Actually, looking back at this week muru, I did evocate. I wonder how I forgot about this. M'uru is really the extreme case of dire mana situation given spellsteal. It drains you more OOM than normal AOE fight because there isn't any downtimes in the fight.
Manly - WWS

But yeah, for the most part, you don't adapt your gear based on DPM issues, you adapt based on consumables, since destro pot/flame cap have relatively very poor ROI (return on investment) compared to mana potions. Sure, a destro pot might give you 50 dps under the most extreme of situations, but is 50 dps really all that much when going 2700->2750 ? I never put much emphasis on flamecaps/destruction potions because their return is relatively poor. Its not something that will make or break a fight. Going OOM at 1-2% of the boss hp isn't something I'm willing to risk. Or worse, try and adapt ahead of time and not dps pre-execute range. In other words, I only burn dps consumables if I know for sure I will end up with excessive mana, otherwise I simply don't.

In the muru example above, its a worst-case scenario. I have almost nothing going to help DPM at all, and I still make it through. This is quite flatly put the worst possible mana hungry scenario I have ever came across in TBC. Arguably there was also shahraz pre-nerf (the first nerf*) but that was a side effect from a long fight and multiple mana burns. Given that, I doubt I can imagine someone consciously changing gear around to drop some haste in order to get more dmg/crit. I just don't see why you would ever do such a thing. As a corollary, if I were 0/40/21, I would have ran into mana issues. But as 2/48/11 I've yet to come across a case where that has been an issue.

Well, not evocating on any fight longer than 5 minutes is simply not an option for me. 1300-1400dps on a shadowpriest is quite unheard of for our guild. Not to mention we run with an elemental shaman instead of resto. At times where I get an innervate I sometimes make it through if I pop mana pots/gems instead of flame caps/destro pots.

Another question, I was looking over the WWS parse you linked. Is KJ fight that bad for a mage? You guys, as mages, were quite competitive over all the fights save KJ. I never read about the fight, so I have no idea how things are like.

Not too sure what to tell you about mana. I've never really ran oom. I could see that being a problem on brutallus and felmyst. If you don't spellsteal on m'uru you should be fine.

KJ is just not a mage fight. Everything is like precisely made to specifically not work no matter what spec you chose. You could probably do decent dps if you were to only nuke KJ and ignore the rest, but at that point you're pretty much useless to the raid if you do that. If you kill the orbs/adds, then you can't do a better job than hunters, and significantly worse than warlocks. That leaves mages in a bind with nothing much interesting to offer.

As for the other fights, well, its kind of debatable. I had really crappy numbers this week due to 1- lack of coe 2- death on twins and felmyst. Sometimes you just get bad days. Usually for new-ish raids getting their brut kills mages do fairly well, then once everyone starts to get full sunwell gear you just can't compete with consistent 2.7-2.8-2.9k dps from rogues/hunters. You might get lucky once in a while I suppose. Another major problem is that threat is an issue once you get more than 1 bloodlust as a fire mage. It kinda depends on your tank too. We're gearing up new tanks and it gets me somewhat in a bind. You might get 1 more bloodlust but that extra bloodlust almost garantees you to cost you a full 5s of dps time lost to invis. Its still a net win, but its really hard to capitalise upon. Last week I got 2 bloodlust, but threat I couldnt stack cooldowns due to threat.

For felmyst, I just expect mages to top meters. If you don't, you did something wrong.
For twins, its a bit hard to say. You do have a hard threat cap to deal with. Usually I can't invis either, and even if I can invis I still have to hold back for a while. I wouldn't expect spectacular numbers on twins as a result of being unable to invis.
For m'uru, it depends a lot on your job. If you spellsteal, you should get top 1 or top 2. If you aoe too (and have the mana for it), then definately first spot. If you're on sheep duty, don't expect a miracle.

If you want a better evaluation of what to expect, I think last week was more fair. For starters, coe, but mostly, not dying on half the fights. Wow Web Stats I would selectively ignore brutallus as far normal numbers go, that's with 2 lust and lucky crit rate. Otherwise everything else is about on line with normal numbers.

