Assuming that they keep the 4% increase in for Spellstone and Firestone has anyone done the hard math on which is better for what specs? I was assuming that if it is 4% increase in direct damage that the firestone is better for demo, destro and hybrids. I was revisiting the numbers and am not so sure anymore.

Since the Glyph vendors are up again on the PTR i decided to buy a glyph of life tap again to test the numbers and see if i could figure out why i was getting 45% and others were getting different numbers and I came upon this with 0 talents the glyph was reading 25% for me and after a few tests I concluded that it was indeed giving me 25% of my spirit into spell power. But when i had been specced into a 0/41/30 when i bought the glyphs this morning it was reading 45% so I wanted to see how i got those numbers and what talent it was that seemed to be affecting the percentage of the glyph and i figured out that it was unholy power as I put points into the percentage on the glyph was going up. I was wanting to see if anyone else was getting this and i can provide screen shots of this If anyone would like

To some people hating the 'new' affliction rotation, here's a comparison of which timers to watch between live and the current ptr:

Live:

1. Haunt
2. UA+Immo having same duration
3. Siphon Life and glyphed Curse of Agony having nearly thesame duration
4. From 25% and down, watching Drain Soul

On the PTR:

1. Haunt
2. UA
3. CoA
4. Drain Soul post 25%
Possibly: 5. Lifetap uptime through Glyph of Lifetap every 20 seconds

So, yeah. Once every 15 seconds and once every 30 seconds you'll use an extra global cooldown's worth of casting Shadowbolt, opposed from recasting Immolate and Siphon Life. with the new Glyph of Lifetap, I think it's going to be worth to keep using a Lifetap every 20 seconds for the buff, but I'm not sure about this.

Ask yourself this: is the spec really dumbed down that much? Just compare yourself vs the other dps classes, it seems to me like we roughly all using the same number of spells. Yes, I've always liked having a whole arsenal of dots to cast and slowly but surely having them all tick away an enemy's hitpoints, but I'm also liking the increased damage done by Corruption and Unstable Affliction, seeying corruption crit for over 6.5k just makes me giggle like a schoolgirl.. The new spec seems to allow more aggressive and faster-in-effect dps time, which IMO is a good thing. Besides, being more dependant on Shadow Bolt and with Corruption and Unstable Affliction scaling better with raidbuffs and debuffs, spellcrit and haste become more welcome stats, aside to spellpower, right?

The thing I do dislike is how much Shadowpriests and Affliction-Warlocks are going to be alike in the future.

It isn't a dps increase to life tap in order to gain the buff, but you should space life taps out by at least 20 seconds in order to gain full benefit of the buff. I'm not quite sure what you mean by "more aggressive and faster-in-effect dps time", big numbers flashing on your screen and random burst doesn't equal an increase in dps.

I believe he was refering to the fact that with less dots, we get into our full rotation sooner, yeilding higher dps earlier in the fight. Currently on live, you have roughly 10 seconds of putting up all initial dots and debuffs; after 3.1, you'll have roughly 7 seconds of putting up all of your dots n' debuffs, giving you an entire extra shadowbolt within the first 10 second window. (Basing these times off of my own haste values ~600rating + Latency) I don't think he was refering to an overall dps increase, simply to more dps in the first few seconds of the fight..

Adding this to the op conflag discussion.
I took this absolutely ridiculous screenshot on the ptr just now.
My sp at the time was 4152, the build 0/40/31, CotE was on the dummy, MC was active, and I was using Spellstone (2/2).

It is the start of the "clip immolate if your trinkets proc"-rotation for conflag-builds!

On a side note, has anyone noticed spellstone and firestone buff disappearing after death ?

Question: when you conflag, is the damage based on trinkets and other temp buffs that were active when the immolate was cast, or when the conflag was cast? IE, does it look to the immolate it popped and get 5 ticks of that, or does it act more like a normal nuke with respect to damage calculations?

Testing on the PTR shows it looks to the immolate. Except for the fact that it doesn't benefit from the glyph, of course. Also, emberstorm is applied to the final conflag damage, meaning it's multiplicative with the immolate talents, not additive.

Can you conflagate twice in one immolate with the conflagate glyph?

yes

I believe he was referring to recasting immolate while you have spell power trinket buffs up, thereby giving all subsequent conflags on said Immolate higher dmg. I'm not quite sure how to go about modelling this scenario to see at what point (how big of a spell power buff) you would need to get to make it worth recasting immolate, seeing as you would only be losing less than a full incinerate cast. I wouldn't be surprised to find that you would want to clip immo with a new cast. I may have thought that through wrong. Anyone have a reasonable model for how to figure the magnitude of the spell power buff you would need to make it worthwhile?

