Thank you for the testing. I was wondering about that.

Now the next question about haste and dots is...

Is it calculated per tick, or per refresh.
With the results you have gotten, it seems as though it might be per tick. (I know there was a blue post saying that they were going to make it so haste, crit, and sp were revaluated every tick)

This would mean that you loose the benefit of bloodlust on your ticks the moment you loose bloodlust (And on a positive note, gain the benefit of bloodlust the moment it gets popped).
For that matter, you woudn't have to worry about popping your haste pot before dot application, and could instead pop it after the dots are up so that you get the full benefit on MF spam *and* your dots.

i'm new to this msg board, so i hope this isnt out of line.

my testing shows that haste is the only exception currently to swp getting updated with mind flay. and that mind flay does not update VT or DP in any way. however debuffs on the mob are recalculated per tick, without any interference from us. crit and damage modifiers are both now updated with mind flay (they were not before).

this means that currently, swp hasted with heroism, pots, whatever, can be rolled the whole fight. but honestly, if haste is the only exception, i would expect it changed at some point. swp also clearly benefits from haste now, but i think we knew that already.

i will preface by saying that while these are small samples, i did run them multiple times with the same result. the only variation being sometimes getting 1 crit while naked, which is expected. i invite anyone else to run the same or similar tests. if these are somehow flawed and i am not understanding something correctly, i apologize.

here are the tests i ran on the target dummies.

Naked time between ticks (expected results)
VT: 3 seconds (5 ticks) (0 crit)
DP: 3 seconds (8 ticks) (0 crit)
SWP: 3 seconds (6 ticks) (0 crit)

Cast Naked, dress to full gear, refresh with mind flay. (same results no effect from added haste)
VT: 3 seconds (5 ticks) (0 crit)
DP: 3 seconds (8 ticks) (0 crit)
SWP: 3 seconds (cant count ticks, reset with mind flay) (4/6 crit showing that crit was indeed updated)

my testing also showed that if i cast vt and dp at the end of heroism (again naked) i would continue to get 2 second ticks until hard refreshed. and that dots cast before heroism also did not benefit from heroism until recast.

to prove the point, i let dots fall off and cast right as heroism is ending (still naked)
VT:7 ticks (+2)
DP:11 ticks (+3)
SWP:8 ticks (+2)

meaning, that mind flay does not update VT or DP with changes in character stats. but does update SWP with crit (new), damage modifiers(new) and spell power (like always), but not haste.

next tests

dot the dummy, then cast curse of elements (while naked)(warlock helped). i did this naked because the damage ticks on all the dots are always within 1 point of damage to each other, excluding the rare crit.

no coe debuff (no shadow form) (control test). damage done per dot:
VT: 224
DP: 231
SWP: 360

after coe was cast (still naked and no shadow form) (no mind flay)
VT: 242
DP: 250
SWP: 389
this shows that dots are in fact updated per tick as the mob is debuffed. no need to recast anything for debuffs.

joining an arcane mage in a group (3% damage) and refreshing with mind flay (before mind flay no change).
VT: 224
DP: 231
SWP: 371 (only change)
this shows that getting a damage modifier power up will only update swp, and only on mind flay refresh.

Good thing CoE was calcuated on tick Pre 4.0. already... Test it with improved Scorch next, that at least wasn't calcuated on tick. Nothing has changed for CoE.

Also to clarify it: NO DoT should be calculated on tick.

What cataclysm changed are that auto refresh mechanics should recalculate all buffs/debuffs that previously could be rolled throughout the whole fight. DoTs without auto refresh are still calculated on cast.

VT/DP are calculated on cast. (CoE being the exclusion, which it always was)

I think you made some minor errors in these tables for the breakoff points. For example in your table for SW:P, the breakoff for 8 ticks is 33.3%, but my total haste along with 2/3 Darkness when I tested (Albeit in my holy spec, so no 5% haste buff from Shadowform) was 1.2891*1.02 = 1.314882 and I still had 8 ticks. Additionally I had 11 Devouring Plague ticks at this amount of haste, but your table specifies 37.5% haste as the breakoff point.

