WotLK - Complete Mage Compendium (3.3 PTR) - Page 114

Roywyn · 10/06/08, 8:45 AM

The problem is that "Focus Magic" is not "Improved Scorch" or "Earth and Moon".
It's not a unique non-stacking raid buff.

It's a talented version "http://wotlk.wowhead.com/?spell=57934" or the perhaps a weak version of the old (2.0) "http://wotlk.wowhead.com/?spell=34460".

For level 70 3.0, we have to compare 11/50/0 to 0/53/8 (or 0/50/11, but that one is worse unless you're mana-starved and low on +hit).
I'd certainly prefer having a good estimate of the total DPS than knowing that "11/50/0 is a bit worse but offers some limited amount of additional DPS that I chose to call utility and not measure".

Ulthwithian · 10/06/08, 8:47 AM

Sondag: Would it be possible to break out the 'theoretical max' dps builds into their own list, and have the 'own scorch' builds in their own list, etc.?

Xentropy · 10/06/08, 8:54 AM

Originally Posted by Roywyn

The problem is that "Focus Magic" is not "Improved Scorch" or "Earth and Moon".
It's not a unique non-stacking raid buff.

What about the other stacking raid buff I mentioned then? Should Simcraft, if hunters are added to a future version, model Ferocious Inspiration as a 15% damage buff to the hunter with the ability instead of 3% to the hunter and four other people in order to accurately model the party buff effect (which does still stack, just like Focus Magic)?

Your argument says yes, but I think that would cause BM hunter dps to appear to blow everyone out of the water on Simcraft even if their in-game personal contribution was balanced. Blizzard themselves are taking a fairly strong "utility should not affect personal dps" tack--note Focus Magic as an 11-point talent adding 6% crit is far stronger than Icy Veins' ~2.22% haste. Most strong one-pointers of this caliber add 3% at most, which appears to be the balance point Blizzard is using--the mage's personal benefit.

As for: "11/50/0 is a bit worse but offers some limited amount of additional DPS that I chose to call utility and not measure". That is *exactly* what should be being said. As it is now someone seeing the list for the first time might say, "Wow, 11/50/0 is ahead even before considering utility, and the fact it also adds 3% crit to someone else means there's no reason for anyone to ever go 0/50/11."

To take a different tack: Basically these theorycrafting tools are supposed to give us an idea of the relative personal dps balance being brought today. Stacking utility shouldn't be included because almost every class brings stacking utility of some kind, and a raid leader will know and understand the utility being brought. A mage adds 3% crit to another caster in the group. A druid brings battle res. A warlock brings soulstones. A hunter brings 3% damage to the rest of his party with about half uptime. That sort of stacking utility is supposed to be considered separately from personal dps contribution. Merging them in the mage's case just makes mages look like we're better off in dps than we really are. As it is, 3% crit for one person is arguable against a soulstone, but if our dps were "balanced" according to your vision of theorycrafting, we'd be "paying" 3% personal dps for our utility while warlocks aren't paying anything for soulstones because soulstones can't be measured in dps.

macbeet · 10/06/08, 9:00 AM

Originally Posted by Xentropy

This wasn't as big a deal (merely a minor cost in gold, as you say) before they implemented the Lexicons of Power. Now you can't just carry stacks of each glyph on you and swap at will; you have to port to town to swap. It's as inconvenient to the raid as a respec. So the issue isn't as much one of cost as one of time and inconveniencing your raid by potentially asking for a resummon every other boss.

Ah, that makes sense, thank you. Well, I guess we will have to see how the announced mana cost reduction will turn out, maybe by then mage armor will only be needed in the absence of proper raid buffs or in special boss fights.

Søndag · 10/06/08, 9:00 AM

Originally Posted by Ulthwithian

Sondag: Would it be possible to break out the 'theoretical max' dps builds into their own list, and have the 'own scorch' builds in their own list, etc.?