Some quick napkin math showed with 0/40/31, assuming your next normal recast of Immolate will benefit from the proc/trinket SP buff, clipping an Immolate (instead of casting Incinerate) to get a better DOT for the next two Conflags is a DPS loss. The SP bonus a trinket/proc would need to give for this not to be true is massive. If you have 2500 SP, 40% Crit (after buffs/talents) a trinket would need to add 4188 SP to make this worthwhile.

Here's the napkin, corrections welcome:
s = spell power
b = bonus spell power from proc/trinket
h = haste multiplier: 1/(1+haste%)
c = crit multiplier: 1+crit%

Notes:
* 40/31 with Conflag glyph
* Assuming MD w/Imp gives 5% additive with ES when calculating Conflag
* Assuming able to get two Conflags in under new Immolate
* Assuming normal recasting of Immolate would have benefited from proc/trinket
* Ignoring Molten Core

Immolate: 1.5(460+0.2(s+b))c

Immolate Tick: -0.5*1.56(157+0.2s)
Half an Immolate tick is lost on average as a result of pushing back the next Immolate tick due to resetting the DOT.

Conflags: 2*1.2(1.36*b)

Simplified: 0.3sc + 0.3bc + 690c - 0.156s + 3.264b - 122.46

VERSUS

Incinerate: 1.2(786+2.5/3.5(s+b))c * 1.5h/2.25h
Scaling this value to the fraction of cast time available

Simplified: 628.8c + 12sc/7 + 12bc/7

Compare:
0.3sc + 0.3bc + 690c - 0.156s + 3.264b - 122.46 > 628.8c + 12sc/7 + 12bc/7
0.3s + 0.3b + 690 - 0.156s/c + 3.264b/c - 122.46/c > 628.8 + 12s/7 + 12b/7
0.3s + 0.3b + 61.2 - 0.156s/c + 3.264b/c - 122.46/c > 12s/7 + 12b/7
61.2 - 0.156s/c + 3.264b/c - 122.46/c > 12s/7 - 3s/10 + 12b/7 -3b/10
61.2 - 0.156s/c + 3.264b/c - 122.46/c > 99s/70 + 99b/70
3.264b/c - 99b/70 > 99s/70 + 0.156s/c + 122.46/c - 61.2
b(3.264/c - 99/70) > 99s/70 + 0.156s/c + 122.46/c - 61.2
b > (99s/70 + 0.156s/c + 122.46/c - 61.2) / (3.264/c - 99/70)

b > (1.4142s + 0.156s/c + 122.46/c - 61.2) / (3.264c - 1.4142)

I'm not exactly how your modelling immolate, it seems your ignoring that the rest of the dot's duration would also tick out at the higher spell power. Other things to consider would be when you were clipping the immolate, clipping it early would be a much larger dps drop than clipping it just before its last tick. I dont think you can model this as a direct damage loss/damage gained of that given second as it has a large affect on your dps during a considerable window (at least as long as your previous immo + the new immolate were both up)

I think the best way to model it would be, compare damage gained from: High spell power immolate ticks, high spell power initial direct damage, two larger conflags, and your sacrificing, 1 high spell power incinerate, remaining low spell power immolate ticks.

Now what I'm unsure of is how to manage the difference in # of ticks between the new immo and the previous. If you were only clipping the last tick of an immolate, your net damage lost is definately smaller than if you clipping an immolate immediately after it was already cast.

Looking back at it, I took two assumptions that somewhat contradict each other. I orginally assumed that either the normally scheduled Immolate or the clipping Immolate would benefit from the proc/trinket so I didn't bother to include the boosted Immolate tick damage because you'd be getting it later anyway if you didn't clip. However, I failed to include that that would mean you'd ALSO get a boosted Conflage later, so over all you are netting only a single extra boosted conflag, at the price of a push back of Immolate ticks and a partial Incinerate damage.

Overall, the previous forumula only almost works for a proc/trinket with a duration that will last you would have normally recast Immolate, and it indicates that you shouldn't bother clipping Immolate. I'll rework for a shorter buff to see if it makes enough of a difference but I'm guessing it won't. In any case, it won't be a number like X, but instead a number based on your SP and Crit%, and so as you get better gear, the proc/trinket required to make this worth it because higher.

Edit: Adding in the boost to Immolate ticks from the proc/trinket, we add to the initial equation:

Gained Immolate Ticks: 5*1.56(157+0.2b)

This makes the final simplified comparison:

b > (1.4142s + 0.156s/c - 1102.14/c - 61.2) / (4.824/c - 1.4142)

That's a lot better, but at 2500 SP and 40% crit (buffed with talents) you would still need to proc/trinket +1460 SP to make clipping Immolate worthwhile.

What he said

And for a dot like Siphon life for example to be better dpct compared to Shadowbolt, it needs to surpass a certain amount of time in order to do so, so if you work with windows of time where you can dps, being able to get damage going faster could be beneficial.

