Stacking Haste as Arcane

[rixa]

Haste is obviously the most valuable stat for an arcane spec mage (aside from spellpower). one of my guildmates is stacking his haste at the expense of other stats (crit and spell power). Currently he has 963 and says that he expects incredible results once he reaches 1000+. I always thought that the best dps resulted from favoring the most valuable stat but not stacking it to the point that it is detrimental to other stats i.e. having sp/crit/haste somewhat balanced (assuming hit capped). I'm not sure how to prove this

[Raistline]

I always thought the cap for haste was 720 something?

[rixa]

I always thought the cap for haste was 720 something?

This is the cap at lvl70 pre-wotlk i believe

[Malentra]

Haste is obviously the most valuable stat for an arcane spec mage (aside from spellpower). one of my guildmates is stacking his haste at the expense of other stats (crit and spell power). Currently he has 963 and says that he expects incredible results once he reaches 1000+. I always thought that the best dps resulted from favoring the most valuable stat but not stacking to the point that it is detrimental to other stats i.e. having sp/crit/haste somewhat balanced (assuming hit capped). I'm not sure how to prove this

The value of haste increases with your spellpower.  Therefore it never really makes sense to sacrifice spellpower to gain haste until you reach the value of spellpower that makes a point of haste rating worth more than a point of spellpower. 

The only real way to tell if you're at that point is to use simulation tools like Rawr.  Select your typical buffs/debuffs/fight length in rawr, load his raiding gear profile from armory, and make some of the equipment trade-offs that he has recently made to increase his haste and check to see if they are actually a dps increase or decrease.

What you're trying to prove seems like it would be simple to show, but its complicated because of the more complex mechanics in WotLK and because values of every stat change as you have different values of each stat.

Also the ratings come into it in a big way.  If you have 20 points, and could spend them into increasing damage and increasing haste.   Say you put 10 in damage and 10 in haste.  You now do 1.1 x 1.1 = 1.21 or 21% more dps.   If you'd put them all in haste, you'd do 1.0 x 1.2 = 1.20 or 20% more dps, so splitting them netted you more dps.  But say  haste cost you 0.8 rating points, and damage cost you 1.  Then if you put them all in haste, you'd get 1.0 x 1.25 = 25% more dps, and stacking haste is all of a sudden more useful than splitting points evenly between damage and haste.    Its obvioulsy more complicated than just that but you get the idea.  (for example, if we introduce mana and evocation downtime into the equation)

[Tronacus]

Stacking that much haste will just drop ones DPS. To be able to stack that much haste you have to give up to much hit, spell power and crit.

Even if he is just a little bit under hit cap the chance of him missing go up do to him casting that much.

Also to get 1k haste he has to give up over 500 spell power and 7% crit (278 crit rating) just to get 450 haste.
the loss in stats alone will lead to dps but giving up 4T7.5 will lead to even more DPS lose.
He is also making thing like BL and IV useless.

[Dibbles]

There has always been a "soft cap" for certain stats and stacking this much haste in my opinion will be going beyond that cap. I can imagine that he is going to have some mana problems even with this content, let alone 3.1. Sustaining a high AB spam cycle and using AM proc when up is argued to be the highest dps, provided mana, and he would simply be using more mana and hitting for a lot less while waiting for AM to proc. I personally believe balancing haste and sp is the way to go something around 500 haste is probably enough, arcane gets lots of + % to crit from talents so i have never really worried about crit as arcane.

[c4tuna]

There has always been a "soft cap" for certain stats and stacking this much haste in my opinion will be going beyond that cap. I can imagine that he is going to have some mana problems even with this content, let alone 3.1. Sustaining a high AB spam cycle and using AM proc when up is argued to be the highest dps, provided mana, and he would simply be using more mana and hitting for a lot less while waiting for AM to proc. I personally believe balancing haste and sp is the way to go something around 500 haste is probably enough, arcane gets lots of + % to crit from talents so i have never really worried about crit as arcane.

I'm sorry but you don't understand how soft caps work then. Soft caps work for things like hit (where you cap yellow attacks but not white ones) and defense (where you cap crit mitigation but not parry/dodge). Haste doesn't, unless you do things like operate filler spells in between other cooldowns (like getting enough haste to squeeze 2 mind flays in between each mind blast cooldown). Arcane mages don't have something like that to the best of my knowledge.

