[Mage] 5.2 PTR Discussion (based on Feb 19 notes)

[jtmzac]

I don't think its that simple. From what I know the only reason haste outweighs int is because of LotC and NT. Once we start hitting caps and now we have some different trinket options it might end up being a dps gain rather then a loss to stack int.

[Deleted]

Haste is king for Frost because of Invocation and the interaction it has with Haste. This will be diminished for 5.2 but Haste will still be important due to trinkets moving onto the PPM model.

Currently trinkets do not function on PPM model; it's only Windsong and Jade Spirit (iirc); but in 5.2 they will and this automatically makes Haste a stronger stat for all caster specs. Whether it will make Haste stronger than the current strongest stat (for example, in the case of Fire) remains to be seen; but as it's currently the strongest stat for Frost I don't see that changing.

[Deleted]

We don't have scaling for 5.2, yet. (as far as I know).

For 5.1, except for full BiS invoc/NT/LotC, haste wasn't near 2x Int. It was a misconception. But for the vast majority who is more around 490+, it's still better to gem int/haste than full haste.

And you really overlooking haste effect on rPPM system. Haste increase proc rate but because it's also reduce the "time_since_last_proc", the impact of haste really minor. If it wasn't the case, all spec would have a better scaling with haste. haste have only 2 purpose with rPPM : reduce difference between spec and nearly guaranteed a proc with big haste proc (like BL).

[oraz4000]

They may have DPS numbers from H T15 equipped Mages, something that isn't available to us on the PTR iirc.

So non heroic equipped mages are gonna be permanently nerfed?

[Swizzle]

So non heroic equipped mages are gonna be permanently nerfed?

No, they tune for the end game, not middle of the road. They want optimal geared people to be on equal footing, which means they balance around heroic geared players and full Tyrannical geared players for PvP. Anything below that and you risk better gear blowing up all over the place.

[Deleted]

We don't have scaling for 5.2, yet. (as far as I know).

For 5.1, except for full BiS invoc/NT/LotC, haste wasn't near 2x Int. It was a misconception. But for the vast majority who is more around 490+, it's still better to gem int/haste than full haste.

And you really overlooking haste effect on rPPM system. Haste increase proc rate but because it's also reduce the "time_since_last_proc", the impact of haste really minor. If it wasn't the case, all spec would have a better scaling with haste. haste have only 2 purpose with rPPM : reduce difference between spec and nearly guaranteed a proc with big haste proc (like BL).

Ah, my bad regarding scaling.

To an extent, yes I am "overlooking" the effect of Haste on RPPM, but look at it this way:

20% Haste: 3.33% chance per Fbolt cast to proc X effect.
35% Haste: 3.33% chance per Fbolt cast to proc X effect.

Same % chance to proc, but you're casting faster.

Fbolt: 2s base cast. Cast w/ 20% haste = 1.67s, cast w/ 35% Haste = 1.48s.

1.67/1.48 = 12.8% faster, so ~ 12.8% more Frostbolts cast during the fight (assuming perfect play).

With 12.8% more chances to proc, you will get a better uptime. It's difficult to put it into Maths for me right now (tired); feel free to do it yourself though. With better uptime on trinkets you'll do more DPS, assuming you haven't sacrificed too much to get to 35% raid haste. As Haste is Frost's best stat (and will likely remain so in 5.2) you're sacrificing next-to-nothing; ergo it becomes a higher DPS gain.

[Skidd]

Arcane Missiles damage and SP scaling reduced by 2%.

Arcane
Arcane Barrage base damage reduced by 3% and no longer lists SP scaling.
Arcane Blast base damage and SP scaling reduced by 3%.
Arcane Missiles damage and SP scaling reduced by 2%

And down we go, its like being in a submarine approaching crush depth lol

[Shangalar]

The scaling removed is a datamining error otherwise the spell would be weaker than in Wrath.