That would not be a problem at all, and i think i'd have to maintain a seperate list for the lvl 70 builds anyway. I'll try to split things more up tonight so i can sort things better soon. How many lists do we want?

1. Theoretical highest dps
2. Theoretical highest dps without a raid slave
3. Realistic builds
4. Lvl 70 3.0 builds

Or did i miss something?

Roywyn · 10/06/08, 9:09 AM

If ferocius inspiration stayed as party-wide stacking 3% damage, then it should be modeled at around 15%.

It is however changed to a raid-wide non-stacking 3% damage though, competing with Ret paladins 2% raid wide buff.

Call of the Wild - Spell - World of Warcraft would be a good canditate that I would model as truly raid-wide gain and assing to the hunter (or their pet).

Assuming you can make a rotation with it, no Callbearance and if it stacks with Unleahed Rage, etc.
It requires some special pet class (Ferocity) which may or may no be DPS loss compared to other pet classes and options.

So, yes, I would assign the raid wide gain there to the pet if it's raid-wide and gets the full benefit.

Xentropy · 10/06/08, 9:14 AM

Originally Posted by Roywyn

If ferocius inspiration stayed as party-wide stacking 3% damage, then it should be modeled at around 15%.

We're just going to have to agree to disagree then, because I don't believe utility should be modeled in personal dps reports, whether or not it stacks. Some stacking utility has a dps effect and other utility doesn't; making stacking utility that affects dps directly (Focus Magic, Ferocious Inspiration) increase "personal" dps while stacking utility that can't be measured in dps (Battle Res, Soulstones) does not just confuses the issue of relative class dps balance. We should all be equal *before* that stacking utility if Blizzard's vision is to become a reality.

Maje · 10/06/08, 9:26 AM

Originally Posted by Roywyn

For level 70 3.0, we have to compare 11/50/0 to 0/53/8 (or 0/50/11, but that one is worse unless you're mana-starved and low on +hit).
I'd certainly prefer having a good estimate of the total DPS than knowing that "11/50/0 is a bit worse but offers some limited amount of additional DPS that I chose to call utility and not measure".

Wouldn't 14/47/0 be better at 70? Something along the lines of http://talent.mmo-champion.com/?mage...00000000000000

you loose 30% crit bonus damage but gain 6% to fireball (arcane explosion etc.) plus focus magic.

As to the focus magic extra 3% crit, in my opinion only the 3% you gain should be modeled the other 3% is someone elses "problem", same way as JoW isn't modeled for paladin DPS though without it rdps would suffer considerably.

Searix · 10/06/08, 9:28 AM

Alternative (and higher dps?) Builds I've come up with:
For the purpose of testing, please try Simcraft or something similiar on these builds.

17/0/54
http://talent.mmo-champion.com/?mage...00000000000000

Deep frost with arcane med regen. In addition to Focus Magic, 6% extra base damage on fireball AND ice lance should make this a rival (better?) to Ignite Frost (Note: Current Beta allows Frostbolt, Frostbolt-Lance on a Finger's proc)

31/40/0
http://talent.mmo-champion.com/?mage...00000000000000

Arcane Power. Combustion. Presence of Mind. Most of the crucial Fire dps talents.
Now don't be so quick to toss this spec aside. Numbers haven't really ever been tossed around as to how good AP and Combustion can be, especially with Molten Fury, and Presence of Mind.
Arcane Instability is about as much dps as Hot Streak. The question comes down to is Arcane Power, 12% int, Arcane Potency, and PoM better than Burnout and Living Bomb? I dare say it might.

ravenndude · 10/06/08, 9:43 AM

I've been playing around with Simcraft for the last couple of days now testing level 70 builds for the 3.0 patch. I pretty much plugged in any spec that I could think of. Here are the results that I came up with:

naming scheme is Mage_xxyyzz_nuke, where xx, yy and zz are talents in arcane, fire and frost, respectively. Nuke is what that mage is using for a nuke, if it isn't easily determined (remember, no frostfire bolt until level 75).

