Kavan

I'm not using int gems just because I would have some strange idea that mana is very important. I based all decisions on sound analysis and optimization. Granted I made most gemming decisions before I resocketed to Relentless meta gem, but even now Potent Noble Topaz only cought up with Brilliant, not really surpassed it. For the subset of encounter parameters I deem important Brilliant Dawnstone scores 11.02 vs 11.16 for Potent Noble Topaz. So technically you're correct and if I made the gemming decisions today I'd use Potent in most yellow slots. But the difference is small enough that I'm not willing to resocket in case they nerf Relentless gem. If you're talking about using Runed Living Ruby disregarding the socket bonus that's not wise in most cases. It's value is 12.21. Some quick math can tell you that for example for Trial-Fire Bonus it's better to match colors to enable +5 spell damage bonus.

Ok let's talk about how important total mana is. Let's look at a 7 minute fight which is about middle ground. Looking at marginal value of different stats I get that 174 total mana increase results in 1 overall dps increase. You could say that this doesn't guarantee going for int. And you're true, this on itself does not guarantee it. But int also increases crit and spell damage. When you add all of this up you end with 2.68 int needed for 1 dps increase. On the other hand you need 2.61 spell damage for the same 1 dps increase. As I posted earlier, in general ratio of int to spell damage will range from 0.8 to 1.5 based on fight duration.

 

Vhad

Galz; I really don't get why you are so strung on trying to defeat everyones desire to play arcane spec. I've said it before, it's not some absurd pvp spec that makes you underperform in raids, and definatly isn't holding my guild back that I'm specced arcane. I more often the not outdps everyone/stay in top5 on all content that we do - would I do better with fire? No, I wouldn't, I've had it with fire spec, it bores me after 2+ years with scorch -> fireball spam. The differences I've found from ingame experience with 2+ months as arcane and 2+ years as fire are so miniscule and insignificant that it comes down to preference for me. Theorycraft should be treated like Religion, it should be a guideline and not the end all be all approach.

I suggest you go watch the Last Resort movie of their first Illidan kill, they had a 10 0 51 mage on top of dmg meters, do you really think he would've done better with a fire spec?
I'm getting a little more then tired of this fire or nothing attitude you present and really don't think it has any place on these boards - it's not helping discussion at all.

To add to the discussion; looking at your gear you have 4 items + trinkets that could be optimized for Arcane spec. Helm/Offhand/Legs/Ashyens Gift. And eventually 3 more replacing the spellfire set.

One thing that I think is a huge flaw in your approach to this is that; Stats matter. You can't ignore stamina and a large portion of your gear is opted to only gear towards dmg and crit and hit. I think you'll find that as further you progress in PvE the more valued Stamina will be. Picking up stamina gear usually also increases your intellect and spirit which in the end favors an arcane build as we benefit directly from those stats (both spec do, arcane just more favorable). I'll lose about 43 dmg going from full spellfire to my desired full T5 + belt from Anetheron, however I'll gain 50 something intellect, more stamina and spirit + the T5 bonus which will far outweigh the 43 static dmg.

Your note about AoE is surely not based on experience of any sorts. Arcane offers far better AoE for several reasons, we have better regen and a bigger mana pool, which in the end allows for more aoe. Arcane AoE consists of AE spam more or less, which by default causes less aggro so you can start earlier, it crits higher and with arcane potency it scales incredibly well with the amount of mobs that have to die. If there is any kind of AoE it again favors using items with stats on as a few HP here and there means you'll have a better chance of surviving.

To Aoife; That's not entirely true, with mind mastery initial mana is a result of your stats and the higher total you can get your mana pool the longer burn period you will have with AB for an arcane mage. With my wishlist items I will be close to 520 spirit evocating and near 13k mana buffed, that's alot of mana to burn.

What!?

 

Nyuu

If I could suggest either changing this thread or starting a new one to provide discussion on how to maximize DPS as arcane rather then another arcane vs fire thread. I see enough people with an interest in playing as arcane rather then fire despite the theorycrafting showing fire as the higher dps spec. At least for me, I will be staying arcane because it is more fun for me at the moment then when I was fire, I may eventually switch back but I mainly wanted a change.

I'm more interested in seeing discussion with mages giving advice and tips for arcane DPS/rotations/gear on different fights and encounters and how to maximize the spec then another min/max argument about which is top dps. After using arcane for a few weeks now in SSC and the Eye and comparing meters I've come to the conclusion i'm not being a liability as arcane vs fire, my dps is still very good, so the spec is useable for raiding. We've established that arcane damage does work, even if it's not the equal of fire in all regards and vice versa. So i'd rather focus on how to tweak it.

 

Aoife

I'm not saying ignore the spell damage bonus, I'm saying gem for damage/crit, not intellect. I'd say that base mana is more important to fire mages than deep arcane mages. This is because other than scorch spam, fire mages have no way of tailoring their spell rotation to accommodate different remaining mana in order to conserve more or dump it for more dps.

