Mojache

If you guys think AM is such a bad spell you might want to relook how much haste gear is being put in Sunwell. I haven't decided if I plan to build for a 2P T5 with 4P T6 but is something that looks like a very possible combination. Need to also note that currently you can't use a flask of light and sunfire enchant with arcane and frost damage, small difference but it's things like that make the difference.

I really hope they put the 2P T5 into the arcane talent tree to better make the spec. It is quiet fun to play managing your mana and added a different element to the spec back in the day. Either way I'll be most likely Fire or Frost for learning sunwell so I can save my pot/stone timers for Hp consumables while we work on progression. Kalecgos is a bad mofo.

 

Pintofbrew

How much haste is in Sunwell is irrelevant mojache. The fact remains AM scales badly with... Everything. There are reasons we're arguing AM rotations vs Fbolt or Fball rotations in arcane, but the loot in sunwell is not one of them.

I assure you, we all have seen the loot list in sunwell.

Vontre: Could you elaborate on this int "sweetspot"? I'm more intrigued with why it drops sharply than with where it is. It clearly isn't a case of "too much mana, AB spam >100% impossible" so I'm wondering if there's some strange limit where the ratio of damage from base spell to damage from spell power becomes such that dropping more spell is distinctly inefficient.

Perhaps it's even due to diminishing returns if you will from Int in terms of mp5.

 

Alvira

I am curious why there would be a sweet spot also. Everything scales (pretty much). So if Int was the better stat earlier on, it should stay good even pst 700 int. And as Pintofbrew said, there isn't any cap that is being reached at 700 Int like too much mana or something. So, theoratically, there shouldn't be a drop off.

Mind mastery, increase in crit ratings, more mana for AB spam, further increase in MP5 from spirit. All these doesn't cap or stop as you stack Int past 700. If big sacrifices had to be made such like sacrificing a ton of +dmg, crit just to raise Int by a bit. Then I would understand. But if we are just increasing Int while keeping the other stats consistent, then all the benefits that it gave before 700 Int should still apply.

Its like how increasing +dmg is always good. There shouldn't be any kind fo drop off at all.

 

Alvira

Great job Vontre, I applaud your efforts in trying to get up a working model so quickly.

I am assuming the changes you mentioned and the change to the spell routine has been put in and all I have to do is refresh the site?

Based on vontre's changes, I checked the mana consumption again. It still seems a bit off. For example, since all 3 arcane specs are ABing until OOM in phase 1 with gems, potions and evoc. Their evoc should happen at nearly the same time. But based on my inputted numbers for my current gear.

The 40/18/3 spec has me evocating at the 298 second mark (nearly the end of the fight). This is amazing and quite impossible because it means I am spamming AB for 5 minutes? Hmmm.

The 48/0/13 spec has me evocating at 225 seconds. This is a lot later than the first arcane spec. But it also means essentially I could have AB the whole fight. And why would there be such a difference? Both specs have the same rotation in phase 1. Just AB all the way. There shouldnt be such a big difference in the time they run OOM and need to evocate. Possibly, icy veins came into play here. But Does icy veins have such a big impact that it causes this spec to run OOM and need to evocate a full minute earlier?

The last spec 40/0/21 is even odder. This spec runs OOM by the 60 second mark and needs to evocate. This is a massive difference between the first two specs for essentially the same routine. Just AB spam in phase 1. No doubt this spec has more icy veins it can use. But is the use of more icy veins so powerful that it runs this spec OOM and needing evocate a full 2 minutes faster than the first two specs?

Furthermore, it goes into negative mana later!!! Because it runs OOM by 100 second mark, and it then goes into negative mana territory. It should have switched to the high DPM mode by then, but it didn't and instead, when into negative 6000 mana which is impossible.

In conclusion, the first two arcane specs lasted so long till evocate that it seems like they can AB the entire 5 minute fight. Which shouldn't be possible at this stage even with the buff. The last spec actually goes into deep negative mana territory which seems off also.

My inputted stats called up from my current gear:

Int: 595 Spirit 321 fire dmg 1024 frost dmg 974 arcane dmg 1024 spell crit 293 spell hit 114 spell haste 0

Trinkets: eye of mag and serpant coil brand.


Another comment. Is it possible to add flask of distilled wisdom to the selectable buffs? If Int scales better for arcane spec, perhaps that flask (+65 int) is a better choice than the +dmg flasks.

Once again thanks Vontre for all your efforts, great job so far. And I am sorry that I am raising so many issues. As I said, its not easy to simulate a well played arcane spec.

 

Kavan

Actually the best consumables for arcane are Adept+Draenic, Brilliant Wizard Oil and dmg/spi food.

 

Alvira

True, and I use that for farm fights. But for progression fights where you expect to wipe a lot, flasks will still come in handy.

 

galzohar

As far as I understand the formula for the regen from int is proportional to the square root of your intelect. That means the extra regen gained by additional int goes down the more int you have.

