Fire vs Arcane number crunching...

Foodspec · 03/07/07, 6:07 PM

The last couple days, in my spare time, I've been working on making a mage spreadsheet to compare specs/spells/gear mainly to see what is optimal but also because I am curious to see what point value each type of stat has as far as mage dps goes.

I'm using an 8 minute fight as the test situation, and still have some work to do. If people are interested I'll post some of the actual math etc here once I'm done.

The main reason I'm posting is because today while I was doing some math everything started pointing towards fire not only being higher sustained dmg but higher dps than arcane as well. I guess I had always been thinking that arcane was innefficient but much higher DPS than the other schools of magic (and always thought that was the way it ought to be as well).

I've got an inkling that I'm doing something horribly wrong in calculating the dmg that arcane missiles does; here is what I have:

Arcane Missile
Spell dmg coefficient: 187.86%
Cost: 832
Cast time: 5 seconds (channeled)
Damage: 3248
Crit Damage: 5684
Crit chance: 20.16%
Calculated Average: 3707 (4470 on clearcast)
DPM: 4.454
DPS: 741.2

This is including talents and +dmg gear (933 spell damage after talents). Anyone see some mistake there? Because going by that math fireball has higher dps and higher dpm than arcane missiles (however that is before calculating spell hit / resists, arcane missiles benefits quite a bit from arcane focus, whereas fire spells benefit from elemental precision + curse of elements).

Mordinm · 03/07/07, 6:12 PM

What did you get for fireball numbers? What about sustanibility with 5 full seconds outside the five second rule with every clearcasting proc using Arcane Missile spam? Also curse of elements is off set by curse of shadows.

◊ Vontre · 03/07/07, 6:19 PM

Ignoring huge components of dps (arcane focus) is not a good way to go about things. Also, arcane benefits from curse of shadows, which is effectively no different from curse of elements, it's variable which you may have in a small group.

That being said, fireball definitely has higher potential dps than arcane missiles, it's suppose to. Arcane missiles is not a primary nuke, it's a strong rotational nuke with some limited pvp applications. It will also handily outlast fireball in the mana department if judgement of wisdom is on your target.

The point of arcane specs is to use arcane blast and rotations.

Axira · 03/07/07, 6:20 PM

Allready posted this in another thread, but i'll post it again:

I've been doing some calculations myself with rough numbers.
(These numbers are based upon a mage with +750 spelldamage (mind mastery not included ofc.))

Fire mage's Fireball:
Average damage: 1800
Mana cost: 400
Casting Time: 3 secs

DPS: 600
DPM: 4,5
DPSPM: 1,5
Mana/second: 133

Fire mage's Scorch:
Average damage: 800
Mana cost: 170
Casting Time: 1,5 secs

DPS: 533
DPM: 4,71
DPSPM: 3,14
Mana/second: 113

Frost mage's Frostbolt:
Average damage: 1400
Mana cost: 270
Casting Time: 2,5 secs

DPS: 560
DPM: 5,17
DPSPM: 2,07
Mana/second: 108

Arcane mage's Arcane Blast 3x rotation:
Average damage: 1500
Average Mana cost: 350
Average casting time: 1,8

DPS: 818,18
DPM: 4,21
DPSPM: 2,7
Mana/second: 195

Arcane mage's Arcane Missile:
Average damage: 650
Average Mana cost: 148
Casting time: 5

DPS: 650
DPM: 4,39
DPSPM: 0,88
Mana/second: 148

I've given the worst elements compared to the other spells a red color, and i've underlined in green the best elements compared to other spells.

I like arcane myself alot too, but I can't seem to be able to ignore the high mana costs of all the spells in there, even Arcane Blast rotations, compared to the other spells of other builds. Frost seems to have the highest DPM spell which only becomes better with all the +crit the tree gets.

Foodspec · 03/07/07, 6:25 PM

O wow, I completely blanked on curse of shadows, forgot it did arcane heh. That will change things slightly, I knew I was missing something significant. I did include the extra 5 second tick from clearcasting mana regen, but the gear set I've been using thus far for the math does not have much mana regen at all (once I start doing more work with frost I'm going to make a slightly different gear set and include more regen type stats in it, I'll also add that the gear I'm using in the math is by no means "AMAZING" but it certainly holds it's own and wouldn't be hard to achieve).