I had to take my haste down to 1.2165*1.02 = 1.24083 for Shadow Word: Pain to go from 8 ticks to 7. That doesn't indicate an exact breakoff point, but it does indicate a silly math mistake. For some more precise proof, I had 11 ticks of Devouring Plague at 31.49% haste, and when I dropped down to 31.18% total haste I only had 10 ticks.

These are the results that I have for breakoff points. Let me know what you think, or if I've missed something.



Devouring Plague - 8 ticks base

Ticks	Haste
9	6.25%
10	18.75%
11	31.25%

Mind Flay - 3 ticks base

Ticks	Haste
4	16.6666...%
5	50%
6	83.3333...%

Shadow Word: Pain - 6 ticks base

Ticks	Haste
7	8.3333...%
8	25%
9	41.6666...%

Vampiric Touch - 5 ticks base

Ticks	Haste
6	10%
7	30%
8	50%

honestly, i was not aware. i went back this morning and tried to test with the 5% crit debuff mages give. the results arent definitive, but they point to the debuff not applying until a hard recast for VT / DP. but mind flay was updating swp. i'm not sure how else to test except a larger sample size, which i did not have time for.

casting dots before critical mass, refreshing with mind flay, repeat.
VT: 15 ticks (0 crits)
DP: 24 ticks (0 crits)
SWP: 41 ticks (2 crits)

casting critical mass first, and keeping it up.
VT: 15 ticks (3 crits)
DP: 24 ticks (1 crits)
SWP: 37 ticks (2 crits)

i then tried 3 more tests.
100 ticks of swp, no debuff, 5 crits.
100 ticks, with swp cast first, then refreshed with mind flay after debuff goes up, 7 crits
100 ticks, with swp cast after debuff put up, 8 crits.

still not very definitive. but seems to indicate that debuffs were updated with MF for swp. meaning self haste is the only thing not updated with MF.

As of 4.0.1 always MB on 3x orb w/ t10 4set, always on cd without.

I just did some quick tests on target dummies. Haste numbers were achieved with a combination of haste pots, black magic and hyperspeed accelerators. You can see my spec and gear here
The World of Warcraft Armory - Elegen @ Mannoroth - Profile

1140 Haste
9 SWP
7 VT
12 DP

1390 Haste (my normal base before procs)
9 SWP ticks
8 VT
12 DP

1630 Haste
10 SWP
8 VT
13 DP

1880 Haste
10 SWP
9 VT
14 DP

2130 Haste
11 SWP
9 VT
14 DP

2380
11 SWP
9 VT
15 DP

Strikethrough wasn't showing up, but I see that you see it's off now... :P Here's what *should* be a correct list.

At 0 haste rating you have 9 ticks of DP
At 6 haste rating you have 7 ticks of SWP
At 57 haste rating you have 6 ticks of VT

At 322 haste rating you have 10 ticks of DP
At 511 haste rating you have 8 ticks of SWP
At 663 haste rating you have 7 ticks of VT
At 701 haste rating you have 11 ticks of DP

At 1016 haste rating you have 9 ticks of SWP
At 1080 haste rating you have 12 ticks of DP
At 1269 haste rating you have 8 ticks of VT, and are GCD capped

At 1459 haste rating you have 13 ticks of DP
At 1522 haste rating you have 10 ticks of SWP

At 1838 haste rating you have 14 ticks of DP
At 1876 haste rating you have 9 ticks of VT

There are a few fairly large plateaus here, which I spaced between. I left in values in the 1800's because it's obtainable through a haste potion if you're heavily haste stacked, though realistically I'd probably cut your haste as close to 1269 as possible. These values are all approximate, they might vary by 1 haste either way.

You gain 1 SWP tick every 505 haste, with the first starting at 6 haste
You gain 1 VT tick every 606 haste, with the first starting at 57 haste
You gain 1 DP tick every 379 haste, with the first starting at 322 haste

Note that 1258 is actually a bit too low to reach 50% haste:
(1+(1258/3279)) * 1.05 * 1.03 = 1.4964, or 49.64% haste.
(1+(1268/3279)) * 1.05 * 1.03 = 1.4997, or 49.97% haste.
(1+(1269/3279)) * 1.05 * 1.03 = 1.5000, or 50% haste and GCD-capped.

