That would be true if Mind Blast had no cooldown. In practice a garbage GCD (that is, one that doesn't fit cleanly into your rotation) is worth closer to 2000 damage, or about the amount you'd lose by flaying through an MB cooldown.

In practice, you buy yourself a MB cycle. I would never call any GCD garbage unless you aren't casting anything productive in it, since those mindflay casts are also running through the cooldowns of your other spells.

Compare
VT-MB-SWD-MF-SWP-MB-DP-VE-MF-MB (on fixing MF, you can cut in DP at the 2 second MF mark)
against
SWP-VE-VT-DP-MB-SWD-SWP-MF-MB

You gain 1 tick of VT and nearly a full cycle on MB in trade for 1 tick of your weakest dot + small percentage on the first MF MB. VT ticks harder than pain ticks, and having an extra MB is better than 10% of MB and 20% of MF.

But this is a strawman because you're comparing a good VT-first rotation against a really terrible Pain-first rotation. Of course Pain-first is going to look worse if you decide to slot in VE and DP early instead of starting your MB cooldown. Also, I'm not sure what you mean by "on fixing MF..." because you should have enough haste to fit 4 GCDs between your MB cooldowns. If you don't, you're much better off just eating the increased effective cooldown. Blizzard says they fixed the timing of the damage ticks; they lie.

The problem here is that casting VT right before MB sets you up to have to make a subprime decsion the next time you need to cast VT. ie you will be in a situation with around 1sec to go on VT right when MB comes back up? What do you then? Well VT is higher DPCT so maybe you recast it first then MB. However that prepetuates the cycle you have created and the one I used to of pushing every otherMB back an extra second or so.

That and you are created an apples to oranges comparison with the different timeing of VE. you are comparing a static position cycle to a cycle that is clreay tuned for movment although it would need VT and DP switched to give the msot movement time.

If you want to truely compare cycles compare apples to apples.

VT-MB-SWD-MF-SWP-MB-DP-VE-MF-MB

Should be ran through a sim against something like:

SWP-MB-VT-DP-MF(2)-SWP-MB-VE-MF-SWD-MB-VT

Although I am winging it with the 2nd it appears to match on time with 3 total MB each. The first has the advantage of getting VT in earlier as the stronger dot. The 2nd has the advantage of 10% more MB damage on first cast and better alignment of VT and MB cooldowns.

I am tryin to convince myself it is really ok to open up with a MB out of the gate. We do have great tanks but as a BC tank and raid leader, I would never hear the end of it if I cause a wipe being overagressive on my first cast. However if I get over that I think that this would be a valid opener on static fights.

MB-VT-DP-MF(2)-SWD-MB -VE-MF(2)-SWD -> normal priorities based rotation.

Due to lag and haste rating you're probably going to end up chopping ticks on VT regardless of what you do. My instinct is to delay casting MB instead of allowing dots to fall off if it happens, but since the dot tick timer no longer resets when you cast it over an existing dot, chopping 0.5-1 seconds off of some VTs won't hurt you very much.

A lot of us run with [Dying Curse] or [Sundial of the Exiled]. At the start of the fight, they will proc. I usually try and hold off 3 or 4 GCDs before applying VT or DP because I want to take advantage of the bonus damage from those trinkets and shadow weaving. Some maths on the opportunity costs of this decision follow.

[Dying Curse] has a 35% rate of activation, with a 45 second internal CD. This puts its activation chance:

on 1 cast at 35%
on 2 casts at ~58%
on 3 casts at ~72%
on 4 casts at ~82%
on 5 casts at ~88%

At 0 haste you would use a sequence like this:
Pain, MB, MF (clip at 2), MF (clip at 2) and then VT, MB, DP, Pain

At normal haste levels, this would obviously change.

The sequence gives you a significant chance of having your trinket active for the application of VT and DP, as well as having a full SW stack for the whole VT. Over starting with VT, a later VT gets about 3.5k extra damage.