I only have the fire stats on paper so far, not fully inputted into the computer (when I'm done I'll put the whole project up here for folks to look at and hopefully find any flaws, once that's done I'll program it all and make it into a calculator for easy usage). But from what I have on paper for fireball (did more than just fireball spam in the fire casting cycle though)...

Fireball sits at around 1015 dps and 6.97 dpm.

I'll play with it some more today and fix arcane spells to include curse of shadows, that will make a big difference between arcane and fire since fire doesn't get anything nearly as significant as arcane focus.

Foodspec · 03/07/07, 6:32 PM

Originally Posted by Vontre

Ignoring huge components of dps (arcane focus) is not a good way to go about things. Also, arcane benefits from curse of shadows, which is effectively no different from curse of elements, it's variable which you may have in a small group.

That being said, fireball definitely has higher potential dps than arcane missiles, it's suppose to. Arcane missiles is not a primary nuke, it's a strong rotational nuke with some limited pvp applications. It will also handily outlast fireball in the mana department if judgement of wisdom is on your target.

The point of arcane specs is to use arcane blast and rotations.

Oh no, hehe I didn't mean it like that, I included Arcane Focus in my final math, I was just talking about the spells base dps/dpm before taking it into account. Arcane Focus is HUGE most people don't realize that. I've still got quite a few spell cycles to hash out including one which takes advantage of the fact that if you cast arcane blast and the debuff wears off while you are casting it that you get a speedy cast for the slow price.

The next spell cycle I'm gonna do will probably be frost bolt with arcane missiles casted on the clearcasts (but honestly that will look WAY better on paper than it will in reality since you won't know you've clearcasted till you're almost done with the next frost bolt).

◊ Vontre · 03/07/07, 6:43 PM

It's buried in the mage thread but I dug up to whereabout discussion on arcane blast cycles starts:

TBC Mage Theorycrafting

A sticky thread on arcane blast cycles might be a good idea since it's a pretty important new concept that may not necessarily be "common knowledge".

Foodspec · 03/07/07, 9:41 PM

While I'm at it. There are a couple more factors that I want to make sure I've got as close to accurate as possible, if anyone knows where I might find the math to them or they know the answer themself I'd really appreciate a response:

-How to calculate hit/miss/crit rate for wands.
-What the proc rate on judgement of wisdom is.

How the new mechanic effecting improved fireball and improved frostbolt talents works EXACTLY. I've heard from a number of people that getting the -.5 second cast from those two specific talents effects the spell damage coefficient for the spells they represent. I need to know exactly what the modifier to the spell damage coefficient is and if this change is in fact verified.

Gerronimo · 03/07/07, 11:32 PM

JoW is generally accepted to be a 50% procrate.

The improved fireball/frostbolt talents reduce your spell damage coefficient by 2% each point. So 10% all up. Unsure on whether the frostbolt coefficient is multiplicative or subtractive from the 81% untalented cefficient, so i'll leave that wisdom to the mage thread. Its buried in there somewhere.

◊ Yes · 03/08/07, 12:03 AM

I'm in the process of making a spreadsheet based on (another) Warlock DPS Spreadsheet

Correct me if I am wrong:

damage = (avg dmg + spellpower * cast / 3.5) * modifier

Modifier for Fire Power / Playing with fire mage = 1 * 1.1 * 1.03

All x% more spelldamage effects are multiplative?

735 Fireball
333 Scorch
725 Fireblast
264.5 Am
720 Arcane blast
623.5 Frostbolt

Spellpower With ignite:

175% crits, 40% of 175% dots.

With ice shards:

225% crits

Arcane potency is 3% crit net,

combustion with fireball is around 6.5% crit, with scorch around 2.3~

I am still a little confused on where the nerf on imp fireball frostbolt goes IE does it + empowered fireball = 5% more spelldamage or is it multiplative

Anything else I might be missing?

Tempestra · 03/08/07, 1:56 AM

Originally Posted by Gerronimo

JoW is generally accepted to be a 50% procrate.