I used the stats from my current 40/0/21 mage, but changed them as if I re-gemmed for hitcap.

gear_stamina=409
gear_intellect=419
gear_spirit=204
gear_spell_power=1116
gear_spell_crit_rating=282
gear_spell_hit_rating=176
gear_haste_rating=69
tier5_2pc=1

Glyphs used for all mages are Mage Armor (even if the mage isn't using it) mana gem, and scorch. They are the only glyphs supported by simcraft and will be available at 70. If you want the text file I used for this, let me know where I can upload it to and I will.

Samuel · 10/06/08, 9:51 AM

Originally Posted by gaerthe

Use the mana gem like the imp as a mana battery. It regenerates mana based on some percentage of your spirit based mana regen (tuned based on testing). Put the mana gem use on the GCD and make it pull mana from the gem based on your spirit similar to how lifetap works in LK for warlocks.

...

Curious to hear other people's opinions (not that Blizzard will listen anyway).

Spirit does seem even weaker that it was in the past when it wasn't that great. Basically if you don't have all the talents to maximize it you are likely to get about the same mana by adding int. Kind of sounds intriguing but there would be a lot of details to work out and this isn't really the place to do that. It would make the gems scale with gear which is generally good. It would take it off of the shared cooldown though so not sure of the total impact.

Jonny_Monroe · 10/06/08, 9:54 AM

Just been on the PTR testing some 53/18/0 (MoE) spec. I'm not sure how many of these are reported but I'll throw in my observations here;

-Blizzard currently does not consume Arcane Potency, but does gain the benefit of it. The AoE DPS this results in on Theramore target dummies can get pretty silly.
-Arcane Potency is consumed on Hit rather than on Cast. This gives 2 results in a standard arcane rotation - 1)ABar and the Off-Nuke hit at almost the same time. One procs potency and the other consumes it. 2)The 2nd spell to land procs potency, both spells in the next 'cycle' benefit from Potency.

For 1) in the above I would assume the spell consuming Potency does not benefit from it as the spell has already left your hand and therefor already calculated its crit chance, but I'll get more samples to confirm.

-Arcane Missiles is working with MoE. as expected, its returning 30%/5 for each wave that crits.

Edit: Whilst Fireball/ABar gets normal results from Potency, spells with 0 travel time can get the double benefit, essentially a 'shatter combo', on Arcane potency. Given that, would it be worth working FireBlast into the arcane cast cycle? if Potency is up its possible to fireball/ABar/fireblast and get all 3 under potency (assuming range isn't an issue). Fireblast's DPCT is competitive with fireball's in an arcane/fire spec, so you're using a slight bias towards spells with a higher crit damage ratio and hopefully milking a slightly better overall crit chance

Edit 2: whilst I am able to consistently get a +30% crit chance for 3 spells per potency proc, the DPs i'm getting isn't showing to be much better and it hurts mana way more.

Xentropy · 10/06/08, 9:57 AM

Originally Posted by Maje

you loose 30% crit bonus damage but gain 6% to fireball (arcane explosion etc.) plus focus magic.

Focus Magic is present in both specs, so the difference is the Burnout vs Spell Impact.

Some quick math seems to put 40% crit as the breakeven point. Above 40% crit you gain more from 5/5 Burnout and no Spell Impact, below 40% crit you gain more from 2/5 Burnout and 3/3 Spell Impact.

2/5 burnout crits (w/ CSD) = 231.56% damage
5/5 burnout crits (w/ CSD) = 254.45% damage

1.06 * ((1 - critrate) + 2.3156 * critrate) = (1 - critrate) + 2.5445 * critrate

Boils down to critrate = 0.400096... Basically 40%.

Note that's final crit rate after all raid buffs. With the +10% crit of Winter's Chill or Improved Scorch I think you'd be hard-pressed to find a mage that doesn't have at least 40% crit in level 70 raids.

