Vand1

At the risk of being shot down for asking a totally stupid question, are we 100% sure that penetration does not overcome this resistance at all? I have often puzzled over why Bliz insists on putting such a seemingly useless stat on raiding tier equipment. Could this possibly be the reason?

 

Northerner

Extensive testing showed spell penetration as useless for overcoming level-based 'resists' back some long time ago and I've never seen contrary evidence since then. It is theoretically possible that things have changed but I've never seen any data suggesting that.

Moreover, people have bitched about Spell Pen on gear for quite a while and I would expect Blizzard to have at least popped in with a "Oh, it's more useful than you think! =) =) /wink =) =)" or whatever by now were it to actually have a purpose in PvE.

 

• Navaash

For what it's worth, the unmitigatable partial resist thing might be the byproduct of some ancient game mechanics.

In the original game's alpha, Arcane, Fire and Frost (the class skills you can see in your Skills tab) had numbers. (Same with all other caster classes.) Originally, you had to cast spells repeatedly to raise the numbers to the current level, just like how weapon skill works. (You can guess how fun this was.) Blizzard quickly realized this was a bad idea and in beta, changed it so that it auto-leveled the number as you leveled up. Eventually they hid the number.

(Of course, I'm not sure why languages still show 300/300...)

The main point I'm making though is that, unlike weapons-based classes who have had weapon skill (until 2.3), casters have never had a way to raise these skills. It's always your spell skill 300/350 vs. the opponent's 315/365.

 

Pintofbrew

Perhaps I missused resilience; A post by Aeus many months ago stated that Boss type mobs have an in-built 5% chance to partially resist spells which could not be mitigated by any means.

My proposition was not that resilience has anything to do with hit, it was that this effect (which I wrongly called resilience) was working wrongly. Here is my evidence:

1) No simulation I've ever looked into takes it into account, yet simulations are often within 1-2% accurate with reality. Thus perhaps:

2) It may be broken, registering as full resist on normal spells but not registering as such on Binary

However, it seems that evidence is mounting towards Empowered Frostbolt granting 5% hit, making this eventuality rather remote and wrong to boot.

Vand1: Yes, Blizz post at the time stated it was 100% unmitigatable (is that even a word?). Penetration does not affect it, or so they say. Poor old penetration... The stat that everyone forgot...

Navaash: This stands up to scrutiny; When WoW was being developed I believe it based a whole lot more than we think on Diablo 2 LOD. There, appart from wanting high armor a player would quite badly need max resist in each of the 4 elements. Particularly for PvP and certain high-resistance encounters, "-% enemy resist" gear became massively powerful. I believe the whole resistance shebang started out to become a big issue in WoW but totaly failed, leaving us with this relic of a half-assed design we have to cope with now. That is, 99% of the time resistance is (ironically) futile. Unless (1) you're a tank on a massive-elemental-specific encounter or (2) you're a lock in PvP, and then only because you can get tons of it just for having a pet out.

 

Rouncer

Well there has always been one way to remove those level based resistences, binary spells.

The mechanic for Binary spells means that having +hit over the cap nullifies the target's resistences even if those resistences couldn't be affected by spell penetration. It was proven by a couple of Warlocks just before TBC came out that the 99% cap only applied to the second equation and that the first one is uncapped in the case of binary spells.

The famous Eyonix the mage example

What the warlocks proved was that the first part of the equation:

0.83 (83% for +3 levels monster) + 0.06 (+6% spell hit) = 0.89

is actually uncapped so that if you have +hit over the cap you can get above 1.00 and then when applied to the 2nd equation it will function as spell penetration gear, since the 0.99 cap is applied at the end of the equation. Which is the only way those unpenetratable resistences can be penetrated. I think they worked out that every 1% hit over the cap functioned as ~14 spell penetration.

True this probably has nothing to do with what is currently being seen with frostbolts and boss resist rates, I just think its an very interesting mechanic.




What I'm more worried about is if we find the mechanic functions the same when the patch goes live and then everyone starts gearing to hit that much lower cap as Frost and then Blizzard goes "whoops that's a bug, hotfixing it tomorrow" and now we got to run around and regem and what not to find our way back to the normal 164 hit cap. Also if it is shown that it is associated with Empowered Frostbolt I would probably spec without elemental precision and gear to hit the 140 cap for bosses.