 

◊ manly

Well, first thing first.

If you need to attack your interlocuter in a conversation, you know you're not talking about the topic at hand. While I may frown upon some gemming practices, it doesn't mean anything more than me just not agreeing with the choice.

Personally, there are few guideline I use to balance my gear. First and foremost, I prioritize damage over stats for the simple reason that having more spell damage/crit/hit means I get better use of my mana, whereas the reverse is not true (since not all specs can realistically go OOM at will). A long time ago, I used to give some 'value' to things like int and spirit because shadow priests didn't exist back in the day. Now that they do, and that realistically I always have one when I raid, I have come to realize that, as fire spec particularly, I cannot go OOM. What I am trying to say is that shadow priest have made spirit totally obsolete as a raiding stat for a mage. Intellect is very very close to obsolete for the same reasons. You could argue that if you play arcane spec, and thus, have the ability to go OOM at will, you will be able to make use of your extra intellect/spirit. I will also make another bold statement; shadow priests are the true source of an arcane mage power, without them, arcane spec would not really exist as a raiding spec. As such, if anyone dares mention that they don't raid with a shadow priest and still top meters on an arcane mage, I will simply call bs on the matter. I know that arcane spec has the best high DPM/low DPS rotations available, but under *every single fight in TBC (yes, everything up to Illidan), there is not a single encounter where that 'highlight' of the arcane spec remotely becomes important. So please, while I do enjoy speaking of theorycraft all day, lets be honest; if were speaking about raiding, and were speaking arcane mage, were talking of high DPS rotations. Nobody realistically cares about low DPS/highDPM rotation if you want to speak purely about 'the best' raiding spec.

While I was personally expecting too much out of arcane spec, fact still is that even someone with relatively no experience with arcane spec I was still able to give not too bad performance with it. I have came to realize that, to me, the downfall of arcane spec is that as soon as your end up with a boss having a task that isn't dpsing, it forces you to break your spell rotation, which arcane is not able to perform efficiently (or at least - unable to do with any rotation including arcane missile). This is the reason I specced out of it. If you want to play an arcane mage, you need to plan ahead your spell rotation and plan ahead your movements for them to coincide with your spell rotation. This means, if you just casted 3x AB, at that point you cant move for 'too long', all you have is about 2 seconds to move, or polymorph, or something. What you CAN DO is stop after 2 AB, do your 'other non dpsing action', then resume your rotation, and its going to work. Some people may appreciate the challenge; I didn't. I think it's counter-productive to what my end goal is.

So, with all this said, since I assume that shadow priest will give you a lot more mana than any amount of gemming or +int you can fit in your gear, I figure its probably better off making all that sweet mana give you more bang for your buck. You *could realistically start a debate on whether that int gear will give a better output than the replacement damage gear, but then i'll point out that the solution is scaling out with consumable usage. Mana potions will give you far far more of a damage boost (assuming you can use that mana of course) than any available potion (destruction pot). I think personally if you need to use int gear, then youre probably better off using better potions (or more often). If that still isn't enough, then I have nothing else than telling that person that no matter what the spec, you need to be able to evaluate at some level your DPM and plan at least a bit ahead of time.

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.

 

Vhad

I'm not sure this goes in here or just in the general mage theorycrafting, anyway: IF you have 2 set Tirisfal, how viable is it to switch to full out AB nuke sub 20% for a 10 48 3 specced mage? We often see fire mages saying that mana is NP and sometimes end on full or near, have anyone tried to model this?

40% permanent extra dmg to AB seem's incredibly powerful, unless ofc all your gear is +fire dmg.

What!?

 

◊ manly

Good question, although I will make a minor note that AB spam doesn't work too good with combustion. Not that it stops you from ab spamming regardless.

And while of even smaller note: most gear is fire and arcane damage The only exception I can think of, besides offhands that can be swapped in combat, is the hellfire encased pendant. But 50 damage is not what will make or break an arcane spam.

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.

 

Vhad

If it works on any kind of level it could be the answer to what the last 3 talent points should be used for in the deep fire spec; 13 45 3

What!?

 

Copernicus

My back of the envelope calculations say that Arcane Blast spam > Fireball spam for the last 20%, but it's very dependent on lag. With zero lag, Arcane Blast was worth about a 200 DPS increase. With 0.2 lag, Arcane Blast was worth +100, and with 0.3 lag Arcane Blast was down to +60 DPS or so. That assumes the mage is using a stopcast macro, not round-trip lag. I also didn't include rotating in a Fire Blast for the Fireball spam.