I have also raided arcane for a while, but that's not especially relevant. Nobody's personal experience or sense of things is determinative. Do you use Rawr? Do you doubt the mage module's accuracy? I am very confident in its assessment. Try the following setup:
standard 40/0/21 spec
[Cowl of Tirisfal] with [Chaotic Skyfire Diamond], [Brilliant Lionseye], [Glyph of Power Word: Barrier]
[Brooch of Nature's Mercy]
[Mantle of Tirisfal] with [Brilliant Lionseye], [Sparkling Empyrean Sapphire], [Greater Inscription of the Orb]
[Tattered Cape of Antonidas] with [Runed Crimson Spinel]
[Robes of Ghostly Hatred] with 2x [Runed Crimson Spinel], [Reckless Pyrestone], 6 all stats
[Bracers of the Tempest] with [Runed Crimson Spinel], Spellpower
[Gloves of the Tempest] with [Reckless Pyrestone], Major Spellpower
[Belt of the Tempest] with [Brilliant Lionseye]
[Leggings of Calamity] with 2x [Runed Crimson Spinel], [Brilliant Lionseye], [Runic Spellthread]
[Boots of the Tempest] with [Brilliant Lionseye]
[Ring of Omnipotence]
[Band of the Eternal Sage]
[Hex Shrunken Head]
[Serpent-Coil Braid]
[Sunflare] with 40 spell damage
[Heart of the Pit]
[Wand of the Demonsoul] with [Brilliant Lionseye]
buffs: Improved Divine Spirit, Improved Mark of the Wild, Blessing of Kings, Inspiring Presence, Arcane Intellect, Wrath of Air Totem, Mana Spring Totem, Totem of Wrath, Improved Blessing of Wisdom, Adept's Elixir, Elixir of Draenic Wisdom, Brilliant Wizard Oil, 23 spell damage food, Curse of Shadow, Curse of the Elements, Misery
fight parameters: 0.05 latency, 360 sec duration, 0.95 dps time, 250 shadow priest m/5, heroism
(In short, Sunwell level gear for an arcane spec, with a typical raid setup)
Feel free to exchange some items up or down Rawr's list and consider variations on the raid setup and fight parameters.

For this exact setup or a range of similar possibilities, what does Rawr tell you about the relative values of [Potent Pyrestone], [Reckless Pyrestone], and [Brilliant Lionseye]? For the exact setup I listed, it rates them at 9.30, 9.13, and 8.84, respectively. Change one of the Lionseyes to a Pyrestone, and the relative value of a Lionseye will rise. Change a couple [Reckless Pyrestone]s to [Potent Pyrestone]s, and Reckless will edge back on top.

Pintofbrew, you have a bad habit here of talking down to people and making coarse statements like "your assumptions are all wrong". Sometimes people do show up with old questions that have been answered or muddled approaches to situations, and you are correct in responding, if not always in the style I would use. Sometimes, however, consider that I or others may indeed know very well what we're talking about. Your blanket declaration "At no point is Crit a competitive stat. In any way." is wrong, unless Rawr is way off here, and I have no reason to believe the latter.

(One might expect you especially to be more careful about such dismissive declarations. Most mages a couple patches ago would have mocked the idea of ever socketing intellect and spirit gems. Many still would. Now we have the heretofore unexpected result of crit rating becoming competitive in some setups. Don't be one of those unthinkingly dismissive people.)

As for the appropriate language to describe this phenomenon, I believe I was perfectly clear in explaining what the other questioner might have meant about crit rating increasing in relative value. On the preferred idiom of calling this "increasing crit's relative value" or some other terms, I can only ask that you reread my previous post and think about how "increase the value of __X__" could mean by reference to some quantity other than the one you assume.

This sprang up in response to somebody asking about mana and damage tradeoffs for fire spec, anyway, not arcane, and Roywyn's post that I linked above, along with the same sort of experimentation in Rawr, should sufficiently address his questions.

