◊ manly

Im sorry, but you can't use theorycrafted numbers and expect the same results from actual testing. Spell rotations are good to give you a good idea what to expect, but given the bosses in BC you will never get to 'play by the book' and follow precisely those numbers. The closest fight where i could maintain a perfect uninterrupted spell rotation would be void-reaver *if* you manage to never need to move at all, which is unlikely to happen given that your totems are likely to move.

If you want to give yourself a benchmark to compare against, check WWS parses. Those will take into account the things dps spreadsheets cannot.

The short short version: Make sure the scorch debuff stays up. Otherwise fireball. Stay in range of totems. Try to pop cooldowns when it makes the most sense (ie: combustion at 20%-, icon of silver crescent during bloodlust).

 

◊ Vontre

Yeah it's just for spamming one spell. Generally with the exception of arcane specs (which can only really be handled by expert mages in the first place) a mage's optimal casting cycle is one-button spam.

The best way to confront your problem would be to link us to the armory profiles there and I'm sure the posters here will pick up any obvious problems rather quickly =).

10/48/3 is a great raiding build, the spell cycle would be fireball spam with the occasional scorch to maintain the debuff.

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

 

Zyrexana

I'm not new to the game or theorycrafting, you know, or even caster classes. I'm new to mages. I want to use the spreadsheet to project information to give me an idea of how numbers in general can look like and stuff... I never suggested they'd directly translate...

I simply requested a translation of what the chart meant when it said this or that, and extra help if anyone is willing to give it. And... Thanks a bunch, Vontre. =)

Well, two of our main mages include...

http://armory.worldofwarcraft.com/#c...rgeras&n=Drink

http://armory.worldofwarcraft.com/#c...argeras&n=Bhal

I'm going to play a mage a while if that's what it takes, I'm serious. =x

 

Tempestra

As Manly said, checking out WWS parses (or better yet, give us their 620 dps mag parse) would be the best way to see how mages handle each fight.

Don't use, or tell your mages to use, any rotation except 1 scorch 6-7 fireball. That's how mages are now, for better or worse, and don't let anyone convince you otherwise (I'm assuming you're spec'ing for 25-man raids).

As for Drink - why in god's green earth does (s)he not use spellfire vest but has the other two pieces on? Also, +8 int gem in the shoulders... really?

As for Bhal - his gear is fantastic for pre-ssc non-tailoring. No excuse not to do well.

 

Zyrexana

Hmm, I'll see what I can do about Drink with spellfire, but Bhal's dps is now lately showing some significant improvements, thankfully (the spec is new for him, he was some hybrid mid-arcane mid-fire scorch spec before).

http://wow.virakar.com/stats/wws-20070529-2044/

This is a parse of Magtheridon where Drink's dps was 628, fairly recent.

Besides that, is that truly the end-all be-all spec? Are there any viable frost specs available? Several of our mages keep experimenting with pve frost. Any thoughts on that?

http://armory.worldofwarcraft.com/#c...rgeras&n=Carea This guy is trying to make pve frost work.

And any thoughts on deep-arcane fire?

(Again, thanks a MILLION for all your help, thoughts, and input. I really appreciate it.)

 

Jayde

That is quite cool. I'm sending the link around to my Mages to have them take a look at it. Thanks.

Vontre: I mentioned it briefly a bit ago, but I think it got lost. Have you at any point done any calculations of Scorch ramp-up, and when/how many it becomes worth it on shorter fights? It seems like this somewhat depends on, for instance, if you have multiple fire mages or not--but there has to be some point where the time it takes to ramp up is too long to be worth it. Sadly, I don't know the inner voodoo of your spreadsheet quite well enough to figure out where that is yet.

 

Copernicus

A minor update to my item spreadsheet - http://www.ainur-guild.org/stuff/magegeardps2.htm

I'm waiting on doing a zero stam version of it until I figure out the macro to automatically sort the spreadsheet with a button.