Basically going by first order for small changes of int:
regen increase =~ change in int * someconstant / int^0.5
That means the more int you have the less regen you gain from additional int, which means with enough int spell damage will start beating it out. This will happen faster than the linear stats obviously, as we all know that with enough spell damage crit becomes better but that amount of spell damage isn't realistically attainable. Since int regen isn't linear reaching the point where it drops off enough to be weaker than other stats is totally realistic, especially when with current itemization int levels int is worth getting.

 

Tharia

Might be those manarestore weapon oils everyone was complaining about might come in handy too, especially if they work with every AM hit and provided they don't have an internal cooldown. I know the text says on spellcast and not spellhit, but blizzard was never really good at coming up with consistent tooltips :x
Did anyone test these yet?

 

Flinx

I do really like the casting Simulator, however it would be great if one could pick different gear for different Talents.
It is easy to swap hit for spelldmg with 2.4 gear.

How would different Metagems affect Arcane

Insightful Earthstorm Diamond
+12 Intellect & Chance to restore mana on spellcast

Ember Skyfire Diamond
+14 Spell Damage & +2% Intellect

Chaotic Skyfire Diamond
+12 Spell Critical & 3% Increased Critical Damage

 

galzohar

Since the simulator is probably already doing an average of many results, can we have it calculate their deviation as well? Using the formula:

{[(x1-avg)^2 + (x2-avg)^2 + (x3-avg)^2 + ... + (xn-avg)^2]/(n^2-n)}^0.5

with avg = (x1+x2+x3+...+xn)/n

After all in order to verify the simulator is giving realistic results in a test (which people will most likely perform) you have to know the expected deviation. Say I do on a 3 min fight 1000 +-200 dps, if I get 800 dps I'll be calling the simulation faulty if I wouldn't know there's a +-200 deviation, since my results would be 20% off of the expected value, while in reality getting a 20% off result is quite likely without any flaws in the theory.

 

Kavan

From my understanding this is not a simulator in a usual sense. It basically simulates imaginary average spells, so each fireball hits for the same given the same buffs, all the variations are averaged out. This way you'll get the same results every run.

 

eaglesrock

since i looked for this in a few of the mage threads and couldnt find the answer i have a few questions regarding the 40/0/21 spec.

1. What spell rotation do people use when they arcane power? 3x ab 3x fb or all ab or all fb?
2. What spell rotation do people use when they icy veins?
3. Do people use icy veins and arcane power together (which seems logical)? In my own testing it seems that i cant get the ab debuff to wear off using a 3x ab 3x fb rotation. do you just waste your mana and ab spam or do you frostbolt spam?

 

BrTarolg

Due to the mechanics of arcane power, and even though earlier its been said haste is near useless, that doesn't mean you can abuse AP+IV+(bloodlust? Does it cap here?) So you can get as much out of the 30% extra DPS as possible (and ofc, dont forget that the mana cost is only 30% of base, not including debuff.)

 

Roywyn

You should pair AP with trinkets and haste.
AP under bloodlust for 15 seconds means that you have the same effect as AP without bloodlust for 19.5 seconds.

Haste for AB spam caps at 50% (1s cast and GCD). 150%/130% = 115.4%. So if you have +15.4% haste (~240 haste rating), you'd cap with bloodlust.
With the haste on sunwell gear, it seems that you can reach that point without trying.

Bloodlust and Icy Veins is 130%*120%=156% haste. Above the cap for AB spam.
If you do BL and IV afterwards, you have 40s of 130% and 20s of 120s cast speed.
If you stack them, you have 20s of 150%, 20s of 130% and 20s of 100% speed.
So, stacking/delaying BL/IV breaks even at zero passive haste. If you have passive haste, you should split BL and IV, if you have curse of tongues on you, you should stack them.

 

jogjog

I'm not 100% sure but i averaged "must have" a passive haste of ~225 to make the following cycles:

without IV or Bloodlust : X AB - 3 frostbolt

with IV: X AB - 4 frostbolt

with BL : X AB - 4 fb


Because we cant substain an AB spam durring the whole IV and bloodlust, because loosing the "haste debuff" of AB is a waste...

What do you think about that?

 

Docjowles

Am I reading the Cast Sequence correctly for arcane specs? It looks like Arcane Missile spam, with the occassional string of Arcane Blasts. Are you saying that AM actually is the optimal nuke in 2.4 for deep arcane? Or just that you haven't had time to model an AB rotation?

 

Pintofbrew

Jog, the whole point of rotating AB with anything is to Drop the AB buff because it's unsustainable

If you do not want to drop it off but want to lower your mana consumption there are a million other ways other than getting 225 haste, which will cap you for AB and thus be wasted i-points. If for some bizzare reason you insist on not dropping AB debuff you can for example rotate FB/AB or FB*2/AB*2 or any other combo other than getting enough haste to cap one spell and only benefit the other.

