I'm sure you'll see a hotfix for that once blizzard finds out.

I've given up trying to determine what's intended behavior and what's not. And Blizzard isn't very punctual about fixing anything unless it has a dramatic, immediate impact on gameplay. (I'm looking at you, Savory Deviate Delight!) It took them years before they zone restricted the resistance buffs you could get from mind controlling orcs in UBRS, for example.

This remind me the age old debacle of 'whats intended, whats not', then I point out 'what was the supposed ouro strategy that blizzard had envisioned?' It's highly debatable that the Y positioning destroyed the intent of the fight (everyone was meant to eat the sand blast?!).

My current profile lists 1197 damage with AI self buffed, add to that Flask of supreme power (+70), food buff (+23), wizard oil (+42), improved divine spirit (~+20). I think I may have also swapped out some gear for pure +dmg instead of stamina on the wand/necklace/rings so maybe another +30 there.

Throw in WoA totem (+101), you get a total of 1483 +damage.

Add to this misery and CoS for a +5% and +10% modifiers respectively.

I didn't actually reach a 1k base AM with that setup, it was more in the 920ish range, but throw in spellstrike procs (which have a fairly consistent uptime with AM) and it was a bit closer. With icon and spellstrike active I would probably be fairly close if not over the 1k mark before arcane power.

Keeping in mind of course, that this is my alt mage, who isn't fully enchanted/gemmed for spell damage, and doesn't have a single piece of raid loot. I have attended a handful of raids with him when we had a surplus of healers, but for the most part I do not raid with that character and most probably would not receive loot even if I did. I am sure that if I had access to T5 equivalent raid loot and really pushed for it, I would be able to break a 1k base AM. With T6 gear and gems, it would be cake. Your crit rate would be total crap, but arcane mages scale best with spell damage and intellect anyways.

Disclaimer - my main used to be this mage, which is why I haven't let his gear go totally to crap. I also have quite a bit of pre-TBC raiding experience on my mage, but switched to my druid when the expansion was released.

My experience is quite different. It started that way, but the fact that AM has such synergy with on crit effects pushed the balance quite a bit. Add in the Relentless meta gem and I'm already at a point where 1 damage equals about 1.1-1.2 crit rating.

Partially because, at one point, this was intended behavior, and a CM said as much.

Over maybe overestimated the coefficient of fireball with the nerf? Emp Fireball only gives +5% coefficient now.

Well, my old spreadsheet is here:
http://www.sigilguild.net/hamlet/Mage3.xls

I don't generally post it often, due to it being incredibly user-unfriendly. I believe it's very accurate, though.

Try as I might, I can't justify speccing out of Fire. The highest-damage Arcane spec is the 43+/10+ that Manly tried (compared in the posted spreadsheet against 10/48/3). But even with 2/5 Tirisfal, the idealized spreadsheet DPS of the deep Arcane doesn't match up against the Fireball (and everyone here knows that Fire behaves more ideally in practice than Arcane does). The mana usage of ABx3-AM-Sc is 1.5 times that of Fireball spam (before JoW, but still).

So, admittedly my perusal of this thread has been cursory, but is there a real contention that deep Arcane specs can be made to actually outperform Fire using any setup?

EDIT: Aside from the Mage sentiment (which I share) that we wanted the arrival of AB to mean we did something besides spam Fireball in TBC.

I used 500 int, which is sort of what I have fully raid buffed, as I use spellfire and not T4, which at the end gives more spell damage even with arcane mind anyway.
If I put somewhere between 650-700 int in my spreadsheet, the builds break about even. But I really don't see myself getting 650-700 int anytime soon.

Remember reducing your spell damage to up your int (which you're probably not doing, but just saying) beats the whole point as I counted the 40/18/3 spec to have the same starting spell damage as 33/28, so you'll have to have close to 700 int after deducting the difference in base spell damage X4 from your int. As in if getting your int up to 700 costed you 20 spell damage, it's as if you only got your int up to 620, which again means 33/28 comes on top. But if you have 700+ int without gimping spell damage (or after factoring the lost spell damage) then I agree 40/18/3 does come on top. I'm just not at those level of int, and don't see myself getting there with the gear I will have in the near future, although it's not completely impossible once I'm decked in SSC/TK gear. But I'll definitely not even be at 600 int when I get my 2/5 T5, which means 33/28.

Note: that is all using the exact same rotation for both builds and gives the same results both with a 0-latency rotation and a some-latency rotation (as I find it extremely difficult but not impossible to do ABX3-fireballX2-scorch).