I can't imagine that your SPs are THAT bad that you can't make it through a 6 minute fight without evocation. Last week I received 2 heroisms, used 2 destro pots and 2 flame cap and used no mana consumables at all and I was fine on mana (was also using an elemental shaman). Considering in a 6 minute fight you can use 3 mana potions and 3 mana gems - an additional 15k mana - I can't imagine your SPs giving 15k less mana than mine. Ours usually pull 1300-1550 DPS and generate between 21-26k mana over the course of the fight.

Truthfully, the only time I find myself evocating at this point is when I don't have a SP (sometimes on Kalecgos) or on AOE heavy fights (like Felmyst), and even then, I only evocate so I can just do more AOE dmg, but it's not going to make or break the fight if I just let the warlocks do it and tap afterwards.

Pintofbrew... How do you remember all that?! That's nuts.

Anyway... Thanks for the extended reply. I was doing most of my "thinking" while preoccupied at work, so I obviously forgot some basic info when I made my assumptions.

I'm still over the frost hit cap now, but I intend to knock it down a bit for some extra spell damage.

I agree (now) that I shouldn't directly compare another mage to myself if his gear is just similar. To make it work, everything has to be the same. It did make me wonder, though... but there are too many variables. I really should know better... I sort of throw all my scientific knowledge of experimental design out the window when I play WoW. I prefer to just "pew pew."

I still think the main point of my message was misunderstood. I know that DPM is a useless characteristic. If a fight is longer than 5 minutes, it is a "survival fight" and the usage of mana will not be as intense due to movement.

I think Manly answered me correctly, simply by stating that the benefits of a Flame Cap or Destruction Potion are vastly outweighed by the presence of mana at the end of fights. However, the hunters and rogues in my guild, apparently cannot push 2.6-2.8k as others can. Our WWS parses show myself and another mage with the rogue and hunter around 2.2k for Brutallus kills.

My real question was whether or not dropping haste on certain fights would be beneficial. M'uru P2 comes to mind, where either a ton of haste or none would be most beneficial. However being that my guild has not reached P2 yet, I cannot personally comment on what cooldowns will be up at a certain point in the fight, but once we get there it will be pretty obvious which should be used. There will be no time for evocation during Entropius so mana must be at a sufficient level by the time he comes out.

Once again, Manly hit the topic I was trying to express. The benefit of a mana potion vastly outweighs a destruction potion.

2/48/11 is thought to be the best fire build. But what is the reason for the 2 points in arcane tree. Why not place them in the fire or frost tree. If you are only taking 2 points in arcane subtlety and you rarely use a arcane spell what is the gain?
Why not have a 0/50/11 or 0/48/13. I am trying hard to justify the 2 points in arcane subtlety and just seems better spent some where else.
How often does spell penetration really effect spells in PVE encounters.

It has nothing to do with spell penetration and everything to do with less threat on Arcane Explosion.

So why waste the points in arcane tree then?

Why not place them in a talent that may net you a better return.

Because it's generally thought there isn't one. AE is the best aoe spell to spam as a fire mage, as flamestrike is balanced around having the dot tick, and blastwave / dragon's breath are on cooldowns. Theres a lot of encounters needing aoe, and even with the reduced threat it's still easy to get aggro if your going all out (lots of haste gear, exalted BT trinket etc).

It would seem to me that the remaining options (or lack thereof) are the reasoning behind the points in Arcane Subtlety. Even though you say that you were not pulling off your Paladin tank without points it in, the fact still remains that you are less likely to pull off of any tank with a 40% reduction in threat. Staying alive longer, obviously, gives you more of an opportunity to deal damage and, from what I can see, none of the other remaining talent options raise your damage output/survivability enough to replace Arcane Subtlety with.

Any of the Arcane talents
Any of the frost talents (except possibly Improved Blizzard for those fighting Illidan)
Impact - Stun that does zero dps
Improved Fire Blast - Ignite bug makes this less viable
Improve Flame Strike - Used less than Arcane Explosion
Molten Shield - Marginal damage increase and a non factor in raiding
Blazing Speed - Zero damage and a non factor in raiding

Agreed. Most of the other options are helpful in PVP but not in raiding. I'd say if you have one point in imp blizzard already, maybe you should put two more there if you can't pull mobs off the paladin with AE, otherwise think about what you do that isn't raiding and put the two points into something that helps there.