(+I am aware of the differences between burst and sustained dps)

New Immo/conflag #'s for this patch

1910 spelldmg

no talents/glyphs

Hits 842
ticks 539

Imp immo
hits 1094
ticks 701
conflag 3503

aftermath
hits 1159
ticks 801
conflag 4005

emberstorm
hits 1221
ticks 814
conflag 4215

F&B
Hits 1360
ticks 872
conflag 4513

WITH Glyph

no talents
hits 758
ticks 647


Imp Immo/conflag
hits 1011
ticks 809
conflag 4042


Aftermath
Hits 1011
ticks 841
conflag 4204

Emberstorm
hits 1137
ticks 922
conflag 4835

F&B
hit 1265
Ticks 987
Conflag 5178

Latest notes from MMO-Champ give these as the T8 set bonuses. At first glance, 20% on UA and Immolate looks pretty ridiculous as a set bonus, especially for only 2 piece.

If 2pc does effect conflag than based on the numbers I gave above conflag would be hitting for 6200 noncrit at 1910 spellpower w/o full talents.

This would basically yeild 30k conflags fully buffed/geared/talented.

Yeah, this is frankly getting beyond silly. My guess is conflag will not be able to crit in the end, and Pyroclasm will be reworked.

Actually, you don't need to worry about clipping anything other than a partial tick. That is because Immolate can only tick at best every 3 seconds no matter how often you cast it. Whether you cast it 20 times of 21 times in a fight won't change that. Compare the following cases for a 5 minute fight with perfect Immolate uptime:

1) Baseline, you will cast Immolate 20 times, and it will tick 100 times
2) You clip one full tick of one Immolate, giving you 21 casts but still 100 ticks
3) You clip four full ticks of one Immolate, giving you 21 casts but still 100 ticks

In all three cases, the total periodic damage from Immolate is 100 ticks worth. But in both 2 and 3 you get an extra Immolate front-end damage in. Which would seem good, but it replaces an Incinerate. Basiscally, the 21st Immolate in these examples must stand up against other spell choices without including its DOT because you've already got the required minimum 20 casts needed for maximum uptime.

Actually, I just kind of assumed those numbers weren't reachable, but if they are, then the number needed can become slightly lower. Best case you can have is for your proc/trinket to go off while Molten Core is up, but it wasn't up when you cast the previous Immolate and Molten Core will still be up when your next Conflag comes off CD (you can't be sure MC willl be up for both Conflags when you are deciding to clip or not).

Best case leaves you with the following final formula:
b > (1.5371s + 0.056s/c - 1259.14/c - 54.8) / (5.06/c - 1.5371)

2500 SP and 40% Crit = +1439 SP makes clipping Immolate on proc/trinket+MC w/Conflag benefiting from MC worthwhile

But it scales with gear heavily with gear. Consider the following cases:
2500 SP and 45% Crit = +1545 SP needed
3000 SP and 40% Crit = +1819 SP needed
3000 SP and 45% Crit = +1949 SP needed

I think these cases are extreme enough that realistically, only an addon could watch for this event convergence and make the call. Not only that, but you'd want the addon anyway to calculate the exact bonus SP needed. I honestly doubt if you'd see more than a tiny handful of DPS total over an entire fight from clipping at the right times. The penalty for making the wrong call would likely wipe at gains from multiple correctly called clips.

Napkin:
Immolate w/MC: 1.6c(460+0.2(s+b)) = 736c + 0.32sc + 0.32bc
Immolate DOT tick clipped:-0.5*1.56(157+0.2s) = -122.46 - 0.156s
Conflag w/MC: 1.3(1.36b) = 1.768b
Conflag w/o MC: 1.2(1.36b) = 1.632b (@)
Immolate DOT bonus damage w/MC: 5*1.66(157+0.2b) = 1303.1 + 1.66b
Immolate DOT base damage increased by MC: 5*0.1(157+0.2s) = 78.5 + 0.1s
Incinerate w/MC: 1.3(786+2.5/3.5(s+b))c * 1.5h/2.25h = 681.2c + 13sc/7 + 13bc/7

(@) We can't assume the second Conflag will have the benefit of MC so we'll assume it won't.

Solve:
0.32sc + 0.32bc + 5.06b - 0.056s + 736c + 1259.14 > 681.2c + 13sc/7 + 13bc/7

Yields:
b > (1.5371s + 0.056s/c - 1259.14/c - 54.8) / (5.06/c - 1.5371)

Ok that makes more sense. It proves that basically, if you have the trinkets that can proc for high amounts of spell power, more than likely the rest of your gear will also be of a quality making it not worth clipping. It was surprisingly feasable at first glance though. Thank you for the mathing.