The values I used for damage-equivalent points were-

Crit Rating 0.67
Hit Rating 1.00
Haste Rating 1.00
Stamina 0.50
Intellect 0.20
Spirit 0.11
Socket 9.00
Meta 12.00

I changed how much I valued hit rating based on a the hit rating thread here. It's still more valuable on boss fights, but due to fights with adds I slightly lowered how much I thought it was worth.

 

Densor

Hey, I was doing some thinking and some programming and have two things to contribute to the spreadsheet, if you haven't already applied or decided to fully ignore them.

First, the Mark of Defiance proc is effectively a 150 mana regen proc with a 15 second internal cooldown and 15% proc rate. Thus, the easiest way to model it on an extended fight is to apply it to the mp5 section. 150 mana regen per 15 seconds, plus proc time, which is 6.66 spell hits on average, which can be determined based on talents and/or cycles or whatever. For myself, as an arcane mage, it's 150 mana per 25 seconds. The 25 seconds comes from 15 seconds of internal cooldown plus 1 second of leeway plus 9 seconds of average proc time, calculated from a ABx2 AMx1 cycle ( (2.5s + 2.2s + 5s) / (1 hit + 1 hit + 5 hits) ). For a fireball spam, it'd be closer to 34 seconds per proc. I hope it's not too hard to add to the sheet, if you haven't already, because I'd think 30 mp5 would have a big impact on many fights. BTW, the 15 second internal cooldown is from examination of my combat logs after spamming rank 1 scorch on a friend for a while. That's the closest it got, and I've never seen it proc within 15 seconds of another proc. I don't have a mod to keep track of it over a long period of time, though, so that's the best I can give you.

Second, I was writing a quick and dirty simulation for Combustion for fun and figured I'd share my findings. The easiest way I can think of to add it in would be as a flat bonus, like what was done with Molten Fury. The cooldown for combustion makes it slightly harder to work with, given that it is 3 minutes plus the time combust lasts, which is based on your crit rate. I'm not sure if it's linear, but you might be able to come up with a good approximation of the line I have for data points. Below are some data points, hopefully enough. The base crit rate is what the normal crit rate is, including gear and talents. The adjusted crit rate is the observed crit rate during combustion's effect. The average time is how long combustion would last. And the average crit bonus is the crit rate gained from combust multiplied by the percent uptime.

The base Crit rate was 0.2 and the adjusted Crit rate is 0.48728.
The average time per combust is 18.4699 and the average crit bonus is 0.0267346
The base Crit rate was 0.25 and the adjusted Crit rate is 0.516378.
The average time per combust is 17.4291 and the average crit bonus is 0.0235159
The base Crit rate was 0.3 and the adjusted Crit rate is 0.548141.
The average time per combust is 16.4191 and the average crit bonus is 0.0207427
The base Crit rate was 0.35 and the adjusted Crit rate is 0.58006.
The average time per combust is 15.5156 and the average crit bonus is 0.018257
The base Crit rate was 0.4 and the adjusted Crit rate is 0.61434.
The average time per combust is 14.6499 and the average crit bonus is 0.0161318
The base Crit rate was 0.45 and the adjusted Crit rate is 0.649407.
The average time per combust is 13.8588 and the average crit bonus is 0.0142554

As you can see, that ranges from 2.6% crit to 1.4% crit between a base crit rate of 20% and 45%. Of other interest, and not shown there exactly, the crit rate gained from combust is about equal to 35% of your non-crit rate ( .35 * ( 1 - baseCrit) ). It's a bit more variable than that, but is close enough to 35% in the region of 20% and 45% crit that anyone who cares will be close enough, since it isn't important info for the spreadsheet. I can give you more figures, if needed, but I think that's enough. Also, I ran this for 10000 combusts (easy since it's a simple program) for each crit rate.

I hope this is useful info to you.