[Rethys]

In order to fix frost scaling, I agree, something needs to be done about stat capping. Perhaps add a bonus damage coefficient if your crit chance is >100% after shatter? Sort of similar to how critical block works for warriors or chaosbolt for destro locks.

[Deleted]

Ah, my bad regarding scaling.

To an extent, yes I am "overlooking" the effect of Haste on RPPM, but look at it this way:

20% Haste: 3.33% chance per Fbolt cast to proc X effect.
35% Haste: 3.33% chance per Fbolt cast to proc X effect.

Same % chance to proc, but you're casting faster.

Fbolt: 2s base cast. Cast w/ 20% haste = 1.67s, cast w/ 35% Haste = 1.48s.

1.67/1.48 = 12.8% faster, so ~ 12.8% more Frostbolts cast during the fight (assuming perfect play).

With 12.8% more chances to proc, you will get a better uptime. It's difficult to put it into Maths for me right now (tired); feel free to do it yourself though. With better uptime on trinkets you'll do more DPS, assuming you haven't sacrificed too much to get to 35% raid haste. As Haste is Frost's best stat (and will likely remain so in 5.2) you're sacrificing next-to-nothing; ergo it becomes a higher DPS gain.

It don't work like this. If you only cast Frostbolt :
@20% : FB = 1.67s so with 0.5 ppm, proc rate = 0.5*1.2*1.67/60=1.67%
@35% : FB = 1.48s so with 0.5 ppm, proc rate = 0.5*1.35*1.48/60=1.665%
If you only take one spell, it will result in an even change.
But, the same with an instant spell :
@20% : GCD=1.25s, with 0.5ppm, pr=0.5*1.2*1.25/60=1.25%
@35% : GCD=1.11s, with 0.5ppm, pr=0.5*1.35*1.11/60=1.248%
@75% : GCD=1s, with 0.5ppm, pr=0.5*1.35*1/60=1.458% (BL, metagem proc, etc)

So what append with haste and ppm?
Haste increase proc rate but reduce the time between proc chance. Mathematically, it will result in no change (+/- rounding). But some spell have travel time, isn't affected by haste (frozen orb), can be GCD capped, or is simply affected by haste differently (DoT). It result in a "time since last proc chance" that is reduced less than in theory. It's the same things that explain why the DPS isn't increase by 30% with BL, but only by 20% or 25%.

It's explain why up-time or interval formula is more wrong than true (and they are formula that I used).

Conclusion :
Haste increase the proc rate more than reducing the time between proc event. But the difference isn't really significant.
For comparison, the increased FoF proc rate will be far more important in determine haste scaling than rPPM influence.

[Deleted]

It don't work like this. If you only cast Frostbolt :
@20% : FB = 1.67s so with 0.5 ppm, proc rate = 0.5*1.2*1.67/60=1.67%
@35% : FB = 1.48s so with 0.5 ppm, proc rate = 0.5*1.35*1.48/60=1.665%
If you only take one spell, it will result in an even change.
But, the same with an instant spell :
@20% : GCD=1.25s, with 0.5ppm, pr=0.5*1.2*1.25/60=1.25%
@35% : GCD=1.11s, with 0.5ppm, pr=0.5*1.35*1.11/60=1.248%
@75% : GCD=1s, with 0.5ppm, pr=0.5*1.35*1/60=1.458% (BL, metagem proc, etc)

So what append with haste and ppm?
Haste increase proc rate but reduce the time between proc chance. Mathematically, it will result in no change (+/- rounding). But some spell have travel time, isn't affected by haste (frozen orb), can be GCD capped, or is simply affected by haste differently (DoT). It result in a "time since last proc chance" that is reduced less than in theory. It's the same things that explain why the DPS isn't increase by 30% with BL, but only by 20% or 25%.

It's explain why up-time or interval formula is more wrong than true (and they are formula that I used).

Conclusion :
Haste increase the proc rate more than reducing the time between proc event. But the difference isn't really significant.
For comparison, the increased FoF proc rate will be far more important in determine haste scaling than rPPM influence.

Your maths is the same as mine...