The improved fireball/frostbolt talents reduce your spell damage coefficient by 2% each point. So 10% all up. Unsure on whether the frostbolt coefficient is multiplicative or subtractive from the 81% untalented cefficient, so i'll leave that wisdom to the mage thread. Its buried in there somewhere.

Your +dmg will be:

[base + (spell dmg)*(.9)*(.81)]

Unless you have empowered frostbolt/fireball, in which case:

[base + (spell dmg)(.9 + empowered)*(cast time modifier)]

Foodspec · 03/08/07, 3:46 AM

In response to the last 2-3 posts regarding spell coefficients. They are addative not multiplicative, as well as +dmg% effects. Tempestra your math is slightly off for that reason.

Gerronimo: Thanks I'll use that 50% proc rate for the time being until I see evidence otherwise, but I'm fairly positive that is accurate.

For any of you who were curious, yes it is -2% coefficient per rank of imp fireball/frostbolt.

Anyone have a link for math related to wand hit/crit rate and which stats (if any) effect it?

Athinira · 03/08/07, 4:30 AM

An important point most people often overlook is lag, and it is certainly important to incooperate in the equation.

Take a look at this screenshot taken on my paladin:
http://i73.photobucket.com/albums/i2...107_181549.jpg

Note 3 things:
1) My Ping is 96 ms
2) Im casting a 1.5 second casttime spell (Flash of Light)
3) Even though there is a 1.5 second global cooldown on FoL, the global cooldown is almost over (0.05 seconds from completion or so) while there is still 0.6 seconds left of the spell.

This means that when you want to do the DPS number crunching, you should, in regards to lag, add 0.3-0.4 seconds or something similar to the casttime. This means that a mage with improved Fireball will not cast a Fireball every 3 second, but rather every 3.3-3.4 second on average. Same with Scorch, 1.5 + 0.3 or 0.4 = 1.8-1.9 seconds. Same is applied to every other spell.

Most people often overlook that fact, but my screenshot clearly shows the lag is a major factor in overall DPS, and this is lag even in low-ping enviroments. This massively favors spells with a longer casttime (including channeled spells).

Edit: Instant cast spells that aren't channeled (Fire Blast, Ice Lance etc.) are not subject to the rules about lag above. They will most likely have 0.1-0.2 lag-cooldown added to the global cooldown max in a low lag enviroment, down to 0.0 with correct button spam timing.

Foodspec · 03/08/07, 4:46 AM

Originally Posted by Athinira

An important point most people often overlook is lag, and it is certainly important to incooperate in the equation.

Take a look at this screenshot taken on my paladin:
http://i73.photobucket.com/albums/i2...107_181549.jpg

Note 3 things:
1) My Ping is 96 ms
2) Im casting a 1.5 second casttime spell (Flash of Light)
3) Even though there is a 1.5 second global cooldown on FoL, the global cooldown is almost over (0.05 seconds from completion or so) while there is still 0.6 seconds left of the spell.

Ya thats where one of the biggest differences between theorycraft and practical "real-world-craft" comes into play. I'm just going to do the math without taking that into play, it's just one of those variables that differs from player to player, we all get the occaisional lag spike etc.

The lag + the travel time for most spells really play a big factor into cycles which involve casting a specific spell on clearcasts. I've decided today that I'm not gonna even bother with those types of cycles just because I know that I will never be able to utilize that type of cycle personaly (frost bolts w/ AM clearcasts).

That's as far as I'm going to go as practicality/implementation goes, lag is just one of those things that is so hit/miss / sketchy that I'm not gonna try to factor it in heh.

On a side note I'm up late right now reworking some things with various new things I found out today. Looks like my inclination was right earlier today and that arcane missiles was A LOT better than my math was giving it credit for.

Foodspec · 03/08/07, 5:47 AM

Alrighty, getting ready to get some sleep now, but I completely did my arcane build tonight and did one of the cycles for it as well (the most basic of the cycles, pure arcane missile spam). I included a number of things which I had not factored in previously, most importantly of which was judgement of wisdom (which made a VERY noticeable difference). I'm not including any consumable buffs, but I am including super mana pots.

The result that was that Arcane Missiles was much more effective, and was about where I expected it to be prior to starting this project. However, I have a strange feeling once more; this time that the regen seems a bit high...