I would just like to find a logical rationale for the lower hit cap for binary spells without it being something like "Blizzard forgot to remove the +5% hit from the Empowered Frostbolt talent".

 

◊ Vontre

I didn't make a real thread. Profanities in the title were probably not looked upon well.

www.magegraf.com

Raiding is full of challenge. Sometimes there is fire. You have to not be in the fire.

"We agree with Communism." - Greg Street 2009

 

Qbert

Fact: Penetration does not affect level-based resistances.

Realistically penetration is entirely a pvp stat. My main gripe with it though is that it is entirely useless when your target has no resistances so you're never going to realistically wear penetration on your gear just in case and cannot change gear in combat after you realize you may need penetration.

Good news is that they seem to have wisened up and finally implemented what I've long considered the only true way to make penetration valuable which is to pile it onto weapons (re: season 3 staff & offhand). Penetration is only good when you need it and you never know you need it until you're being resisted since there is no way to ever see a target's resistances ... thus the only viable implementation for penetration is making it swappable in combat, a la weapons.

 

Athemeus

First, you would at least have the abillity to perform well at those encounters, where fire is less useful, whereas with fire as the only tree, you could only sit out. Although I don't really see those encounters around (except for the resist ****).

Arcane sucks in PvE solo, to a degree where I often die farming, while fire gives you gimmicks like blastwave, dreagons breath and range. Arcane specced, you can basicly just frost nova once and then take the beating until the CD is up again.

One of the best things of arcane is the synergy with other classes or one could say, the lack of it. You just need to worry about CoS, which you can rely on anyways (also malediction).

I agree that the AB mechanics are not so easy to balance around, but at the moment a talent in the arcane tree like 30% more mana cost/damage to AB (instead of prismatic cloak or emproved blink+removal of 2T5)) and a change to empowered AM should be sufficient to solve the situation and make the trees scale similarly.
I would have to do the maths for this first though, before suggesting it seriously.

 

Kir

Arcane sucks for soloing? huh?

Maybe at other gear levels, but for me, one arcane blast and one arcane missiles kills most any lvl 70 mob that I would farm (I actually take off my MSD helm because I've got a better overall dps helm, since half of the focus procs are wasted). If clearcast procced on the death of the last mob, I open with AM on the next, as the crit chance will often kill the mob with a single AM, with a frost nova+ice lance as a finisher if needed.

With mage armor up and mushroom farming trinket from sporeggar, I can farm effectively non-stop. I use evocation when it's up, and it's almost always up again by the time I would need to use it. But, I do take a few hundred damage per mob, as they hit me a few times before AM finishes them off, but my health goes down slower then my mana as I farm. So, I occasionally bandage/cannibalize if my health is lower then my mana. If my mana is real high, but my health is low, I will sometimes AB/Slow/back up/AM or AB/Frost nova/backup/AM for a few mobs.

Only when my health and mana are both low do I sit and eat/drink. This happens like once every 20-25 minutes. I often summon the highest two ranks of mana gems while in town, and use those while I farm if mana is low and evoke is on CD. Basically, I can delay my mana issues long enough that I have to visit a vendor/town first.

As for taking down 'harder' mobs for a quest or something, well if it's not susceptible to slow, it's not snareable, and you probably need a group anyway.

edit: and if you are on a pvp server like me, slow + AP/Pom/Pyro lets you take down a lot of would be gankers if you aren't oom when they attack. A critted AP pom/pyro + one AP AM takes down a lot of classes by itself. If a warlock ganks you, well.. they are hard 1v1 anyway, especially if they get the jump on you. If you are in a bad position, slow can often let you escape at least. It's still not as strong for pvping as frost, but it's better then 10/48/3, that's for sure.

 

Tyrian

Is this a joke? Do you realise how strange this comment sounds... are you sleeping while you farm! :S

Shadow priest, Shaman? For serious guilds that want to make the most of raid comp setup to maximise player dps, you will want to consider more than 'just CoS' when it comes to helping arcane spec mages.

I like that arcane and fire require different gear and other classes to get the most out of. However, like others said - Blizzard needs to come clean with their intentions for Arcane.