 

◊ Vontre

Version 2.6 is out. In addition to a few fixes I added calculations for several trinket proc abilities, such as nexus horn, lightning capacitor, etc. New url for the download, www.radiationnow.net/wow/mage_dps.zip

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

 

galzohar

For meta you don't want to use 12 damage meta or all helms with a meta socket will simply suck (for their Ilvl and effort needed to obtain them). You should always use the 5% chance for next spell to be cast in 1/2 time meat or not use a helm with a meta socket. It means gimping gems, but usually not even close to what you gain from the meta. While it provides no efficiency, it's 2.5% faster fireball casting on average, which to me with my gear is slightly over 50 spell damage in terms of DPS. Much better than 12, although you do need to reduce a few points of spell damage (how much depends on your setup) due to meta requirements of more blue than yellow gems.

 

Suesse

Very cool spreadsheet!

Would it be possible to model the 1/2 cast time meta gem? Or does anyone have any pointers to a good working theory of how one might model this gem?

 

galzohar

It really depends on your spell. If you're just spamming 1 spell and ignoring all others, it's easy as hell to model:
1 in every 20 spells is 1/2 cast time (or 1.5s cast if it's limited by the GCD). Time to cast 20 spells is 19*castime+newcastime and the time it would've taken you to cast 20 spells is 20*castime. Speed increase equals old time / new time. For fireball it's 2.56...% and for frostbolt it's 2.04...% speed increase. Of course if you're using multiple spells that some don't get as much effect, it'll proportionally be used on both (if they're not instant) as just like timing clearcasts, lagX2+response time makes it impractical to stop your spell to cast another that would benefit from it and you're better off just finishing casting the spell even if it's a 1.5s AB that goes to 0.75 but is limited by 1.5s GCD and therefore gets no benefit.
Turning it to effective spell damage points is more complicated - you'll have to actually know how much spell damage you need to increase your DPS by 1%. For me that's 20.5 spell damage when fully potted and buffed for fireball.

 

Suesse

I think the modeling of the Icon is wrong. The gear selector has a value which includes the use effect, but you also model the use effect in the calculations tab. I believe the appropriate fix is just to have the gear selector use the passive value.

Your excel powers are very strong, I have no idea how half of this stuff happens.

For the meta gem, would it work to modify "Frostbolt Cast Time" in the "Game Stats" tab? What would be the new value? 2.95

Edit: Yes, that gave her 18 dps which is what she said it would be anyway. So, I think that's right, thanks.

 

Liebestod

Appreciate the update, the trinket feature is good, but it's missing a couple I'm interested in, namely Darkmoon Card: Crusade and Mark of Defiance. I can emulate the effects of the former by just adding 80 +dmg, I guess, but... yea.

Might also be nice to see a premade 33/28 or 34/27 scorch build.. maybe that's what 43/18 is supposed to be, but I've never seen that one before.

 

murgos

Vontre, is it me, or is the Lightning Capacitor not showing in your trinket menu? I have the most updated version of 2.6....

 

Zwink

There is an option to account for the 2 piece Tirisfal bonus (20% damage and mana increase on Arcane Blast), but what about the 4 piece Tempest bonus (5% damage increase on Frostbolt, Fireball, and Arcane Missiles)?

http://vodka-guild.com

 

Flick

I'm an excel noob, can I just type stuff in to make my own rotation, I've been using:
Scorchx5 to start (obviously)

Fireballx2
Fireblast

repeat 3 times, then once:

fireball
scorch (to reapply buff)
fireball
fireblast

This is an extremely mana efficient rotation, but raid buffed and with a shadow priest I'm able to keep it up and fireblast's DPS is higher than fireball by an amount noticeable enough to warrant adding it to a rotation.

That rotation is generally what I use on DPS heavy encounters in 25 mans, where I can count on atleast 1 other mage to help keep scorch up, since the rotation cuts it a little close, an unlucky resist + bad reaction time can mean having to reapply 5 in a single mage situation. The disadvantage ofcourse it its short range, but on encounters where that makes a difference I generally stay away from this rotation and go back to fireball spam.

EDIT: neveremind, I figured it out, but wouldn't mind discussing said rotation. According to the spreadsheet I'm gaining over 50 DPS with the gear you have there.