For the life of me I can't think of any other reason to instigate rotations which drop AB (or indeed mix it) for anything but economy and not dropping it is the epitome of bad economy.


Edit: Doc: No, rotate AB/AM/AB/AM/AB/AM/Sc ( I shorten it to (AB/AM)*3/Sc )with AM on clearcast. This rotation is under examination as to whether it's superior to the old AB*3/AM/Sc and it definitely -can- consistently AM on clearcast. It's not clear yet if it's superior to 3*AB/3*Fball with 40/18/0 or 3*AB/3*Fbolt with 40/0/21. The goal however is to maximize AB spam time, the rotations are simply fillers.

Edit II: I just realized you might be refering to Vontre's simulation, in which case disregard what I typed.

 

Docjowles

Yes, I was talking about the simulator. The cast sequence for deep arcane is literally just AM spam, which seems a little odd given Votre's undying hatred of the spell.

 

jogjog

I meant to partially drop it:

AB, AB(0), AB(1), fb*3, AB(3 - base cost), ab(1), ab(2), ab(3), ab(3), ab(3), ab(3), ab(3), ab(3), ab(3) ....
------------------------------------------------------------------------| IV, cold snap, trinket, AP*************|

, fb*4, AB(3 - base cost), ab(1), ab(2).....,fb*3, AB(3 - base cost), ab(1), ab(2), ...
|- IV ***************************....|----------------------------------------- ...

 

eaglesrock

this is all under 2.4 assumptions.

What about 2.3?

and if i am interpreting you correctly you are saying to AP + IV + Arcane blast spam and just deal with all the mana consumption?

 

Hibbo

Hi Guys what do you think about my Spirit/Int Mage Build? (40/0/21)

chardev.org - Template 23084

The calculated MP5 Reg you can find here:


There i have 641 I5SR from Equip,Buffs,Elixir, for Mana Potion and Stone i add 2 times x 95,8 mp5
and a shadow priest i calculate with 300 mp5

This will give my Spirit/Int Mage

1132 mp/5 I5SR and
1502 mp/5 OO5SR

The Usage of pure AB Spam needs about 2240 mp/5

2240-1132 = 1108 when AB Spaming

This will give us 65 Seconds of AB Spamming till OOM

then our Spirit/Int Mage uses Evocation for 10 Seconds (T6 Bonus) restoring 75% of the Mana and 1 Time I5SR and 1 time OO5SR mp/5 Tick what will give us ~13500 Mana from 14501

next AB Spam till OOM

now we get 1 Innervate what will give us ~20K Mana and in my opinion also more than the 14501 Mana we have, because we try to burst our Mana out in the time of the Innervate ticking with AP and maybe IV when it will lower the GCD for our AB Spam.

so we get about 3 Minutes of only AB Spamming

For the Time beeing OOM we can use Frostbolt for lower DPS but restoring about 500-600 mp/5 for filling up the Mana Bar. Or even use Frostbolt earlier to have full Mana for the last 20%.

Doing 50 Seconds nothing is the same like Evocation with this Mage ............

I hope you have some ideas to enhance this Build, or an other opinion so please let me know.

Sorry for my bad english I´m still learning

 

◊ Vontre

Haste is bad for arcane because it reduces effective dpm. Yes, reduces, you get less spirit regen per spell cast. Your efficiency cycle sucks anyway so improving it is not really effective. What you want is more AB spam.

Every stat scales against other stats, intellect and spell damage are no different. Since they are so closely the matched, the correct ratio is actually achievable. Also intellect mana regen has diminishing returns due to a square root formula.

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

 

◊ Vontre

It should be at very least an AB/AM mix. Post your stats so I can examine?

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

 

◊ Vontre

By the was the simulator hasn't been using a rotation since yesterday. I wrote a new logic method "arcane3" for the arcane specs. Basically it works like this. Pick a filler spell: fireball/scorch, frostbolt, or arcane missiles. If arcane power is on, spam AB. If you are low on mana, spam the filler spell. If the filler spell is arcane missiles, alternate between AB and AM instead of just spamming so you can pick up full spirit ticks. There isn't any logic for a switch on clearcast, however I imagine this would be smart if you are doing AM at any time. +1 for arcane missiles.

I would say the full arcane version (arcane/arcane lol) is preferred now. If you are consistently getting CoE fire is still better overall, but it's a close call and arcane does have more throughput control so I may think about speccing it... if I can ever get my hands on another T5 piece. -_-

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

 

Docjowles

It would probably be easiest to just pull my info from the Armory, but i'll attach a pic with my settings.

Edit: It looks like it's partially because I set the encounter duration to 600 seconds, which is probably a little high even for T5 (with a few exceptions like Al'ar). If I set it down to 300 I get at least some AB's mixed in, though it is still mostly AM spam.

Edit II: Yeah it was a function of mana regen and encounter duration, things look more normal if I tweak those.

Attached Thumbnails