I ran with blazing speed for a while and it was mildly nice in non-raid environments, but I finally just decided to have a PVP/solo spec and a raiding spec and cough up to respec as needed. These days I spend most of my time NOT in raid spec, but shift to 2/47/12 for a few hours on saturday when raiding. My other spec works fine in 5mans, arena, battlegrounds and solo-world stuff and has advantages over the raid spec. But its DPS is 13% lower on fights that last more than 20 seconds.

If you don't need -40% threat to aoe, you're probably not using icy veins to aoe. Or combustion.

But despise that, even if you don't have a 'practical' use to it, -40% threat allows you to start much much earlier your aoe.

Theres always a gain to be had from -threat.

Combust only works with fire, but if you lead with dragon breath/blast wave before AE spam it's relevant as it will likely fully proc with just those two spells. I guess the goal here is to use up whatever threat lead the tank has before going to low threat AOE spam, as this trick will be quite high threat.

Icy Veins...quite honestly it never occured to me to use this when AOE'ing trash. I usually don't blow all cooldowns on trash, although there really isn't any reason aside from threat cap to not do that as long as a boss fight isn't about to happen.

Now I want to try this out, but it looks like my next raid will be with a roleplay guild rather than a raiding guild, and I'm not sure I want to experiment with higher threat AOE spam until I have a sense of how solid their AOE tanking is.

Combustion -> IV -> Trinket -> What-have-you + Flamestrike-Blastwave will net you quite a bit of threat. When you go back to spamming AE, you'll be thankful for the -40% threat. Unless you are just not trying when it comes to AE, in which case you might go 0/47/14 for imp blizzard to help out on Illidan or something.

My normal start out to trash pulls in hyjal is.. scroth twice - Combustion -trinket - IV - flamestrike - blast wave- AE till mobs are dead. I never get any aggro.
This I was putting this question out there. Just seems like with a good pally tank. The points could be place somewhere else.

As people have said many times now, yes you could put the point somewhere else, but anywhere else you put those points is equally a waste in PvE. I wouldn't want to do Felmyst without the talent.

With six tics of consecrate+paladin passive stuff to build threat I'm not surprised you don't pull threat. I have a tendency to open up on AOE about 3 seconds earlier with my paladins. I'm not saying it's a better practice to do so, I really lack the experience to say but the two scorches != AOE, they're similar to doing nothing for 3 sec on all the secondary mobs.

The last two ticks of 5k+ ignite...Is it because of the self stacking ignite mechanic (or bug) ? Note, this is with 130ish haste +IV.

4034 crit
5420 crit
1307 ignite
5349 crit
3892 crit
5082 crit
5256 ignite
5256 ignite

total damage: 23777
expected ignite: 9510
actual ignite: 11819

Not much to see here. First, you see combustion fading after 2 crits, which means theres a previous crit that you're not showing up. And also, I know that there is a trailing ignite from before the start since the first ignite tick damage doesn't match the numbers. This is the telling lead that allows us to understand what happened.

Let's find the previous trailing ignite...
x = previous trailing ignite
x + (40% of 4034) = 1307 * 2
x = 1000
previous crit (not shown in screenshot) => 2500.

So now fixing the sequence, and re-ordering them in the order they happened.

2500 crit
4034 crit
1307 ignite
5420 crit -- this one caused a rolling ignite, which means the above tick (1307) was free damage.
5349 crit
3892 crit
5082 crit
5256 ignite
5256 ignite

So now fixing the calculations taking into account the above:

total damage: 26277
expected ignite: 10511 (counting free ignite => 10511 + 1307 = 11818)
actual ignite: 11819

If you had read my thread about the bug, it should have been pretty obvious.

Yes, that log is not complete and doesn't show 1 previous crit (although I don't remember what it was, probably 2500, partial resist, like you found out). Manly, could you please link your post about this bug?