EDIT: I looked at 42/19/0 and the DPS of AM simply isn't worth it even after you manage to replace each hit rating with 0.8 spell damage until you have at least 6% hit left. You still end up being better off using fireball, so might as well toss those extra points in fire - either 33/28 with lower int or with extremely high int 40/18/3 and either way use AB-fireball rotations. Not having to keep scorch up has a hefty price I'm not willing to pay in terms of DPS loss. It's really not that hard to keep it up, 10s into the fight and the effect of having to put scorch up becomes almost neglicible (as it doesn't do extremely lower DPS than fireball and you can time it so you don't cast it much at all). And besides I compared the AB-fireball to 10/48/3 fireball spamming, so while you lose slight DPS to keep scorch up, so does 33/28. I could calculate it more accurately but that's a place I really don't want to go into as it's starting to mess with fractions of %s of DPS... I find going into accuracy of anywhere more than 0.5% difference compeltely pointless, and even moreso when it affects both rather equally.

EDIT2: I looked at what kind of int I would be getting at best after getting decked in SSC/TK gear and I will still be well under 600 int fully buffed, making 33/28 still the best spec.

Did you not factor in arcane mind that increases total intellect by 15%? I still use full spellfire and have 573 with only AI - 650 something when raid buffed. Add to that, you are a gnome, how could you possibly get a set together with arcane mind and not get 600 int raid buffed?

You're right I forgot to add in arcane mind, which with a non-missiles build is possible to get 5/5 of (although not possible with a missiles build), which with FULL SCC/TK gear would still have a hard time bringing me to the 650-700 int spot. And I multiplied arcane mind with the value I got after factoring kings which actually gives me a few more int points than I'll really have. Remeber 650-700 is the break-even point, which means you need MORE than that for 40/18/3 to really beat 33/28. And not only I'm not in SSC/TK gear, I'm still waiting for our guild to recruit some more people who aren't total nubs and show up for raids so we can actually kill stuff successfully in SSC and get a 2nd piece of T5.

Again even if 700+ int with arcane mind is achieveable (without gimping spell damage), it's not something I will have anytime soon. But it's good to know there may actually be a point where arcane mind + mind mastery actually become useful, but then again with all that gear I will also have to recalculate the whole DPS thing as your INT isn't the only stat that is changing. Up your damage/crit as well and suddenly the % increase talents also increase in power, not just mind mastery, so it's just a requirement to re-compare specs and DPS cycles as your gear gets dramatically upgraded.

For example if I add myself 4% crit and 200 spell damage (just to see what it does), the break-even point between 33/28 and 40/18/3 increases to 750-780 int. I'm not gonna make a different line in my spreadsheet for every possible item combination though, I'll just update stats as my gear upgrades and make choices according to that ;p which for now means 33/28 once i get another piece of T5.

The real reason I don't optimize my gear around the spec when comparing, is that even if I take the inferior spec, I see if it can do better with different gear of equal quality - and it never does. While the ratios between stats change a little between different specs, they stay pretty close to eachother. One time 1 crit rating will be worth 0.72 damage, another time it'll be worth 0.83, and those are huge extremes in terms of gear and spec changes. The only time you'll actually change an item becuase you changed your spec is if the items were already incredibly close in quality to eachother, in which case the difference is MUCH MUCH smaller than the difference in DPS between the 2 specs.

So basically if you optimize your gear for each spec and THEN compare, if spec X was >2% DPS increase compared to spec Y with the best gear for spec X, if you compare it to spec Y with the best gear for spec Y, you're most likely to *still* have spec X doing more DPS. Maybe the difference will go down by 0.5-1% but even that I highly highly doubt. The itemization options in the game really aren't big enough to warrant those huge differences.

Of course I assume you don't do silly stuff like frost damage gear on a fire spec etc, and anyway I already took into account swapping out my pure fire damage offhand to the all-damage offhand when comparing the specs, as it's my only fire-only item ATM (after finally replacing my "of fiery wrath" bracers with the new pvp bracers, as before those there was simply no upgrade until al'ar).

Also note that when comparing DPS you don't really look at the fight much. Different fights basically change how much you move, how much time you have to burn mana out of how long of a total fight duration and what range and AOEing abilities are required. However moving fucks every spec over in a pretty similar way, although arcane does get screwed a little more if you have to move at the exact same time as your 1.5s AB needs to be casted, but I doubt it's a major issue. Gonna have to play around with it though, as having to often do 2.5s cast ABs will gimp your DPS to the point of being roughly equal to 10/48/3 but consume more mana, in which case 10/48/3 would be better for mana efficiency, DB/BW and range.

Obviously the changes in gear can be quite substantial if I have 850 int raid buffed with gear up to Mag while you can't see more than 700 even with full SSC/TK gear. I think your assumption that gear doesn't significantly change performance is not well founded.

It is more that most of the mages are tired of spamming just fireball/frostbolt throughout a fight usually and are looking for arcane to be viable dps in terms of matching coming close to a deep fire/frost dps. For arcane to work it need not outdps deep fire but provide dps close to that of a deep fire spec as currently theorycraft shows deep frost to be. The extra goodies that come with slightly reduced dps as frost/arcane more than compensate for it.