Although you should definitely run a spreadsheet (available at elitistjerks.com) to determine when crit, etc. beats haste.

[Malentra]

I'm sorry but you don't understand how soft caps work then. Soft caps work for things like hit (where you cap yellow attacks but not white ones) and defense (where you cap crit mitigation but not parry/dodge). Haste doesn't, unless you do things like operate filler spells in between other cooldowns (like getting enough haste to squeeze 2 mind flays in between each mind blast cooldown). Arcane mages don't have something like that to the best of my knowledge.

I would imagine an Arcane haste "softcap" (or essentially amount at which a given stat drops signifcantly in value) would occur when you've GCD-capped (1.0s cast time) major spells in the dps rotation.  However, since ABar is no longer a necessary part of the dps cycle (although it can be), ABlast is our first major haste-capped spell.  It requires 150% total haste to haste cap ABlast (or >>1600 haste rating while under both BL and IV, and >>4500 haste rating without any external haste boosts) , which I think he's safe from doing as long as he's under a AB3AM cycle.

[Dibbles]

For fire mages in BC there was a "soft cap" on stacking crit because after a while getting more sp or haste was more valuable, that is what I was talking about. It is the same concept for many classes. I am not referring to an actually cap of haste, but a cap at where haste becomes no longer the most valuable stat it has often been called by some people a "soft cap".

[Malentra]

For fire mages in BC there was a "soft cap" on stacking crit because after a while getting more sp or haste was more valuable, that is what I was talking about. It is the same concept for many classes. I am not referring to an actually cap of haste, but a cap at where haste becomes no longer the most valuable stat it has often been called by some people a "soft cap".

I doubt that's the intended definition of softcap.  A softcap is esentially a certain number when a stat diminishes in value, but not to 0. 

For example, hit is *hard-capped* for casters because its value goes to 0 when you reach 17% total hit.
Meanwhile hit is *soft-capped* for enhancement shaman at 17% total hit, because it still has *some* value after this mark (because the dual-wield melee hit cap is higher).

Another example is the softcap on crit for Frost mages.  Normally the hardcap for crit is 100%.  But because of Fingers of Frost and Shatter, when a Frost mage reaches 50% crit chance, he essentially hits the 100% hard-cap when FoF is procced.  Therefore crit rating is more valuable up to the point that the frost mage reaches 50% crit chance raid buffed, then diminishes in overall value after that point.  (but not to 0, since it still increases crit chance when FoF is not procced)

A *cap* is more about the mechanics of how a stat works in relation to your class/spec abilities, and less to do with when its value is surpassed by other stats due to your gear item level.  I think you are confusing "optimal amount or range" with "softcap".  There was probably an optimal range for crit% in TBC, but it certainly wasn't a softcap.

[Caeldeth]

Order goes hit until capped, sp > all after that.... to sacrifice even spell power for haste is a loss in dps, and so you know for arcane, intel and haste are virtually identical 1 intel roughtly = 1.08 dps, while 1 haste roughly = 1.07 dps..... balance the 2... but never sac spell power for haste in equal mounts.

I dont care how fast you hit me if you hit like a limp noodle... hit like a truck first ...then learn to hit like a truck driving 100mph.

[Raidenwrath]

i currenty play with exactly 600 haste (2518 spellpower, hitcap) unbuffed
during heroism, IV, trinket proc (spider or egg, not both together) and raidbuffs my arcane blast needs roughly 1.0 sec casting time, as well as my mbar, so it would be rather stupid to go way over 600

spellpower > all for arcane, especially after reaching 600 haste and hitcap

[Bentusi]

Stacking that much haste will just drop ones DPS. To be able to stack that much haste you have to give up to much hit, spell power and crit.

Even if he is just a little bit under hit cap the chance of him missing go up do to him casting that much.

Also to get 1k haste he has to give up over 500 spell power and 7% crit (278 crit rating) just to get 450 haste.
the loss in stats alone will lead to dps but giving up 4T7.5 will lead to even more DPS lose.
He is also making thing like BL and IV useless.

