Using Rawr with my gear and my stanadard Brutallus setting. I'm hitcapped (0.12% above) with Zhar'doom, and nearly capped (-0.19% below) with Tempest/Chronicle without Precision (or without ToW).

Comparison with all consumables allowed

2/47/11 - 2283 DPS with Zhar'doom with 2 pots, 1.74 gems, 0.06 Evocations. 2268 DPS with Temp/Chron (and far over cap) using 2 mana pots and 1.43 mana gems.

33/28/0 - 2215 DPS with Temp/Chro (0.19% below cap, that's a ~4 DPS loss compared to being capped), using no mana pots, no gems, no evocation and doing Fireball/Fire Blast for 181 seconds.

It uses pure fireball spam during cooldowns (135 seconds), fireball spam for another 43.64 seconds, and then fireball/fire blast rotations for 181.36 seconds.


Comparison without Destruction Potions and Flame Caps

Removing Destruction Potions and Flame Caps from the list of allowed buffs, it looks like this:
2/47/11 - 2255 DPS with Zhar'doom, 2235 DPS with Temp/Chro (and far beyond hit capped).
33/28/0 - 2169 DPS with Tempest/Cronicle.


Scaling Comparison

Scaling of 33/28 is as follows:
10 haste = 11.76 DPS, 10 crit = 9.45 DPS, 12 damage = 10.74 DPS, 10 int = 4.03 DPS, 10 spi = 0.69 DPS, 4 mp5 = 0.64 DPS.

For comparison, 2/47/11 with Zhar'doom scales as follows:
10 haste = 10.86 DPS, 10 crit = 8.17 DPS, 12 damage = 12.13 DPS, 10 int = 3.64 DPS, 10 spi = 0.00 DPS, 4 mp5 = 1.46 DPS.

The scaling with haste is a bit better because there is more mana to burn.
Crit scales better due to spell power, I expected crit to be a lot better though.
The scaling with spell damage is worse because Empowered Fireball is missing.



First impression on 33/28 from theorycraft

In terms of raw DPS, it's 4% behind fire/IV.

It deals less damage but has much more free mana to burn. This allows the use of more Destruction Potions and Flame Caps, but they only add slightly more damage.
33/28/0 gains 46 DPS from full pots/gems, while 2/47/11 gains 28-33 DPS from partial pot/gem use.

Accounting for the more liberal use of DPS pots/gems, it is still 3% behind fire/IV.

The difference shrinks to 1.7% if you remove the shadow priest.

Serpent-Coil Braid is shown as pretty bad because the spec already has "too much mana".


The spec probably isn't as bad as its reputation, but I don't really see it's use.
It's a slightly weaker fire spec with better AoE capabilities and I'd say comparable burst capabilities (AP + PoMPyro vs. IV/Comb + 15% better scaling with +damage from DPot/FCap/Trinket).
If AoE matters, you're probably better off going 40 Arcane, and if it doesn't, your normal spec (Fire/Fireball or Arcane/Arcane Blast) is the better choice.


Edit
// Going with full Sunwell gear, 33/28 overtakes 2/47/11.
// 33/28 - 2805 DPS with Illidan hat, T6 gloves, Chronicle, using no mana consumables, no Fire Blast either.

Edit: I forgot to remove Mind Mastery in the spec. Doh. Same situation in end sunwell gear, fixed the numbers.

33/28 - 2621 DPS with Illidan hat, T6 gloves, Chronicle, using no mana consumables, using 0.92 seconds of Fireball/Fire Blast rotation.
2/47/11 - 2730 DPS with T6 hat, Sunfire gloves, Heart of the Pit, using 1 mana pot, 2 mana gems, 0.04 Evocations.

(This models the M'uru trinket as 240 damage for 20s every 2 minutes. It is actually 320 damage for 15 seconds every 2 minutes, allowing for better cooldown stacking. It shouldn't make a major difference though.)

I am using the old Rawr though (13.1) because I didn't want to remake my itemlist (new/unknown loot and better gems). So, someone might want to double check things.

Major problem there. b14 had lots of fixes for mage calculations. You probably need to redo these comparisons in b14. You can use your b13 itemcache just fine, just copy it into the b14 folder. Same with characters, enchants, buffs, etc.

Erm, this is all fine and well, but i'd rather have you tell something about where specifically i made judgement mistakes, instead of generally dissing me.

So, if i understood correctly, the spec in question scales better with haste and crit than standart fire, but the loss in +dam scaling makes it lose out in the end where single-target dps is concerned. In return, it has stronger AE spam.

If this is the case, it might be a decent spec for MH raider who somehow didn't manage to get his hands on 2xt5. But that's about the only scenario i can think of

Since you ask...