This is also just basically looking at full fireball spam. The fact Spell Impact doesn't affect pyroblast but Burnout does just puts the math even more in favor of 11/50/0 over 14/47/0.

Roywyn · 10/06/08, 10:21 AM

Originally Posted by ravenndude

I've been playing around with Simcraft for the last couple of days now testing level 70 builds for the 3.0 patch. I pretty much plugged in any spec that I could think of. Here are the results that I came up with:

Hm, all those numbers seem pretty low.
Do you have a shadow priest / elemental shaman / moonkin modeled in the raid as well?
They are needed if you want things like ToW/Misery/MKA/EaM as buffs/debuffs around for simcraft.

Also, is it possible to upload the full output graphics?
I'd be interested in the "damage per execute time" graphs and the "damage sources" pie chart.

I'm mostly wondering why 0/53/8 is *that* far ahead of 0/50/11. (And of almost everything else.)

Samuel · 10/06/08, 10:33 AM

Originally Posted by Xentropy

We're just going to have to agree to disagree then, because I don't believe utility should be modeled in personal dps reports, whether or not it stacks. Some stacking utility has a dps effect and other utility doesn't; making stacking utility that affects dps directly (Focus Magic, Ferocious Inspiration) increase "personal" dps while stacking utility that can't be measured in dps (Battle Res, Soulstones) does not just confuses the issue of relative class dps balance. We should all be equal *before* that stacking utility if Blizzard's vision is to become a reality.

Xentropy makes more sense to me but I guess as long as people are clear about what they are doing I'm thankfull for the work they are doing even if it does seem a bit confused. When I see these dps simulations I was just assuming they are intended to show what a damage meter would show and I suspect I'm not the only one.

ravenndude · 10/06/08, 10:56 AM

After you mentioning the huge lead 0/53/8 had, I decided to double check some thing. It seems that I had every mage with Focus Magic casting it on the 0/53/8 mage. So that data is incorrect. Here is the correct data still no raid help besides a couple pally buffs (blessing of kings, blessing of wisdom, sanctified retribution, swift retribution)

The pally buffs are all that I can add without modeling other players. If someone can let me know what specs the other support classes are going to use for 3.0, I'll try my best to model them.

Putts · 10/06/08, 12:32 PM

I'd be curious to see the build that you're using for 51/10/0. Going down the Arcane tree and grabbing only the essential DPS talents (and I'm even skipping Arcane Flows and Slow), I still seem to be sitting at 53 points, and I'm not sure what needs to be sacrificed at this point.

andastra · 10/06/08, 12:33 PM

Originally Posted by Samuel

I was still thinking about the current 8 minute cooldown so you are right but using it 3 times in one ten minute fight sounds kind of silly and even using it twice would probably mean you wouldn't be getting replenishment in a 25 man some of the time so some of these hypotheses about maximum regen from int are mutually exclusive to some extent.

Also, Student of the Mind is the 10% spirit talent so that is 450 * 1.1 * 1.1 with BoK so the number is right if we want to maximize talents just for comparison.

Lastly, JoW restores 2% of your total mana so 49.5 mana for the extra 2475 mana on every attack that hits and at most once every 4 seconds. With fireball that would be once every 6 seconds assuming haste cancels out lag and hit is maxed out. With frostbolt and arcane blast once ever 5 seconds. I don't see how a mage could get JoW once every 4 seconds. Roywyn used the 6 second assumption I think and that sounded reasonable as fire is expected to be the dps raiding spec again.

If you check any of the builds posted here, there's no room for student of the mind outside of maybe arcane specs that would also take arcane mind. Maximizing spirit talents is just unrealistic.

And you leave out the biggest advantage of int. On fights where you can get away with using molten armor, spirit provides only 30% regen. If my int is high enough to get away with using molten armor, I'm going to use molten armor.

dedmonwakeen · 10/06/08, 12:45 PM

Originally Posted by ravenndude

The pally buffs are all that I can add without modeling other players. If someone can let me know what specs the other support classes are going to use for 3.0, I'll try my best to model them.

