a little off the actual topic question :

is it true that spell penetration still does nothing in pve ?

some mages in my guild did lots of testing and tell they prefer 13% hit and 60 spell penetration rather than 16% hit, but I'm not sure it was made the right way, and I believe it as already been tested here, but I can't find where except the post in the thread :

Working Theories of Theorycrafting as of 2.1

that says :
Spellcasting
- Spell penetration does nothing in PVE once a Warlock curse is applied. At all. Ever. It is a good PVP stat and nothing more

Sigh. For about the million-th time (and it's been mentioned at least 10 times in this thread).

Spell Penetration does nothing in PvE except on Supremus for fire mages. On every other 25 player raid encounter, it is completely worthless in all respects.

The working theorycraft thread is kept updated, and it is there for a reason.

First you don't have to respec fire to make a lot more dmg, you can and it will help a bit, but it's problematically with your frozen shadowweave and it's not the reason why your dmg is low actually.

About Maulgar kill:
An average frostbolt dmg of 1,38k is ... well ... ok for lvl 60 but for lvl 70 even with much worse gear you should do more dmg unbuffed against solo mobs.
I can only assume that you had to downrank often wich shouldn't happen.
Get the ogrila mana pots and flask and chain chug the pots, they are extremly cheap.
I'm not sure if this wws is complete, but i don't see even 1 curse of elements (or shadows) on the mobs, but sometimes curse of agony, fix that, it will up your raid dmg a lot!

Gruul kill:
Better frost average dmg (COE!) but still way to low, also it seems you got 2 battle rez in this fight - stop dying it's bad for dps.
At gruul fire will always out dmg frost a lot because you can't really use your iceblock and winters chill sometimes falls of at ground slam (icelance when flying and directly after shatter to prevent that).

You should aim 1,7k -1,8k average frostbolt dmg (that is non crit dmg and includes trinkets and procs), get all consumables and some gear improvements and it shouldn't be a problem.

Gear:
-Resocket helm with spell dmg meta from alleria hold, chain chug ogrila or unstable mana pots and gems and don't tell me that you have a mana problem as deep frost.
-Put best available sockets into frozen shadow. you will wear that until WotLK.
-Resocket the sta/int and spellcrit gems first.
-Get Bracers of Havok - Items - World of Warcraft or pvp, they are cheap and enchant them with spell dmg.
-Get rep with aldor for better shoulder enchant
-Enchant your Chest and Boots!
-Farm Spellstrike pants and echnant them!
-Get icon trinket for badged
-Get a weapon and take a 1h + offhand if possible.

The Felhunters on Azgalor trash have +all resist, and Illidari Council has +all resist.

Rage Winterchill also has very high Frost resist.

As far as I know though, spell penetration is pretty useless for a frost mage due to the binary nature of frostbolt.

Not true. Let's take an example. For the sake of simplicity spell hit is not factored in and all spells are considered to be hits.

Level 70 caster is firing fireballs against a target with 350 fire resistance, which is enough to reach the 75% resist rate cap. Any individual fireball might get resisted in steps of 25%, so an individual fireball may do 0%, 25%, 50%, 75% or 100% of the damage, but the average damage resisted will remain at 75%.

Level 70 caster is firing frostbolts against a target with 350 frost resistance. 75% resist rate simply means that 75% of frostbolts are fully resisted while 25% do full damage. In a short fight the resist rate might seem random, but it will average to 75% of total damage being resisted.

When you stack up on spell penetration the average resist rate will decrease and thus your overall damage increases. While the effect might be more noticeable on non-binary spells, spell penetration is equally beneficial for binary spells as well.

If I were to make a rough guess, I would say he has 200 FR. Typically I get sub 40% partials with no COE (but wearing 50-ish spell penetration (blaming shoulders)). With COE, that number goes down to sub 20%-ish.

As far as supremus goes for fire spec, just take advantage of your range. An example of a few parses, to show you a bit what fire and arcane can do. Stion/Ishaxa is fire, I'm arcane in all those parses I believe.

Loading...
fire spec: 1037 dps
arcane spec: 1052 dps

Loading...
arcane spec: 1115 dps

Loading...
fire spec: 992 dps
arcane spec: 1023 dps

As you can see, even with 20% partials, its not like fire can't compete.

[Elixir of Empowerment] is also nice for Supremus, but I've never used one, and comparatively it's kind of costly.

Anyone tested AM + BT trinket after patch? Does it work properly?

Yes, and its amazing, it has very high uptime.

I dont know if this has been discussed yet but I have not found any info on this site.

They say that in patch 2.3: "Obliterated. Damage coefficient reduction removed from Improved Fireball/Frostbolt"

Im not in to theorycrafting but I have read most of the posts on this forums about mages and still havent found any information about there beeing a tax that inflicts impovered fireball.