At the start of your first post, you did exactly that: you dissed our beloved Rawr.mage without providing any specifics on why you don't trust it. I found your post ironic enough to feel compelled to respond. The judgement mistake was to use the word 'mathcraft' and then just list the talent differences between the two specs. Specifically: you didn't actually do any math, while saying you were a big fan of number crunching, but not trusting Rawr.mage which is by definition a number cruncher.

For example. Stuff like "So we can conclude Molten Fury and AP/PoM cancel each other out" is spurious. One is 20% for duration X, while the other is 30% for duration Y. Not forgetting one affects mana consumption as well. Depends on the fight, depends on gear, etc. These are things that Rawr can tell you, if you take the time to configure it.

Thank you for your feedback.

I apologise if any people that use and love rawr were offended. I assure you it was not my intention.
I recently read that the rawr version which i used back when i tried it out had mistakes in mage formulas.
Maybe the latest version will be able to provide me with more plausible results.

Since the new pieces off of Kil'jeaden have been datamined (the ones with warlock / firemage stats) I thought I would run a complete check of which of the 5 old t6 pieces to keep. My comparison was merely a stat -> spelldmg equivalence using a value of 12 spell damage for sockets (since no setup has more than 2 blues and yellow ~= red).

Hit comparisons done using next best option method, 5 hit = 6 dmg via gemming. (this also allows me to be neutral with 12dmg / socket comparisons)
Haste comparison using 1 dmg = 0.8 haste (end sunwell gear)
Crit comparison done using 1 dmg = 1.2 crit
It also uses int - crit conversion and adds that in.



Looks like t6 chest wins by a good margin unless I've made a glaring error. (the number on the right is all the stats converted to spelldmg)

Given this method isn't quite as accurate as simply making the entire gearset and figuring out the dps by swapping out the pieces, but this was just some quick work and I think the 10 dmg gap is enough to say which is better. If you use normal hit equivalency instead of next best option the chest just comes out further ahead. The only possible downside is that the chest setup has the least stamina. (obviously this is for non-tailors also since sunfire chest would add 22 spelldmg in equivalence putting t6 chest setup at the bottom)

With using Etherealz calculations and using t6 robe I would Say this:

Helm: [Dark Conjuror's Collar] With Chaotic Skyfire Diamond & Glowing Shadowsong Amethyst
Neck: [Amulet of Unfettered Magics]
Shoulders: [Amice of the Convoker] With Runed Crimson Spinel & Reckless Pyrestone
Cloak: [Tattered Cape of Antonidas] With Runed Crimson Spinel
Gloves: [Item not found!] With Reckless Pyrestone & Runed Crimson Spinel
Leggings: [Leggings of Calamity] With 2 Runed Crimson Spinel & Reckless Pyrestone
Ring 1: [Loop of Forged Power]
Ring 2: [Ring of Omnipotence]
Trinket 1: [The Skull of Gul'dan]
Trinket 2: [Shifting Naaru Sliver]
Weapon: [Sunflare]
Off-Hand: [Chronicle of Dark Secrets]
Wand: [Wand of the Demonsoul] With Reckless Pyrestone

As for tier6 pieces: Chest/Bracers/Belt/Boots

This gear would put you on 159 Spellhit so with using 1 Veiled Pyrestone in a yellow socket ( wand would be a nice idea ) you can reach the perfect hitcap without getting any raidhelp at all. I haven't done any calculations about this gear rly which tier6 piece would be the best to actually use but this gear seems to getting near perfect for TBC.

As for the keeping the meta bonus I would suggest:

[Robes of the Tempest] with 2 Reckless Pyrestones & 1 Glowing Shadowsong Amethyst
[Bracers of the Tempest] with 1 Runed Crimson Spinel
[Belt of the Tempest] with 1 Reckless Pyrestone
[Boots of the Tempest] with 1 Reckless Pyrestone

This will make sure you keep your meta.

I haven't included any professions with this gear.

To the guy making the comparison above between different t6 pieces and deciding which one wold be best to keep..

Did you include the sunfire robe?

Tailors are better off keeping the Sunfire robe and using T6 shoulders.

Helm: [Dark Conjuror's Collar] CSD + Amethyst
Neck: [Amulet of Unfettered Magics]
Shoulders: [Mantle of the Tempest] Veiled Pyrestone + Amethyst
Cloak: [Tattered Cape of Antonidas] Spinel
Chest: [Sunfire Robe] Spinel*3
Bracers: [Bracers of the Tempest] Spinel
Gloves: [Handguards of Defiled Worlds] Veiled Pyrestone + Spinel
Belt: [Belt of the Tempest] Veiled Pyrestone
Leggings: [Leggings of Calamity] Spinel*2 + Veiled Pyrestone
Boots: [Boots of the Tempest] Reckless Pyrestone
Ring 1: [Loop of Forged Power]
Ring 2: [Ring of Omnipotence]
Trinket 1: [The Skull of Gul'dan]
Trinket 2: [Shifting Naaru Sliver]
Weapon: [Sunflare]
Off-Hand: [Chronicle of Dark Secrets]
Wand: [Wand of the Demonsoul] Reckless Pyrestone

The comparison I made was NOT using sunfire chest, and i just noticed in my original post I said spellfire. Also, if you have an elemental shaman the shoulder setup is again superior due to hit cap.

