +2 damage per on-color gem. If there are 3 sockets, then you get +5 damage socket bonus (even worse). This means, given the epic gem values, that any 3 piece socket bonus will never be worth taking (until they begin to put red sockets). So this leaves only 1 and 2 socket. Assuming there are no red socket, which there very very rarely is anyway, this means any time you have 1 blue socket you just annihilated the socket bonus. The cost of taking 1 blue gem in your gear is simply too high to justify the rule of thumb +2 dmg/socket they put everywhere.

With this said, if they had yellow/yellow +4dmg it would be worth serious consideration. But there is no such thing currently. Its always blue +2, or yellow/blue +4, or yellow/yellow/blue +5. Then of course, if they put red sockets that would allow red/blue +4 to be workable and probably worth it. But again, that's the sole possible exceptions.

Yes, it is Vontre's spreadsheet modified for for the coefficient change + a few items from ZA that prove to be better dmg than BT/Hyjal gear.

And i actually just realized that I used the same gear for both calculations...Here is the calculation with 204 hit and a little less dmg.

33/28 (10xfire, 1xScorch, AP): 1733(.91667) + 2453.02(.08333) = 1793dps

10/48/3 gear

Cowl of the Illidari High Lord
Hellfire-Encased Pendant
Mantle of the Tempest
Cloak of the Illidari Council
Robes of the Tempest
Bracers of Nimble Thought
Tempest of Chaos
Gloves of the Tempest
Belt of Blasting
Leggings of the Tempest
Slippers of the Seacaller
Ring of Ancient Knowledge
Ring of Ancient Knowledge
The Skull of Gul'dan
Hex Shrunken Head
Chronicle of Dark Secrets
Carved Witch Doctor's Stick


33/28/0 gear (added 2 6dmg/5hit gems as well)

Cowl of the Illidari High Lord
Translucent Spellthread Necklace
Mantle of the Tempest
Cloak of the Illidari Council
Robes of the Tempest
Bracers of Nimble Thought
Tempest of Chaos
Gloves of the Tempest
Belt of Blasting
Leggings of the Tempest
Slippers of the Seacaller
Ring of Ancient Knowledge
Ring of Ancient Knowledge
The Skull of Gul'dan
Hex Shrunken Head
Chronicle of Dark Secrets
Wand of the Forgotten Star

Ah well, thanks for the answer. I guess it's 10 48 3 or don't raid in 2.3.

The rounding bonus shouldn't be a factor in whether or not you want to use runed gems. The reason one should pick runed gems is that the gem has a comparative advantage in contributing to DPS when compared to the alternatives.

Well, its pretty obvious that if I could pick between 11 dmg and 6dmg/5crit, I would always go for 11 dmg, given a realistic high end spec and gear. 12 dmg just really pushes it over the edge.

Anyway, to me its not even a matter or whipping out the spreadsheet before every gem. Its just common sense. You will reach hit cap without hit gems. Your next best choice is all reds. Reds it is.

(not to speak that hit gem have no effect on sub-73 mobs, which does matter for bosses with adds)

I was under the impression frost does competitive damage in addition to providing excellent utility next patch.

I made my wishlist with WoWEquip, getting 4 set t6 with the Kaz'rogal legs + all the other best pieces. Getting CSD once the patch hits and fitting in 2 blue gems to meet its requirements I get:

Naj'entus boots and Tempest mantle: 2 spell dmg vs 5 crit rating. 10 crit rating > 4 spell dmg unless I'm entirely mistaken?

T6 gloves, 4 spell dmg vs 5 crit rating, this one I'm not entirely sure about but inclined to go for the crit.

[Cowl of the Illidari High Lord] + [Leggings of the Tempest]

EDIT: I'd probably go for pyrestone on t6 gloves. Mostly because crimson spinels are very contested. I expect slightly better dps from pyrestone anyway.

I'm still not sold on the CSD over the nerfed MSD. All of my calculations put them within a couple of dps apart depending on the gear setup. The only difference I can really see right now is CSD will return more mana from critting more often and the MSD will spend more mana since it is around 33 haste rating by my calculations (that's calculated at a 15% proc rate and 90 passive haste). However the MSD should do more damage on fights where there are periods that you are moving around since it benefits more from less casts during the cooldown.


And yes, putting a 6dmg/5crit gem in the tempest gloves does give slightly higher dps than a 12dmg gem. And by slightly I mean .3dps by the spreadsheet =)

Care to share that wishlist please? I would like to see it. Maybe I am blind and or just dumb but I am not comming up to the hit cap for a fire spec.

Item stat value with the old MSD (MSD doesn't affect stat values, but CSD makes crit and intellect ~6% better, if we go from 210% to 216.3% crits. Need some CSD info from PTR.):

http://elitistjerks.com/506021-post2099.html

4 dmg and 5 crit are pretty much the same. Crit should be ahead by 1% for the CSD.

With a blue gem in T6 leggings and 2*red in a belt, you get +32 dmg.
With a red gem in T6 leggings and orange/purple in a belt, you get +28 dmg, +5 crit.
Same as above, 4 dmg or 5 crit.

T6 hat + channeled pants should be a tiny bit better than T6 pants + illidari cowl.


We're talking about 0.05 DPS differences for the gem choices, mind you.

Edit:
Gear setups of choice in the bolded link.

Edit:
4 dmg vs. 5 crit is a wash. Cowl/Leggings is whatever you get first. Hell, I could live with 5/5 T6.
The bigger deal still is MSD vs. CSD, how CSD, and if we whined enough to reduce the MSD cooldown.

Well, I guess that if the difference goes down to 0.05dps we can call it a wash and go for whichever drops first. I do liken the idea that cowl/t6 pants already provide the 2 blue gems for CSD, in addition to have a better look (and not looking the exact same as s3 helm).