Just add the necessary players from the default raid_80.txt and then add quiet=1 to their profiles. This will prevent them from showing up in any of the charts.

grayrest · 10/06/08, 12:49 PM

Originally Posted by Putts

I'd be curious to see the build that you're using for 51/10/0. Going down the Arcane tree and grabbing only the essential DPS talents (and I'm even skipping Arcane Flows and Slow), I still seem to be sitting at 53 points, and I'm not sure what needs to be sacrificed at this point.

It's normally Mind Mastery. Arcane Mind and Mind Mastery have the least impact on your dps. MM is ~25 +dam per point and Arcane Mind is roughly the same amount of intellect.

Edit: I should note that I do not anticipate TtW applying to raid bosses. It's way too much like demonic sacrifice and encourages mage stacking, which blizz is explicitly working away from.

Jonny_Monroe · 10/06/08, 1:12 PM

We think mages are still having mana problems and don't want to reduce Evocation's cooldown anymore. Nor do we want to buff mage armor too much more or using e.g. molten armor seems pointless.

Instead, we reduced the mana cost on Fireball, Frostbolt, Frostfire Bolt, Arcane Blast, Arcane Missiles, and Arcane Explosion.

Posted by GC earlier today. Hopefully this will allow mages to consider Molten Armour, or at very least have some mroe flexibility in choice of spells.

◊ manly · 10/06/08, 1:28 PM

That blue post was posted last week.

dedmonwakeen · 10/06/08, 1:37 PM

SimulationCraft version r680 available for download.

Mostly Warlock tweaks and a couple Shaman fixes.

Biggest change was adding Skyfire Diamond to all the profiles...... which helps close the gap on Aff-Warlocks. In fact, Fire-Mage comes out on par with Aff-Warlock provided the Warlock uses a non-damage Curse. Of course, this still requires an unbalanced Torment talent to balance DPS......

solbergb · 10/06/08, 1:52 PM

Originally Posted by Roywyn

No worries, I had a look at that same thing before, particularly with Loatheb and 100% crit in mind.
Scorch spam is actually getting pretty close there. The deciding factor would be possible differences in Ignite bug behaviour.

Where it gets interesting is if there is any "proc without cooldown" or "proc with cooldown short enough that fireball doesn't max it but scorch does" type bonus/gear/buff out there in future. Similar to the metagem+arcane blast before they nerfed it.

Also the fact that scorch/hotstreak/fireblast is a lot closer to fireball spam than it is now with even moderate crit rates is good on the PVP front, where you can't cast anything longer than scorch anyway and faster casts = more impact procs. When leveling I'm going to be likely doing something like frostfirebolt-scorch/fireblast/hotstreak spam. I don't expect many level appropriate enemies to live to reach me.

Ranch · 10/06/08, 2:08 PM

Some pages back people were wondering why only a few select spells were added to spell impact, the reason being Blizzard spam was competitive DPS already with FoF and frostbite working together, even only 6% more dps to Blizzard was probably too much. (and 12% is way too much).

Even though it is stated "we don't want every raid to have every X to function", I get the feeling they do not want all the buffs to be available in 10 mans.

Also, could we list the probable changes that come with the numbers pass? Or bug fixes?

Living bomb- only 1 per target(bug)
Living bomb-knock up removed, possible snare added? TOW synergy.
FFB- double dip threat talents, none at all?
Arcane blast- New % modifiers might be 35%/150%. (35% dmg per debuff, 150% mana increase per debuff).
Frostbolt added to spell impact.
Evocation added to Arcane Stability.
If you have 3/3 Prismatic Cloak, is the instant Invisibility still giving no benefit?
Does FFB get benefit from Burning Soul?
They could make TOW only give benefit if the target is affected by YOUR snare. IE Reason to use Slow.(maybe if it lasted longer, 30 seconds would make it beneficial)

Edit: I suck at code.