My Rawr model gives me a different "Wishlist"

Helm: [Dark Conjuror's Collar] + Chaotic Skyfire Diamond & 6dmg/7stam
Neck: [Amulet of Unfettered Magics]
Shoulders: t6 +12dmg & 5 haste/6 dmg
Cloak: [Tattered Cape of Antonidas] +12dmg
Chest: [Sunfire Robe] +3x 12dmg
Bracers::t6 + 12 dmg
Gloves: [Item not found!] +12dmg & 5 haste/6 dmg
Belt: t6 + 5haste/6dmg
Leggings: [Leggings of Calamity] +2x 12dmg & 5 haste/6 dmg
Boots: t6 + 5haste/6 dmg
Ring 1: [Loop of Forged Power]
Ring 2: [Ring of Omnipotence]
Trinket 1: [The Skull of Gul'dan]
Trinket 2: [Shifting Naaru Sliver]
Weapon: [Sunflare]
Off-Hand: [Chronicle of Dark Secrets]
Wand: [Wand of the Demonsoul] 5 hit/6 dmg


I'm alliance and I've considered the 1% hit dranei buff.

You need one more blue gem for your meta.

Fortunately a very simple problem to amend.

Couldn't find a proper spreadsheet to work it out & I'm just bad at this game anyway...but yeah.. what works out better in terms of DPS and overall damage done, Battlemasters 40+ haste trinket or the Hex Shrunken Head.? Using the Battlemasters haste trinket puts me at 212+ haste with 1243 spell dmg. Also taking the /use 210dmg every 2 minutes for 20 seconds into consideration.

thanks in advance.

Hex in pretty much every scenario.

The only possible time the haste pvp trinket would surpass it would be if it would tip you to around 684 passive haste for static bugged- extra ignites.

So would that then mean the Mu'ru trinket isn't at all an upgrade from the Hex? :s

I do believe the m'uru trinket's "circle of power" buff can be shared by other party members, if not every caster in the raid (someone please verify this). I would think multiple trinkets used in a rotation would be better than HSH.

It's not a shared buff, but it is still better than HSH.

And in end game setups 1 dmg ~= 0.8 haste hence why dmg gems and haste gems are a wash in terms of dps in a sunwell setup with dmg being favored for dpm.

I sure hope not, that would be quite frustrating

But yeah theres some tidbits that tc misses on. I think the m'uru trinket is interesting, but at the same time I've found it to be somewhat frustrating to use, as is any 15s activations that aren't haste effects. Is it better to cast a fireblast at the tail-end of the 15s ? Or maybe with shifting naaru sliver + destro pot? Its already obvious that you want to refresh your scorch at the tail-end of the buff (or rather, time the activation of trinket so that it times well at the end with scorch refresh). All of this assumes of course that you can't fit an extra fireball at the end, but could fit a scorch.

But then all of this is further complicated by the fact that depending on the cooldowns used this might not be a good thing. I mean, if you have IV + shifting, is it really worth it to finish off the tail end of shifting with fireblast ? Instinctively I would say that it depends on your passive haste and the number of fireballs you can cast with IV affecting the cast, which may or may not make fireblast worthwhile.

I need to work out a model to properly evaluate it. I just hope it won't turn into a total mess for a really minor detail.

The buff 'limitless power' is not shared. However, I found an interesting 'feature' of the trinket in its current form. If you activate it, then step out of it, then come back in, you get another 15s stack. Problem is, when you run out of the circle the mechanics are kind of iffy and sometimes it stays on your for 5-10s longer and sometimes 2s. So its kind of not usable in practice, but still, maybe it can possibly be abused. Keep in mind, getting a new 15s limitless power buff will not change the fact that the power circle will die 15s after activating the trinket, which means your new limitless pwoer refresh will not last much longer.

However, what I don't know, maybe your limitless power doesn't expire as soon as your power circle dies, but rather, as soon as you step out of any* power circle. Let me word it in another way:

Maybe you could do the following.
Activate trinket. Use for a while. Step out of circle. Let 'limitless power' run out. Run back in your circle for 'limitless power' refresh of 15s. Then get another player activate his own trinket (over you - before your circle dies out). Maybe that will allow you to keep the 15s buff ?

No. It used to work like that, but it was a bug and quickly fixed in 2.4.2. The item would be ridiculously overpowered if everyone in the raid (or even the group) could get the buff and use it in a rotation.

I believe the m'uru trinket is also on a 90s timer now, although I have no source for that at the moment. If that is so it would be quite a bit more convenient to time with 3 min cooldowns especially on longer fights where you can't match them as well without ending up missing cooldowns.

It is 90s.