My questions are:

1)Can someone explaion how the tax works?
2)how will the removal of the tax change our damage, any calculations up yet?
3)Some people say that this will boost frost specs more then fire specs, some people also say that arc/fire and arc/frost will outbeat deep frost/fire speccs by far after this is implemented. Any calculations or thoughts about this?

While not stated on the talent sheet, each rank of improved frost bolt and improved fire ball causes the spells to recieve 2% less of your +dmg bonus. Technically this means if you're fire and have 1000 bonus damage, your fireball will receive 900 of that only, and if you're frost a mite less because frost (i) has 3sec cast time as opposed to 3.5, with respect to +dmg calculation and (ii) has an additional snare effect which causes it to lose 4% spell dmg as well.

Long story short, look at your +dmg, divide by ten. ROUGHLY that much is going to be the gain in fireball damage and a bit less in frostbolt damage.

I ran some Lhivera's theorycraft today, for classic full frost pve it represents a gain int he region of 45-55 dps and for fire 55-65 dps. It should also be roughly the same for arcane frost, but I didn't bother with arcane fire, it's not a particularly valid spec imho unless you're talking about 33.28.0 scorch hybrid with lightning capacitor, which while still fun and impressive is not as good as any other more accepted spec.

I have read forum speculation that fire mages will gain less benefit than frostmages because fires are forced to include a scorch in their cycle while frosts spam only frostbolt. To this I cn only answer that (i) frost still can't reach fire (i) however much of a dps "decrease" having to introduce one scorch per 8 fireballs is, don't forget that a good 10%+ of the frostmage's dps is water elemental and that is also not affected by the change in mechanics.

Lastly, arc/frost will never ever out-dps full frost unless you're wearing massive crit gear. The only cases I've seen specs that work better with 40.0.21 than 10.0.51 are with monster crit. Most glaringly painfuly obvious being Mar from <Loot FTW> who had an astounding 37% crit on top of hit cap, and 1200+spellpower in frost. If you're running low crit (which is a lot more conventional for frostmages) 43+frost spec will be better by far than 40.0.21. Don't forget; the tax removal benefits ALL specs except pure arcane, and arca-fire scorch hybrids.

I wanted to check numbers on Evocation scaling, might as well do it here rather than on my computer where I'd never find it again.

2.2 Evocation
1 spirit gives 0.25 mana/2sec, Evocation is +1500% regen for 4 ticks (8 seconds)
That makes is 4 mana/2sec, or 16 mana for 1 point of spirit during Evocation.
It's 20 mana with 2/5 T6, and 17.6 or 22 mana with BoK or for humans, and 19.36 to 24.2 mana for humans with BoK.

Swapping in a Nightstaff (55+20) and a green wand (18) adds 1836 mana for me (gnome, arcane spec, no T6, assuming BoK).

2.3 Evocation
1 point of intellect gives 15 mana, Evocation restores 60% of it, 15% for 4 ticks.
That makes it 9 mana per point of intellect, 11.25 with 2/5 T6. Gnomes get +5%, BoK is +10%, arcane mages get +15%, multiplicatively as far I tested.
So, 9.9 or 12.37 mana with BoK, and 11.95 or 14.94 for gnome arcane mages with BoK.

A green staff is 56 int, a dagger is 24 int, an off-hand is 32 int, a wand is 18 int.
Exodar Life Staff is 64 int gemmed, Apostle of Argus is 59 int, Zhar'doom is 47, Nexus Key is 52. Tempest of Chaos is 22 int, common caster weapons are 18 int and more. Typical mage offhands are 0-19 intellect, wands are 0-11 intellect.

A green set has 56 (staff or MH+OH) + 30 (enchant) + 18 intellect, 104 total.
My current casting setup has 22 + 14 + 9 = 45 intellect.

So, swapping would net me 59 intellect during evocation, that's 705 more mana from swapping. Pretty low returns, but at least +30 int is a cheap enchant!


Conclusion
Looks like a buff to people with off-set gear (no/little spirit) and no swapping macro, and a nerf to people with class sets and macroed swap.
I just fear that this change removes what little incentive is left for spirit on gear.


Time will tell which changes will happen and which ones might be reverted. I'll just wait and see.

Roywyn, don't forget the following 2 facts:

1) Arcane Meditation now grants 30% regen while casting. Combine with Mage Armor that's a breath-taking 60% regen in combat making +spirit a massively more powerful tool for us than ever before. Also note, on that matter, that Blizz has said that Spirit will get looked into, but in a patch after 2.3. I expect spirit to be revamped to a much more useful state either in the next big patch after 2.3 or at the very latest in the expansion.