Anyways here's the nitty gritty, lemme know if you see any errors just by skimming it over or if there are any questions on how I got to some particular stat (I expect there to still be a few errors).

Arcane build 1

Talents: 43/0/0 (with 18 points to spare) http://www.wowhead.com/?talent=oiGV00czxIui
Stats:
Int: 624
Spi: 267
+DMG: 975
Hit Rating: 69
Crit Rating: 159
Base Mana: 2250 (close approximation)
Mana: 11610
Crit Chance: 21.00%
Hit Chance vs lvl 73 mob: 98.5%
Wand DPS: 186.3
AVG wand hit: 279
Active Regen: 247 mp/5
Passive Regen: 79 mp/5
Evocation: 4755

Spells:
Arcane Missile
Spell dmg coefficient: 187.86%
Cost: 832
Cast time: 5 seconds (channeled)
Damage: 3652
Crit Damage: 6391
Crit chance: 21.00%
Calculated Average: 4227 (5049 clearcast)
Mana Consumption: -568 mp/5 ( +432 mp/5 on clearcast )
DPM: 7.44
DPS: 845.4

Arcane Cycle A (Build #1) (0-480 seconds)
Arcane Missile Spam, nuff said...

0 - Arcane Missile x8
40 - Mana Pot + Mana Emerald
43 - Arcane Missile x2
53 - Clearcast AM (10012/11610 mana)
58 - Arcane Missile x10
108 - Clearcast AM (4764/11610 mana)
113 - Arcane Missile x8
153 - Evocation + Mana Pot + 1 extra mana tick (7622/11610 mana)
163 - Wand x36 (Full Mana)
217 - Arcane Missile x2
227 - Clearcast AM
232 - Arcane Missile x10
282 - Clearcast AM (5658/11610 mana)
287 - Arcane Missile x9
332 - Wand x54 (6471/11610 mana)
413 - Arcane Missile x11 (223/11610 mana)
468 - Clearcast AM x2
480 (END)

Arcane Missile: 60 casts > 253620 x 98.5 hit chance = 249815
Clearcast AM: 6 casts > 30294 x 98.5 hit chance = 29840
Wand: 90 casts > 25110

Approximate end damage: 304765
DPS: 635

• Maledict · 03/08/07, 6:03 AM

A sticky thread on arcane blast cycles might be a good idea since it's a pretty important new concept that may not necessarily be "common knowledge".

Please, for the love of god no. Sorry, but I don't know why this has a seperate thread - the mage thread contains *pages* upon pages of identical theorycraft, and unfortunately because everyone uses different methods and no-one writes up how they arrived at their numbers or even what gear they are using, it doesn't really add much to the debate.

Also, stickying something about "arcane blast rotations" would be a mistake, simply because outside of theorycrafting, sticking to a set arcane blast rotation just guarantees you'll be outdamaged by the arcane mage stood next to you who isn't. Cycles are great for theorycrafting, but arcane comes into it's own when you learn exactly how hard you can push your mana, and how you can change your damage output depending upon what's happening around you in your group / raid. Beyond theorycrafting out the most DPM cycle or the highest DPs cycle, everyting else should realy be learnt by the mage - convincing peopel that "cycles" is the best way to play the class is only neutering the amazing power and flexibilty that the expansion brought us.

◊ Yes · 03/08/07, 5:17 PM

Early version up.

Looking for feedback

Grey = not done implementing yet\

IF YOU CHANGE the forumulas, pleeease write somesort of readable changelog of you plan to use it for non personal use!:P

*updated to include frostbolt

◊ Yes · 03/08/07, 8:24 PM

Big update!
Looking for feedback

Updated again to include arcane power.

NEED TO DO: Better molten fury/combustion implementation, include water elemental, include spell rotations (ie fireblast, or arcane blast), Include X length fights, model int > crit, model crit / hit rating into crit.

Gerronimo · 03/09/07, 12:30 AM

I will second what Maledict said- Leave this thread to die. It clogs up forum space for other classes and all the discussion is currently in the mage thread.

Having said that- Yes ( the player) that looks to be a very nice sheet. It will get very nice feedback from other sheet crafters in the mage thread.

I apologise for the irony of bumping this.