Personally I would like them to make arcane as equal dps to fire at face value (without the gimmicks) and focus more of the arcane talent tree on ways to change your playstyle while keeping that dps. Im not going to pretend I know how they would balance or go about this.

AB let us play with rotations - I found that fun. Arcane missiles let me have some fun with experimenting with synergy/haste gear and items. But with those gone and seeing that fire is the 'proper' spec for dps in 2.3 (with frost acceptably close, if you want your survivability instead), all the rotations/haste/synergy fun arcane offered to broaden our playstyle if we chose to try it - is thrown somewhat out the window. Sorry blizzard, we'll have to put aside all those fun playstyle mechanics because fire is 'the serious' dps spec once again.

 

Muphrid

There's a more profound question: given the asymmetric effect this has, to the point where it favors Frostbolt over other spells, I'd say we're not even sure this is really "resistance" in the first place.

 

taciturn

[quote=Tyrian;535461]
Personally I would like them to make arcane as equal dps to fire at face value (without the gimmicks) and focus more of the arcane talent tree on ways to change your playstyle while keeping that dps. Im not going to pretend I know how they would balance or go about this.
/QUOTE]

Due to no ramp-up time, **sharing CoS**, superior aoe, and no issues with resists, it seems unlikely they will be ballancing them around equal dps.

Unless range becomes an issue, it just wouldn't make sense to spec fire when you can save a curse (one warlocks seldom want to cast, and almost never gets malediction) and a debuff slot.


Seems more likely to me that they will try and salvage arcane with some gimicky passive brilliance aura.

 

◊ manly

I already answered how they could scale arcane in a balanced way. Bring arcane damage in line with firespec dps. This will lead indirectly to higher DPS than firespec because of no ramp-up time. But to make this a fair deal, make the arcane tree have a *really bad* DPM. Something a bit worse than AM spam we have right now, but for considerably more dps. You give extreme mana consumption in exchange for low threat and no ramp up time.

If I were to make some very very rough numbers, I'd say increase AM DPS by 10-15%, ind decrease its DPM by 15-20%.

Log on with different model:
1- Create a character of the desired model. Log on/off.
2- At character selection screen, select your actual character; mouseover the new, desired model character, and hold down left click; hit enter and release left click at the same time.
bug Arcane Potency only applies to the first Arcane Missile bolt.

 

Roywyn

Using fire and arcane spec with CSD, Skull of Gul'dan, Hex-Shrunken Head, the numbers are as follows. (No gimmick tricks, not using Ashtongue Talisman and MSD is very noticable. Using a shadow priest and 10% CoE/CoS.)

AM spam:
1488 DPS, 14m with JoW, 6m without JoW
0.621 DPS gain per point of +damage

Fireball spam:
1839 DPS, 12m with JoW, 8m without JoW
0.843 DPS gain per point of +damage

So, we're talking about 24% better damage and 35% better scaling on fireball.

Arcane wins at ramp-up time, spell pushback (and threat, if we were still in BWL).
AM synergises with CoS, and AM longvity has a massive synergy with JoW.
Arcane gains damage from intellect, so the pure DPS/+dmg ratio can be even a tad lower.
Fire has to use scorch, but can compensate for that with using Flame Caps.


If we want AM viable, those differences should be at most 5-10% for the damage, and 10-15% for the scaling.
That means at least 20% better scaling.

AB similarly suffers from bad scaling.
Make 2T5 a talent in the middle of the tree.
Make the mid tier damage talents better - 2% crit on AE/AB but not AM? And crits are only +75% anyway? Make it 3% at least.
AM uninterruptable - 5 talent points just to be able to use that spell.
Add a "crits reduce the cast time of AM by 0.1s for 8 seconds, stacking up 1/.../5 times" to it, and a utility talents for one spell adds DPS and no DPM. Adds a little incentive for crit, and a little ramp-up time so people don't complain about burst in PvP.
15% intellect for 5 points - that's 1.5% crit, 25 damage and 2.5k mana (inc. Evocation) if I stack int. Other trees offer 10% damage for 5 points. Have it add 1% crit per point as well.


It's not impossible to fix arcane, but it takes a lot of thinking and balancing to make it viable without the gimmicks.
Basically, the first 5 lines of arcane talents need to be better to make the tree competitive again.