 

dragnork

I had been thinking along the same lines and rather than using an AB spam at the end during the time when molten fury is active, the more logical thing would be to start off with AB spam. Given the inherent threat reduction there shouldn't be any aggro problems and you could spam it till you could refill your bar with a pot/gem and evocate. This should allow you to be able to sneak in a second evocate too if needed on the slightly longer fights.

This is quite similar to the problem with an arcane spec chosing multiple cycles with varying dps to achieve maximum throughput. Except that you are using fireball and AB cycles.


P.S. A mod that could obtain raid dps from a damage meter and boss health and could calculate

a) number of spells of a specific tree that you could chuck till the boss dies given current damage inc to boss and his health.
b) the number of spells you could cast given your current mana

With such information it should be quite possible to switch cycles dynamically depending on the situation.

 

JasonX

Thats an interesting thought. I've never thought of using pure AB spam for the last 20% HP as a deep fire mage. Well, I'm playing with some 500 latency and is using the /stopcasting macro, but I still think it will increases my dps more than fireball spam (provided combustion is not up). I've converted to arcane spec after getting 2/5 T5 recently. Will probably spec back to deep fire if I can't adapt to arcane spec.

 

Vhad

How exactly do you use the stopcasting macro? Just add /stopcasting prior to your spell and count when your spell is done realtime compared to your castbar?

Edit: And a note to arcane spec, if you use any castbar addon that count's in milliseconds too you can cancel your AM waves a 0,9 seconds to finish since it fires off 5,0 4,0 3,0 2,0 1,0.

Edit2: I'm blind, its more like 4,5 3,5 2,5 1,5 0,5, point still stands though and it works pretty well.

What!?

 

Kavan

You use /stopcasting before the spell yes. In addition it's highly suggested to also use a casting bar such as Quartz that displays the latency so that you have a dynamic visual indicator of when to start the next cast.

 

galzohar

Pretty much, yes. and for 1.5s cast spells it's even easier because the client has an inherent 1.5s timer called the global cooldown ;p hit your spell every time your global cooldown is up if it's 1.5s cast and you'll never be disappointed.
As for quartz etc, I don't find a need for one. I know my ping and how stable it is and have a timer on my castbar that shows how much time is left on my casts. If my ping is 0.3 for example, I'll hit it when my bar shows less than 0.3s. After a while you get used to it to the point where you don't even need to look at your screen to do it decently, although it'll obviuosly give better results if you do nothing but watch your castbar. I dunno how quartz adjusts to ping stability, but that may be another thing you need to take into consideration. If your ping changes from 400 to 500 and back consistently, you have to play by a 400 ping or you'll end up canceling spells (or triggering global cooldowns right as you want to cast a spell, as there's a bug that sometimes your client will not show you casting, and trying to use the macro at that point will trigger a global cooldown but not do anything else - your current spell will still fire off though).

As for full arcane, I already assumed all your +hit is converted into spell damage and arcane missiles still did terrible dps. The only arcane spec I'm considering is one that uses fire spells as fillers between arcane blasts, and if you scroll a couple posts back you'll see what breakpoint I found for 40/18/3 VS 33/28 specs (all with 2/5 T5 of course).

 

• Arawethion

I believe that AB is the right thing, theoretically, to dump your mana into as a boss is dying, as 10/48/3.

The issues are Combustion and lag, as have been noted above. Also, if Heroism is being used sub-20% (as is often correct for Mages), it's easy to lose a lot of the theoretical damage output of AB spam with imperfect timing.

Answers to Moonkin questions (pre-3.3):
0) Read the TTT/use the spreadsheet: http://elitistjerks.com/f47/t66856-moonkin_pve_dps/
1) Maintain high DoT uptime. Use WiseEclipse.
2) Nothing beats 2T8.
3) Yes, sometimes you cast many Wraths and no Eclipse procs. Deal with it.

 

Vhad

If theorycraft puts AB > FB spam sub 20%, couldnt you just opt for using Heroism earlier to get it to 20% even faster?

What!?

 

galzohar

If AB spam is superior sub 20% it is also superior at any other %. Of course if there's no heroism factor you should obviously use your highest burst after 20% rather than before for better overall DPS, but if increasing the effect of this burst means you decrease the effect of heroism by using it earlier, depending on how big of a DPS boost AB spamming is you may be better off using AB before 20% and heroism with the remaining mana after 20%. Will take more planning a head of course, and it *really* depends when is better time to have 20% more damage during AB or heroism. Since fireballs with heroism almost certainly do more DPS than AB spamming you're probably better off with the heroism. Remember though for saving heroism, if 1 person in the party is dead saving that heroism was actually less beneficial than using it early. 20% more effect is great but when it gets 3/4-4/5 of the effect because 1 person is dead by the time you use it makes it not worth it.

 

• Arawethion

Right, but using AB at any time other than the very end of the fight might not be realistic--it assumes you know precisely how your mana usage is going to turn out.