Put it this way. The % chance to proc is per spell hit; PPM is calculated like this which is fine; but because you are casting faster, you simply create more chances to proc, with the proc chance being almost the same.

More chances of proccing = (in general terms) more procs. More procs = higher uptime; therefore increased benefit

[Deleted]

Your maths is the same as mine...

Put it this way. The % chance to proc is per spell hit; PPM is calculated like this which is fine; but because you are casting faster, you simply create more chances to proc, with the proc chance being almost the same.

More chances of proccing = (in general terms) more procs. More procs = higher uptime; therefore increased benefit

Yep, indeed but the increase isn't as great as haste would suggest.
+10% haste won't result in +10% more chance to proc, more 8% (human factor, lag, haste interaction, etc).
And we spoke about trinket proc that will result in +/- 1% dps. So 10% more frequent proc of 1% of the dps, will result in a +0.1% dps increase. Even +0.1% is always interesting (+1k dps).

In term of haste scaling, I can't number it but it will result in what? an increase from 1.49 to 1.50 ?
It could be important if crit is scaling around 1.48. But it's only an example, and I don't think it will be the case.

[huth]

OK, I have made some calculation for Frostbolt.
At 30k spell power, we have :
"+52%" => 191%*30k = 57.3k damages
"+32%" => 155%*30k*1.15 = 46.5*1.15 = 53.5k damages
Base damage ignored (~1.5k/2k). The debuff damage is multiplicative.

In short 1.91 > 1.78 (1.55*1.15).

Easy solution : remove FB's debuff and add a 20% PvP nerf like Frost bomb.

Check your math. 32% buff means 166% base, not 155%. 166%*115% ~191%.

zomg already explained why the easy solution is bad 3 pages back.

[Deleted]

Check your math. 32% buff means 166% base, not 155%. 166%*115% ~191%.

zomg already explained why the easy solution is bad 3 pages back.

Yep, I have redo my math on tweeter but not here. I always check twice when I use WoWhead data, but not this time.

Yep, I have read Zomg post. In the same direction but on PTR forum : http://us.battle.net/wow/en/forum/to...61?page=46#920

[Deleted]

Yep, indeed but the increase isn't as great as haste would suggest.
+10% haste won't result in +10% more chance to proc, more 8% (human factor, lag, haste interaction, etc).
And we spoke about trinket proc that will result in +/- 1% dps. So 10% more frequent proc of 1% of the dps, will result in a +0.1% dps increase. Even +0.1% is always interesting (+1k dps).

In term of haste scaling, I can't number it but it will result in what? an increase from 1.49 to 1.50 ?
It could be important if crit is scaling around 1.48. But it's only an example, and I don't think it will be the case.

Shouldn't really bring human error into maths calculations as human error is subjective. It's less accurate, sure, but if we assume all humans play at the same level; the difference amounts to the same; hence why you don't variate based on skill.

I would think that it would increase the weight by a little more; as the weight would be stronger until you reach the haste breakpoint; but I can't comment on a rough value; we'll just have to see what SimC throws out.

[Deleted]

I will check if I can compile SimC this sunday just to see general result as ppm-trinket seem to be implemented.

[Deleted]

Another nerf, oh boy

[nimryas]

The new percentages on the latest new post on mmoc are on top of the old changes?

so on top of the recent nerfs comes this:

• Arcane Missiles damage and SP scaling reduced by 2%.
• Arcane
• Arcane Barrage base damage reduced by 3% and no longer lists SP scaling.
• Arcane Blast base damage and SP scaling reduced by 3%.
• Arcane Missiles damage and SP scaling reduced by 2%.

?

If true, this is totally fudged :')

[stillwinds]

The arcane charge buff is now 60% damage per charge and 180% mana

[Deleted]

blizzard now will make a new glyph that make Arcane Blast generate a big amount of threat and reduce damage taken and damage done by 60% and make us tanks!!! because we can no longer dps as arcane for sure lol!!!! /FACEPALM