 

Muphrid

I must be pedantic on one point: imagine, for a moment, that at the no-gear level, Arcane deals less DPS than Fire. Now, give them "equal" scaling under your definition. With this sort of scaling, the ratio between the two goes to 1:1, essentially trivializing the difference between the two with equal gear increases.

In other words, knowing the added DPS or raw damage added per point of +damage is convenient for calculation, but without the context of base DPS, it is also an incomplete measure of scaling. It does tell us whether one spec will out DPS another; it doesn't tell us which spec is becoming more viable.

 

Searix

One of the things i cannot stand is arcane's reliance on shadow priests. There are times when i have to absolutely beg to get into the shadow priest group, and then a person who needed it goes oom (We run with 2, try to keep both in but sometimes one has to go for whatever). Doing Kael without a shadow priest or JoW as AM spam spec i started phase 3 (weapons) at full mana, ended up oom after (2pc t6 evocate with spirit weapons included) at the start of p4 spamming super manas and gems. I literally had no other option but to wand for most of the phase to get enough mana for eggs/shield

My point? I'm very dissatisfied with arcane's current role, and possibly that's what devs are seeing too. Why should a spec's role be completely reliant on Another class being grouped with you.

I don't know about the rest of you mages, but not fearing being 12th on dps just because i don't have a shadow priest next patch is a godsend, at least people understand why fire spec won't be as much damage on fire immune bosses.

 

chase

I remember seeing this claimed and every time I asked for a link to the data an answer never came.

I think it is possible it isn't capped in the first equation, but I have never seen proof.

Can you link me the data/post?

 

Roywyn

Yay, I finally got a [Chaotic Skyfire Diamond] on the PTR!

TLDR
Chaotic Skyfire Diamond works like the early RED/resilience crit modifiers.
It's not a straight 1.03 multiplier.
Critical damage talents get amplified with this meta gem.
/TLDR



Untalented AM:
581 hits, 898 crits
898/581 = 1.545611 ~= 1.545 = 1.5 * 1.03
So, the usual 3% increase, yay.

Talented AM
759 hits, 1379 crits
1379/759 = 1.81686 ~= ( 1 + 0.545 * 1.5 ) =/= 1.75 * 1.03
So, it definately doesn't act as a 1.03 multiplier for crits.
It works like the old RED used to work, and how resilience used to work.

I.e. assume a 1000 hit, it becomes a 1500 crit without talents.
The gem increases it to 1545 = 1500 * 1.03. The that damage is split into 1000 hit damage + 545 crit damage bonus.
The spell power talent (+50% crit damage) would then increase the crit damage bonus to 817.5 = 545 * 1.5, and increase the total spell crit damage to 1817.5, instead of the simple 1802.5 = 1750 * 1.03 crit damage.

Checking it with a frost spec Frostbolt:
Talented base damage: 636-686; geared hits: 1735-1782; geared crits: 3630-3733
3630/1735 = 2.092219, 3733/686 = 2.094837

Checking the damage differences (top end minus low end), we see that the hits are 3 points to narrow, and the crits up to 3 points to narrow, giving us an error margin of ~0.35% for each, and ~0.7% when we take the quotient.

So, a 2.09 crit multiplier is very well in our reach, while a 2.06 is far beyond or possible error.


With this, critical spell damage becomes as follows:

150% => 154.5%
175% => 181.75%
200% => 209%
225% => 236.25%

210% => 216.3% for 150% + Ignite (Fire)
245% => 254.45% for 175% + Ignite (Arcane/Fire)

 

Rouncer

As far as I know the link is completely gone.

It was one of the old mage stickies which was just a rundown of all the known TC about the mage class mechanics. On the last page before it locked a warlock linked an experiment he had done with another warlock to test binary spells and resistences.

Basically they alternated CoR and CoW on each other with varying levels of spell resistence and varying levels of spell hit and after 1000s of applications they had enough data to prove conclusively that +hit over the cap affected the target's resistences which meant that the equations were uncapped on the first part and only capped on the second part.

About 2 weeks after the post and the thread locked (it was literally 3-4 posts before the lock) the forums (and all of the stickies) were all erased when they updated them in preparation for TBC.