You're wearing pvp gear so I can't really see, but you can look at my armory and let me know which pieces I could swap for stuff that would better suit arcane. I just can't find items that give noticeably more spell damage for mind mastery specced mages than non-mind mastery specced.

I've done advanced search sorting items by "effective" spell damage on allakhazam based on 1 int = 0.22 dmg through crit and 0.25 dmg through mind mastery for total of 0.47. 1 crit = 0.8 dmg and 1 hit = 1 dmg.

Head: Same stuff I would use with any other spec, no item stands out when you increase the value of int.

Neck: The one from prince is really the only option in the game until you get very far in raiding content.

Cloak: Same.

Chest: I'll still use the one from vashj, and spellfire until then. Not even tirisfal beats
spellfire in DPS even with mind mastery if you don't get 4-piece bonus out of it.

Bracers: Same

Gloves: No difference, spellfire all the way to T6.

Waist: Belt of blasting, and spellfire until then... Or cord of screaming terrors if you're willing to spend DKP on an item when its upgrade is already available at the same level or raiding. Nothing different than normal item choices though.

Legs: Tirisfal>7th circle>spellstrike for any spec.

Feet: Same as well...

I won't keep going into rings/weapons/trinkets you'll probably get the same results, and even if you don't it's not really a big difference.

To sum it up if you have over 700 intelect and are not decked in T6 you're probably using subpar gear which just happened to have a lot of int on it (such as T4 and lots of 5-man blues that simply suck but have a lot of int).


Regarding utility, I would seriously reconsider. Fire has more range and better efficiency and better AOE (unless you have loads of mana for AOEing and only need to AOE once every 3 minutes...). Blastweave and dragon's breath are extremely high DPS for their 1.5s GCD and extremely mana efficient for AOE spells. Arcane doesn't have that AND arcane is limited to 30 yards range. So I would only spec arcane if it would give noticeably MORE damage, which would only be the case with 2/5 T5.

Once again, we're not trying to convince fire mages to spec arcane. We're debating how to optimize arcane specs via gear/spell cycles.

As I said I have gear up to Mag and for example on a 10 min enrage fight with access to shadow priest and totem of wrath I use the following gear:

Ogre Slayer's Cover (+20 penetration)
Anchorite's Robes (+6 all stats/+8 int/+8 int/+4 int, 2 mp5)
Boots of Foretelling (vitality)
Evoker's Mark of the Redemption (+12 spell damage)
Violet Signet of the Archmage (+12 spell damage)
Soul-Eater's Handwraps (+20 spell damage/+4 int, 2 mp5/+6 spell damage, 5 crit)
Collar of the Aldor (+22 spell damage, +14 hit/+6 spell damage, 4 spi/Relentless)
Trial-FireTrousers (+25 spell damage, +15 sta/+8 int/+8 int/+8 int)
Torc of the Sethekk Prophet
Pauldrons of the Aldor (+18 spell damage, +10 crit/+8 int/+9 spell damage)
Icon of the Silver Crescent
The Lightning Capacitor
Gladiator's War Staff (sunfire)
Girdle of Ruination (+12 spell damage/+8 int)
Eredar Wand of Obliteration
Veteran's Silk Cuffs (+15 spell damage/+8 int)

Obviously I'm missing some upgrades, just had bad luck for some slots. I consider this setup to be quite optimized for full arcane build. Raid buffed this would net 846 int, 313 spi, 59 mp5, 1127 spell damage, 6.77% hit, 31.58% crit.

Once again you're only comparing dps. You fail to realize that more mana means more damage.

Calculate how much mana you use in a fight. Then calculate how much mana you get from 100 intelect and see that it simply doesn't make the difference compared to the HUGE amounts of spell damage you're losing. What's 1500 mana when you use 30000 or something in that order of magnitude compared to increasing the effect of that 30000 by 10%+ by using spellfire and spellstrike? And that's in short fights, in longer fights you'll use way more than 30000 mana making the 1500 even less useful compared to the spell damage. Granted spirit keeps working as the fight moves on, but it's still tiny amounts of mana compared to what you already get (from mana spring/tide totem, BoW, JoW, gems, pots, and most importantly VT).

I guess you didn't notice that I'm not tailor. I agree that spellfire is superior, but I'm not interested in dropping my professions.

I wasn't a tailor either, it didn't stop me though. If you don't want to do your best why are you even reading these forums? Just about any decent raiding guild I know has just about any top caster there have tailoring. It makes a much much bigger difference than your little arcane vs fire dillema.

First of all I don't appreciate your tone in this post. I don't have any dillema, I'm perfectly fine as arcane. I'm here on this forums because I get useful information and I think I contribute useful comments. I'm sorry if I don't meet some "hardcore" requirements for posting here.

Well, you still should be re-gemming your gear for +damage gems instead of +int.

Total mana really isn't that important. Mana usage versus mana gained is. Neither of which depends on your initial mana.