Answers to Moonkin questions (pre-3.3):
0) Read the TTT/use the spreadsheet: http://elitistjerks.com/f47/t66856-moonkin_pve_dps/
1) Maintain high DoT uptime. Use WiseEclipse.
2) Nothing beats 2T8.
3) Yes, sometimes you cast many Wraths and no Eclipse procs. Deal with it.

 

Patterns...

I think using AB spam above 20% as 10/48/3 is never going to be worthwhile; trying to optimize heroism/combustion/AB-spam-mana-dump use to that extent isn't realistic. The real question is (as you said, I think) whether or not heroism/combustion pre-20% with AB mana-dumping at the end of the fight is a net benefit over normal 7:1 fireball pre-20% with molten-furied, combusted, heroic (?) fireball spam sub-20%. My intuition says yes, but I don't trust myself to do the math here.

edit: too slow.

 

Hephaustus

Just responding to what someone said about Last Resort's Illidan kill video. When you have multiple frost mages with water elementals, your damage is way inflated because water elementals cant be distiguished between by parsing, so ALL water elemental damage is added to ALL mages, severely inflating their damage even though they may be way lower in terms of actual damage done.

 

galzohar

If you heroism under 20%, most bosses will die about when your heroism runs out, which means you won't get much AB time anyway. it really depends on the fight and your buffs at the end of the day regarding AB timing, but more often than not it's not gonna be worthwhile. If it's not worth it before 20%, it's not worth it after either.

With my gear I get 1351 DPS with AB spam with 0 latency and 1294 with fireball spam and 1285 if I have to keep scorch up for myself. The increase from using heroism under 20% instead of anytime is like increasing the 30% bonus by 20%, so total 6% DPS increase during heroism if you use it under 20% instead of any other time. That's 77 DPS (6% of 1285), or 1362 DPS total, which beats the increase from AB spamming, which much better mana efficiency. So as a 10/48/3 I would only AB spam when the fight is near over and heroism is already used and I still have mana left to burn, and even then, while I don't have it on my current spreadsheet, I remember fireblasting every 2 fireballs (1/3 imp fireblast) increases DPS by ~5% which would also bring you on par with AB spamming, while not reducing mana efficiency nearly as bad as AB spamming. That works both b4 20% and after, the only difference is that if you only have mana to burn for a short period of time, it's best to be done under 20% obviously. Remember though that fireblasting with heroism up is a DPS LOSS, as well as AB spamming for 10/48/3.

 

songster

That's a bit dim, then. Why doesn't it divide all elemental damage equally between all elemental-specced mages? Still wrong, but not quite as egregiously.

 

Vhad

They were only 3 mages in that video, and I really doubt that even if they got dmg from all that Kerp would've gone from first to anywhere not acceptable, and I doubt he would be much higher with fire which was the entire point.

@Galz; How do you define much AB Time? 40 seconds of AB is a long time if you compare it to fireball. With 0 lag you get 19 fireballs compared to 13 w/o heroism, while you get 26 AB - and you don't have to keep up scorch or anything you can just sit there and spam AB.

The entire point was that Fire mages keep saying that Mana is no problem and can end fights with almost full; How about taking advantage of that, adding to the equation they could get 2 set tier5 which will sub 20% increase AB dmg by 40%.

I'm not sure if you counted 2 set bonus or not (or even used molten fury for AB?), if not it's quite evident that it will work to use heroism before 20% and then dumb all your mana for AB sub 20% for the fat 40% increase. 1351 dps seems very low for AB spam even without 2 set bonus btw.

What!?

 

MacrosTheNaked

I've got one of the stranger specs out there I'm guessing, where I'm arc/fire, but spend 43 in arcane and 18 in fire, which enables me to get both mind mastery and imp scorch.

Ideally I rotate AB X 3 + AM + Sorch as this is the best DPS/DPM combination I've found.

However, there's several alternatives, Scorch spam till clearcast + AM is the most mana efficient one and makes me go on for hours, but a pure AM spam (rank8) is also highly effective, but only if you've got the most OP trinket on the planet, the lightning capacitator.

 

galzohar

Had some calcualtion error, AB spamming is actually 1678 DPS. Much more worthwhile. Defnitely more useful than bloodlust *if* you have the mana. However at the end of a 10 minute fight I don't think you'll have mana to anywhere near 40s of AB spamming, but again it's very very fight and group setup dependant. All in all you will probably rarely have mana for AB spamming for full 20%, especially if you had a heroism or 2 in that fight. Also remember that if you only have mana for, say, 20s of AB spamming, the "chargeup" time also reduces the overall DPS of AB spamming as it's not neglicible when your entire AB spamming period is too short.

Worth looking into depending on how much mana you have at the end, but be careful not to end up gimping your DPS instead of buffing it. Of course with 2/5 T5 you'll probably do more DPS as 33/28 anyway.