 

Muphrid

That is to say, the real crit bonus for on-hit crit mechanics is given by b' = 1.09*b where b is the ordinary crit bonus? And thus, for Ignite, it's given by b' = 1.09*b-.036?

 

Darkchani

Nice tests !
What would be the crit rating vs spell dmg under such crit damage bonus ?

 

Muphrid

Well, given the conversion from crit to +damage is...

Substituting b' for b, knowing b' > b, we see that this expression increases with increasing b. Thus, 1 crit rating becomes worth more +damage than it was before. You can easily substitute b' = 1.09*b or 1.09*b-.036 as needed.

Edit: that is, crit is worth (1/b+c)/(1/b'+c) times as much +damage as it was before.

 

Darkchani

Obviously crit rating value increases when the crit bonus is bigger, didnt need any formula for that !
Now a formula is nice to know the exact value, thanks for that, except you're not exactly helping alot without indicating what each variable is :/

 

Muphrid

For original crit bonus b, crit chance c. And, of course, b' = 1.09*b without Ignites, or 1.09*b-.036 with Ignites.

And base damage m, +damage coefficient r, +damage d.

 

Faxmonkey

I'm amazed by both Jayde and Frenzi's comments regarding frost nova. You guys are apparently quite luckly if you have AoE tanks for anything but Hyjal Trash. In my guild, the rule is: Watch where you're standing. There's no AoE tanks on things like Morogrims Murlocks. We just them all cluster at the same point frost nova them and then with a Hunter trap Down, run in circles around the outer edge of the murlocs spamming AoE while the warlocks do SOC from range. If all goes well, everything is dead before it gets off the trapped area -- thanks in large part to frequent frost novas from all the mages.

There are no tanks. They run towards the healing paladins at the trap location, and once they're in a good clump, the first frost nova goes down. The paladins move away -- and from then on, no one gets hit. Maybe one or two people get hit a couple of times if they get too close -- but by and large, the murlocs are sitting there frustrated that they can't reach anyone.

That's how most AoE fights go for us -- except hyjal trash, which involves one warrior tank per every few mobs. If your guild has some sort of different way of doing it, where frost nova is a bad thing, maybe that changes the rules somehow.

This point could be argued indefinitely and we'd never reach a consensus. I specced arcane for DPS until now, but Illidan is dead. DPS is not my top priority anymore and I was only specced for max dps because the difference between arcane spec and "everything else" was so extreme. Next patch, it'll be a choice of frost or fire and the difference won't be very extreme at all. Roughly 100 dps or so according to Lhiveras calc (1450 vs 1570 I think it told me).

I don't know. Maybe you're thinking of things like, "zomg Illidans at 32% lets get him to phase 5 before he does another Parasites" where as I'm thinking "Phase 1 Magetheridon is 2 minutes -- and it's the only part of the fight where DPS matters. Thus, frost mage is much better for learning Mag becuase it does the most dps within the 2 minutes where dps makes any difference." I suppose it can go ether way.

Nothing is optimal 100% of that time. So everything is bad?

This is just silly. Deep Fire's Arcane Explosion is the same as Deep Frosts -- except minus the +50% chance to crit. Your Flamestrike only does like 10% more damage unless you got Improved FLamestrike . . .

At any rate, yes, Shatter MORE than covers the gap of having "slightly better flamestrike". Especialy when TLC was in the Mix, but with it getting nerfed in 2.3 that margin will narrow a bit in Fire's favor.
Still, if shatter can be used, then Frost AoE is far superior.

As I said, if you wish to consider the question of a frost mage's total +damage per second, That is (His scaling), you have to consider The Water Elementals 35% divided by his percentage of uptime and add that to what you get from Frostbolt Spam. Even if you factor in all raid buffs/debuffs, Frost still comes out a *VERY* tiny bit a head. Why? Well remember frostbolt gets +15% to crit from talents (this counts!) and more +damage% since they altered Arctic Winds. It all ends up being almost identical scaling on the two to be honest . . .

The big issues for Frost Scaling are actually the fact that while your pet will get 35% of your +damage every couple of seconds on his water bolts, he gets 0% of your +hit/+crit/+haste rating. With that in mind, if you have lots of crit/haste frost probably falls back behind again. The point to keep in mind though, that contrary to your assumptions it really is a dead heat. The two specs scale pretty competitively with each other . . .