Compare this against your:
VT-MB-SWD-MF-SWP-MB-DP-VE-MF-MB

On this sequence, there is a very good chance that you won't have a trinket activation for your VT reapplication. The buff from [Dying Curse] lasts 10 seconds, and if it activated within the first four GCDs, it will be over before it is time to reapply VT unless you cut off a tick.

Comparing the two cycles, the cost of delaying the application of VT is about 1600 damage. The average ticks of Pain are 900 less than VT in my gear, but that is with all appropriate buffs and trinket activations considered. Also, Pain ticks 1.5 seconds after the GCD for the cast is up, VT ticks 3 seconds after the first cast.

There is also the additional cost of a GCD in reapplying Pain to get the appropriate damage buffs. I am not sure, but I think reapplying pain does not interrupt its damage schedule, unlike other dots. Still, you lose any bonus damage you could have gained from raid buffs from not applying pain sooner. I will be generous and say that this cost is about 1/3 a MF and 1 tick of Pain, or about 5k damage.

It comes down to a gain of 3.5k damage on VT, an average of 1k on Mind Blast, and about 2.4k on Mind Flays. All of this for roughly a difference of 300 damage?

I’ve come to the final conclusion that worrying too much about starting sequences is silly. The differences are relatively minor as long as you obey the following rules:

1.) Get MB on CD early.
2.) Make sure that you get Pain applied (or reapplied) after all applicable buffs and debuffs are active.

Reordering MB to be infront or behind MB doesn't change the fact that you'll end up with conflicts regardless. Cooldown confliction is baked in SP mechanics -- its ALWAYS less than ideal, which is why we have a priority list instead of a rotation.
But I'll look at reducing confliction.

Still though, you don't open with SW: P

Calculating the 10% MB bonus and 20% MF bonus using my own numbers
3999 MB average
1925 MF tick average
363 bonus to MB
641 bonus to MF for 2 ticks
You trade 1000 damage for one GCD. Given that my mindflay ticks nearly double that in less time, it's still a bad tradeoff, even if they both crit.


Whatev:
Mindflay currently applies 2 weaving, instead of 1

I think you're neglecting crit on your numbers. You want (Total Mind Blast Damage)/(MB Hits + MB Crits + MB Misses).

So on my last PW I have 113079/(8+9+0) as about 6652. MB gives a bonus of about 605 (it is not 10% of 6652, but 110% of 6047). Also for mind flay you do the same. For me, its 3024 per tick, or 500 damage each time.

So it is still a little under 1600 damage as a trade off for the GCD, and independent of other considerations, it is bad to have to reapply Pain. That said, if you take reapplying Pain in context it can become worthwhile. I give one potential case above.

SWP is reset on a re-cast as well.

Like Akthal said, this is wrong. Chopping VT ticks is just poor play unless you're sporting a ping over 500ms. You can see how easy it is to set up your spell casting to be more efficient using a mod called EventHorizon, it is listed in the SPriest mods thread.

I did not include critical since it works like a flat multiplier ((1-critrate)*1+critrate*2). If we compare damage between 2 sets of spell damage (in this case, MF+MB bonus versus MF), they use the same crit multiplier. You can simplify the problem by dividing both sides with that crit number. This effectively nullifies crit in terms of the comparison, saving you the effort of calculating in crit.

Of course, if your actual MF ticks are 3k, then I'm impressed.

The expected value of my MF Tick is 3k, which must include the crit rate. It is okay if you are comparing damage proportions, but that was not clear to me.

Furthermore, you cannot escape crit rate, as it is factored in all of your dots. You would have to factor it out, which to me seems like a big pain.

Last, I want to use expected value over a proportion method because I am trying to look at the whole casting sequence, and not some small part of it. It is suboptimal to cast Pain twice, but it can be justified with larger gains elsewhere.

Exactly, SP is taken into account on every tick.

This happens since the time of Lifebloom nerf, when they could cast it with a trinket up and keep the super-Lifebloom up FOREVERE.

Hrm. I apologize, my information was bad then.