 

galzohar

With the builds being so close in T5 gear, I doubt losing either the set bonus or the stats on T6 (by sticking with T5), on top of gaining extra 5% frostbolt/fireball DPS (!!) would leave any point debating arcane vs fire. Unless the effects of intelect/crit difference are a lot greater and different than I though (which is very unlikely but I'll look into it when I have the time to see the breakpoint where arcane can be clearly better than fire if ever), fire should hands down be better than arcane with T6 gear.

 

Zwink

Well you would only be using 2 pieces of tier 5 the rest of your gear would be tier 6 equivalent.

http://vodka-guild.com

 

CHeeSY-CrAfT

Is the Shiffar's Nexus Horn being modeled on the old 15 seconds duration, or the newly fixed 10 second duration on the proc?

 

galzohar

Anyway everything I said becomes quite absolute after doing some math:

I calculated 3 specs: 10/48/3, 40/18/3 and 33/28 all for maxed DPS (no range for the arcane specs since you're in 30 yards anyway and getting more range on fire requires gimping DPS and/or DPM unlike with the 10/48/3 spec).
Spell damage was 1250 for all, int 500, crit on character screen was 30% for 10/48/3 (and calculated based on that the crit with 40/18/3 and 33/28).
For hit values, I did the math and needed to estimate the amount of spell damage you could gain by reducing your hit rating due to extra hit talents and assumed your hit with arcane spells is always maxed. I noticed that it's still worth keeping capped hit rating if the spell damage you can get per lost hit rating on items is a bit under 1 point per point, and I estimate the value in practice to be around 0.7, therefore hit rating is worth keeping at max with all specs. However 33/28 doesn't have elemental precision, and therefore needs more hit rating - so I deducted the appropriate spell damage at 0.7 spell damage per hit rating needed to get capped.
For AP I calculated as if you keep using the same rotation (as if you let arcane spec ever spam AB it will obviously win and no point in calculating anything, just pick arcane if AB spamming is possible at any point of the fight), I multiplied all spells dps and mana cost by 1.025 as well as adding 1 APed pyro and removing 1 APed scorch every 3 minutes.
Combustion was estimated as a 2.5% DPS increase as it's almost impossible to calculate.

Results are as follows:

10/48/3
Fireball spam: 1254 DPS, 10.5 DPM
8XFireball+1XScorch: 1242 DPS, 10.6 DPM
Any other rotation has too low DPS to mention.

40/18/3
3XAB+2Xfireball+scorch: 1305 DPS, 8.4 DPM
Any other rotation has too low DPS to mention.

33/28
3XAB+2XFireball+Scorch: 1324 DPS, 8.8 DPM
2XAB+2Xfireball+Scorch: 1269 DPS, 10.3 DPM
Fireball spam: 1260 DPS, 9.9 DPM
8XFireball+1XScorch: 1255 DPS 10.1 DPM
any other rotation has too low DPS to mention.

As you can see 33/28 comes on top in everything other than the range loss and slight-medium DPM loss depending on the rotation. Most efficient rotation that still does more DPS than the best of 10/48/3 is very little behind it in DPM.

What this really relies on: perfect usage of AB rotations so you never get the max mana cost and never get the max casting time, 2/5 T5 and never losing the scorch debuff stacking (you can easily swap the first 2 fireballs for scroches which is neglicible DPS loss if done once in the fight, but if you have to do it multiple times your DPS *WILL* drop noticeably as scorch is much weaker - notice I didn't include scorch spam DPS for that reason).