Thats what I have been thinking as well, and excessive haste will also give you mana issues in longer fights, especially on AP and or AB spam.

I have tried to balance it but have focused more on SP, hit and crit because of the mana situation. There is another mage in my guild who has done the opposite and stacked haste. What he does is to never go above 2 AB debuffs so the mana burn isnt as intense, but since I have less mana burn at 3 stacks I hit fewer times than he does but alot harder, especially on crits.

Atm he outgears me a little but I am pretty close. I would think that with comparable gear we would be doing virtually the same dps. Of course this doesnt take into account skill and such, it might well be that he is a horrible haste-stacking arcane mage but I really dont think so.

Therefore my conclusion based on careful unscientific observation is that either way works if you approach it right, they just represent 2 slightly different ways of gettin there

[Illuvetar]

Rawr is your best friend when it comes to making gearing choices.

I have personally never played arcane for PvE, but from what I understand spellpower is king, with haste coming in a somewhat distant second. Crit basically doesn't matter as I remember.

So no, don't stack haste to the detriment of spellpower

[Bentusi]

Rawr is your best friend when it comes to making gearing choices.

Crit basically doesn't matter as I remember.

Now at 2150 a noncrit AB on triple debuff, fully raidbuffed, can hit around 5-6k, a whereas a crit can hit for 16-17k.
I believe this is sufficient for me to conclude that your statement is false, of course crit matters.

SP is first I agree there, but otherwise we totally disagree

[Malectrius]

Now at 2150 a noncrit AB on triple debuff, fully raidbuffed, can hit around 5-6k, a whereas a crit can hit for 16-17k.
I believe this is sufficient for me to conclude that your statement is false, of course crit matters.

SP is first I agree there, but otherwise we totally disagree

Every stat matters. You should not ignore one COMPLETELY.

But every theorycrafting shows that haste is better for arcane than crit.

The reasons are that you already get a lot of crit buffs/debuffs from a raid and the difference the percentages you gain from crit gear makes is simply not as much as the percentages you get from haste on gear and gems.

Remember arcane has a chance to proc with clearcasting which increases your crit chance by 30% on proc. Meaning that to some extent the more haste you get the more this would proc.

[Raidenwrath]

Now at 2150 a noncrit AB on triple debuff, fully raidbuffed, can hit around 5-6k, a whereas a crit can hit for 16-17k.

What?
Math isn't your strength, is it?

crit is a rather worthless stat for arcane mages

[Gadirm]

Now at 2150 a noncrit AB on triple debuff, fully raidbuffed, can hit around 5-6k, a whereas a crit can hit for 16-17k.
I believe this is sufficient for me to conclude that your statement is false, of course crit matters.

SP is first I agree there, but otherwise we totally disagree

Uhm... im at almost 2.2k spellpower and a bit over 600 haste (just 2 t7.5 and 2 t7) and i dont think i can remember seeing a 6k AB on a triple debuff, ofc i tend to spend the debuff on MB proc or AM at the very least to not burn mana to much. But also crit for 16-17k.. meh... no.. maybe around 12-13k but unless your kicking Malygos arse and have one spark buff then... no.

[Spomo]

Uhm... im at almost 2.2k spellpower and a bit over 600 haste (just 2 t7.5 and 2 t7) and i dont think i can remember seeing a 6k AB on a triple debuff, ofc i tend to spend the debuff on MB proc or AM at the very least to not burn mana to much. But also crit for 16-17k.. meh... no.. maybe around 12-13k but unless your kicking Malygos arse and have one spark buff then... no.

" a bit over 600 haste"= wasted haste I'm pretty sure the cutoff for icyveins+heroism+haste=590 or so
At this point your casts become 1 sec and the extra haste is useless. I would suggest switching some haste gems out if you can.

[Amandil]

" a bit over 600 haste"= wasted haste I'm pretty sure the cutoff for icyveins+heroism+haste=590 or so
At this point your casts become 1 sec and the extra haste is useless. I would suggest switching some haste gems out if you can.

"Useless" is pretty strong when the only time you don't gain its full benefit is for 20 seconds out of every 5 minutes.