2)While it's a nerf in terms of usability for raiding mages, who can damn-near fill their pool up with a spirit buff, a BOK and a swapgear macro (I for one always thought +15spirit chest chant was superior to the piddly +6all stats or +150hp) It's unreasonable that we have a talent that only functions correctly when subbing as much gear as possible during it's use via macros. Also, from a PvP mage's perspective the damn spell was nearly replacing as little as 22% over its whole duration. For them this change will be a much much needed buff.

Personally I'd also like to point out two interesting consequences of the change:

1) Arcane spec mages will be making better value from Evocation as Arcane Mind will be giving them freebie mana pool increase.

2) More manapool total is better than it ever was before. It used to be that Bossfight mana for disposal = Pool + pots + Evo + gems, now it's (Pool*1.6)+gems+pots thus making a slightly more dynamic biase towards players with monster manapools (damn you gnomes!)


PS. Remarkable insight on the swapping gear idea. The thought of swapping into +int gear for Evo never occured to me and is rather amusing, but as long as i remember when you change your mana pool doesn't its % remain constant and the value change? ie. if you have 10k normally, and swap to 11k for evo, you'll regen 6.6k, but when you swap back to normal gear you'll end up with a 6.1k mana gain no? I might be talking cr@p mind, it's been a while since I took notice of such phenomena.

Thanks alot for the fast help, will be nice to outdps everyone in my guild even more then I do today

The itemisation point is correct, perhaps I hadn't thought of it enough. However I expect that given:
1) you'll only use Evo once in a fight
2) the fight will last circa 10m

I'd suspect that the extra 15% regen while casting -should- break over the deficiency of losing a heap of mana in Evo. Though acrurate calculation is a bit tough; if you're using ab/am rotation you'll be making 2 tics in full regen per cycle, and if a AM procs a clearcast you'll be getting 7ish tics in full regen making it rather complex.

Back-of-cigarette-packet calculation I'd say 15% regen in combat is circa 40mp5? so for 10m thats 4800 mana (at best) over the whole fight. Let's assume you're fully regening (and thus the bonus is nul) is what? 15% of the time? that's still 4k mana regenned. Maybe things aren't that bad?

Note, I'm neither very high-end nor particularly fluent with arcane spec output in raids. But I shouldn't think it's all that bad.

Weapon Switching for Evocation is not away - it's only worse, it's changed from spirit weapons to int weapons.

If you change a weapon with int, this changes your max mana, not your actual mana.
Let's say you have 10k max mana raidbuffed (with BOK, not acrane specced), the new evocation would restore 6k mana.
Take a mage with arena S2 blade + nightbane offhand + magtheridon wand (that's what I'm wearing actually) this is 48 int.
Now let's take some int equipment for evocation e.g. Curator Staff (51int) with 22 int enchant + an int wand (18) = 91 int = 43 additional int. With BOK that's 47 additional int.
47int * 15 = 705 more max mana.
60% of 10000 mana = 6000 mana.
60% of 10705 mana = 6450 mana.

So an additional 450 mana from switching to int weapons for evocation, that's not so much difference than spirit actually but with 10k mana it still changes evo from 60% to 65%.
It's more effective the lower your max mana is.

Come on blizzard and change evo to regen mana to full and change set boni to cd reduction.

You can farm up your own Frostfire Bracers/Belt/Slippers/Ring, can't you?

Right now I have 12256 raid buffed mana and 422 spirit, I have nightbane staff with 20 spirit and a green wand with spirit, I get ~10,5k mana back on a single evocation raid buffed with this change I'll get as little as 7300 mana and stole a spirit staff over an alt

Seems like there are 2 different people making the mage changes, and they aren't talking.

Mage Armor - Spirit Based
Arcane Med - Spirit Based (buffed for 2.3)
Evocation - Was spirit based, but going Int (Mana) based in 2.3?

They should leave Evo spirit based to emphasize the "spirit has a benefit to mages" theme they have going with at a few of the abilities. Otherwise, it would seem to make more sense to convert some of our spirit based abilities/talents into something like Dreamstate. They need to pick a mana regen theme and go with it.

Question: How many mages in the raid does it take to justify CoE?

I recently discovered that the locks in my raid do not always keep CoE up for the mages. They seem to believe the curse slot is better devoted to another curse.

We ordinarily raid with 3 mages, 3 locks and 2 spriests in the raid. Sometimes, its 2 mages, 4 locks and 2 spriests.

By my calculations, even if there were 2 mages, the 10% extra damage of 2 mages would outweigh the additional curse of a 1 warlock?

I really would be grateful for pointers in this regard as I am trying to determine whether my raid will have more dps output with or without CoE.