Holding off 3 or 4 GCDs before applying VT or DP is probably not a good idea, though you're correct that the spellpower value that is present at the initial cast of the spell is the one used throughout the whole duration of the spell, which is why during bloodlust you should pray for VT and DP to be coming off cooldown at the time one of your trinkets procs so you can slam a wild magic pot before refreshing it.

No. Shadow Word: Pain is the only spell that can change tick values mid-duration, and it only happens when you refresh it with a mindflay.

Two things wrong with this opener. First, you didn't cast SWP... though I'll assume you meant SWP where the first SWD is. Second, MB does not apply misery. Due to misery's mechanic and a raid's reliance on that 3% spell hit the only viable first cast options are MF, VT, and SWP. Pick your poison. My personal preference is to MF1 to get a quick 2 SW stacks while instant cast debuffs are going out. Afterwards I VT and by the time the VT cast is finished all casted debuffs (imp scorch) are already up for its benefit.

My opener: MF1-VT-DP-MB-SWP

The lifebloom nerf was that when you reapplied the spell, it would update with your current spell power. If you cast VT with trinkets up, you get 40% of the trinket's value per tick. That seems almost too good to pass up. Over a full five ticks we're talking about an extra 1500 damage or so.

The conclusion from earlier in the thread seems to be that all spell power bonuses are applied on auto-refresh, but anything that involves a multiplier (including crit debuff) needs a full reapplication to get counted.

I see, so it doesn't matter whether the crit is from a buff (potion) or debuff (Loatheb) but neither will affect SW:P unless it is manually reapplied. Does this also mean that if you drink the potion of Wild Magic, you can also keep the extra 4.36% crit (but not the spellpower boost) indefinitely as long as you refresh SW:P using Mind Flay?

I did some tests initially when 3.0.8 went live and came across that the crit bonus didn't fall off via MF refresh after the pot had worn off. I was talking to Dezzimal about this on IRC not to long ago, and am going to do some more testing to see what actually falls off and what stays.

Haste Breakpoints:

I'm looking for the specific "Haste Breakpoints" for a Shadowpriest; what I mean is the amount of haste to do exactly 2 mindflays in between mindblasts cooldowns, and the amount of haste needed, with bloodlust, to achieve three mindflays between mindblast cooldowns. Does anyone have these numbers confirmed?

Apologies if I missed it; I did search the thread but didn't see it outlined.

That's all pretty basic math.
Fully improved MB has a 5.5sec cd, that means mindflays of 2.75 sec.

Means you need (3/2.75 = 1.091) 9.1% haste. Don't forget the raid buffs you get when converting that to haste ratings. In fact, I'm not 100% sure how the shammy totem works into the haste formula; I recall Blood Lust working multiplicative.

Blood lusted you want 3 in 5.5 sec, so that's 1,833 sec per mindflay. 3/1.833 = 1.63, dividing by BL, 1.63/1.3 = 1.258, or 25.8% haste. You can look up the conversion ratios for haste-% to ratings on wowwiki or whatnot I'm sure.

Great, exactly what I was looking for. I checked all of the shadow priest forums and couldn't find the exact math. Here it is, for all those looking for the same thing or just to check my numbers.

Total Haste Needed for 2 Mind Flays between MD Cooldowns: ( All numbers rounded up to the nearest whole number)


No Buffs/Self buffed = 299 Haste rating needed for 2 Mind flays between MB Cooldowns.
With Wrath of Air totem + Imp. Retribution Aura= 37 Haste Rating
Just Imp. Retribution Aura = 201 Haste Rating
Just Wrath of Air totem = 135 Haste Rating

Total Haste Needed for 3 Mindflays between MD cooldowns With Bloodlust:

No Buffs/Self Buffed = 853 Haste Rating
With Just Wrath of Air = 689 Haste Rating
With Just Imp. Ret Aura = 754.17 Haste Rating
Imp Ret + Wrath of Air = 590 Haste Rating

If this is wrong, let me know!

Edit: Wrath of Air and Imp. Ret aura may not stack, Maxdps seems to claim they do. Going with assumption they do, will edit if they do not.