Other things to note:
-AP PoM pyro really tips the tides here (see my old calculations a few posts up). I did take into account the GCD it takes and the one less APed scorch you have. Granted it's still kinda rough but pretty damn close to reality.
-Increasing crit rate significantly (to 40% in the 10/48/3 spec and proportional increase for the rest) does NOT significantly tip the scales for the 33/28 spec. The fireball spam DPS increases by ~1.5% or so, but the best rotation of ABX3+FireballX2+Scorch actually LOSES ~0.2% DPS (does 5.34% more DPS than 10/48/3 compared to the 5.55% extra DPS it would do with 30% crit).
General scaling of specs: 10/48/3 scales faster with both spell damage and crit, but so slightly better than even by increasing crit and +damage by a huge amount 33/28 3XAB+2XFireball+Scorch still win by over 5%.
-Arcane spec gives 15% regen while casting - while not much considering the DPM is lower anyway, when you use the *almost-the-same* rotation DPS/DPM-wise it might be enough to make it more DPM on top of making it more DPS.
-Best bracers in the game until alar are "of fiery wrath" so if you want arcane you're forced to lose spell damage by using bands of nefarious deeds, but then again there's the new BoE epic that's slightly better than "of fiery wrath" anyway so I'll probably by that when I feel I'm anywhere near 2/5 T5 and speccing 33/28. For now, since it's nearly the same DPS and less DPM (not using AB due to no 2/5 T5) I'll keep the range of 10/48/3.

 

Roywyn

How realistic is it to get 7.5s cast time in between the two arcane blasts? I'm currently 40arc/21ice (for Al'ar), and never managed to get 3 frostbolts (7.5s cast time total) between two arcane blasts in a raid without teh debuff falling off before the cast. Yes, I do use stopcasting and Quartz for my casting bar.

So, anyone here who can squeeze in 7.5s easily and get his following Arcane Blast at 1.5s (1.8s) casting time?
And can possibly give any advice to people who don't? :o

 

galzohar

I need to give it a try and see I guess, thought it would be possible considering you have 0.5s to play with. Using AM can really tune down the DPS as it's the only possible way to get >=6.5s casting between ABs without going above 8. Only possible rotations are AB+scorch, 2xfireball+scorch and 3xfrostbolt. If lag is too much for you to lose more than 0.5s every 3 spells it may screw over the whole spec, then again if you lose 0.5s over 2 spells you could ABX2-fireballX2 and still cast the AB so it's finishing casting right after the debuff wears off... But I doubt you get that much delay when using /stopcasting appropriately.
Dunno how quartz works, I just time my /stopcasting according to the latency to which I got used to, or even by knowing how long it takes a fireball to cast and hitting the fireball macro exactly 3 seconds after I hit the last one. Again I'll try do some testing to see how much time you really lose.

 

Stein

I can't. I just use(d) 2xFireball, 1xFireball+2xScorch or 4xScorch.

 

galzohar

It took extreme effort, but I pulled it off. It requires completely perfect use of your /stopcasting and looking at nothing but your casting bar, but after a little practice I managed to pull it off a few times. Then again if you can't pull a scorch off and the AB that comes in its place doesn't need to be waited for, you lose similar DPS to what you would lose with any other spec without optimal use of /stopcasting.
Assuming you should learn to use /stopcasting perfectly anyway, the basic ABX3+fireballX2+scorch rotation still stands as the best DPS rotation. It just takes a lot of practice to master, especially in the heat of battle, I don't even wanna imagine how's that like with my 400 ping. But it's definitely very doable. And if you lose more than 0.5s between the 3rd AB to the next 1st AB, you can just do 3XAB+2Xfireball and your average DPS per time actually cating will actually increase, however your overall DPS will drop simply because you're wasting more time not casting anything, but that would be true for any spec when not using /stopcasting.
The thing about the 33/28 (or 40/18/3 although inferior) 3XAB+2Xfireball+scorch is that while it's doable, if you try it and mess up you lose a lot of DPS. Other specs you just lose the time wasted between casts. At that point it's impossible to calculate as the difference relies on your ability to keep up the rotation, not how good the rotation actually is.

Hopefully someday blizzard fixes all this dead time between casts due to latency issue and just let us que the next spell ahead of time and end this bullshit. My spells are designed and balanced to hit every 3 seconds, not 3 seconds + ping or 3 seoncds + your messup time with /stopcasting, which is total BS as it varies between people. I wonder how mages would fair in DPS if you could actually get every 3s spell casted with a 1/3s frequency...
Wishful thinking I guess, over 2.5 years of WoW and still nothing to adress this.