Also note that I was completely unable to get a 4th tick on mindflay, so I'm assuming that it's not updated on live. With 1945 haste (59.32% tooltip, 72.3% with talents) I still only had 3 ticks. I wasn't able to round up a shaman to test for certain with lust, but I should easily be hitting 4 or 5 ticks with that much haste.

I think I may be misunderstanding your post, but I did test the points at which each SW:P and Devouring Plague get their first additional ticks just to confirm. This is with no shadowform buff, and 0/3 Darkness. Even when not specced into haste increasing talents they appear to match the numbers I've given:

Devouring Plague:
6.16% - 8 ticks
6.47% - 9 ticks

Shadow Word: Pain:
8.23% - 6 ticks
8.54% - 7 ticks

As my previous post would indicate the breakoff points after that seem to be accurate also, even when specced into 2/3 Darkness. From there you can recreate the tables for haste percentages/ratings. For the ratings column I'm going to assume that 3/3 Darkness and Mind Quickening are present, but like Malapropia's numbers and yours, those are approximate:

Devouring Plague - 8 ticks base

Ticks	Haste%	Rating
9	6.25%	0 (Those talents alone put you above this)
10	18.75%	323
11	31.25%	701
12	43.75%	1080
13	56.25%	1459

**EDIT - Mind Flay does not appear to gain additional ticks with haste on live**

Mind Flay - 3 ticks base

Ticks	Haste	Rating
4	16.6666...%	259
5	50%	1269
6	83.3333...%	2280

Shadow Word: Pain - 6 ticks base

Ticks	Haste	Rating
7	8.3333...%	6
8	25%	511
9	41.6666...%	1017
10	58.3333...%	1522

Vampiric Touch - 5 ticks base

Ticks	Haste	Rating
6	10%	57
7	30%	663
8	50%	1269
9	70%	1876

2 Things. This is from shadowpriest.com about MB on the Shadow Orb x3 issue.

Napkin math with some slight testing gave me an avg MF tick @ 2084 and MB(No Orbs) hit at 4055. The cast times for me at 1263 Haste rating and 4pc t10 are 1.66 seconds and 1 second respectively.

3(2084)/1.66 = 3766.26 while 4055/1 = 4055 (~300dps difference, similar to the post above's claim)

This proves with the new spell ranking system, an unorbed MB is better than a MF. Now according to the post from spriest.com, @ 950 haste you have an 11% chance to NOT generate a single orb, so at 1200 or so you are most likely getting at least 1 within the 6.5 seconds of MB's CD. Since each orb multiplies MB by ~1/3rd my MBs would hit at: ~5400/6700/8100 depending how many orbs proc within those 6.5 seconds. This would make MB on cooldown better than waiting for 3 orbs. For example: 3 2 orb MBs is better than 2 3 orb MBs.

Also, I would like to see crit vs. haste numbers now that SW:D has been reduced and the change to the glyph (a 5 or so second icd i think) both have been implemented.

As of just recently, if no one has noticed, there was a nerf to the Shadow Word: Death glyph. They have added an internal cooldown so you can now spam it twice really fast but then must wait the 10 seconds out before using it again, so spamming it isn't exactly an option anymore on targets below 20%(25%?) health.

Though this is true in a base sense, it does not take into account flay's DPCT when the chance to lower fiend's CD is included into it's value. I would be interested to see how much value that added to flay vs an orbed and unorbed MB when you also consider that every GCD not casting flay is a GCD not spent reducing his CD. Keep in mind also that this "extra value" on flay will drop considerably when Cat releases and combat ratings drop making it much less likely to crit on flay than it is now. At this point you can almost guarantee 1-2 10 second reductions per cast at the 277 level.

Keep in mind I am just trying to get through the last 6 weeks of Wrath, so I don't care about cataclysm stuff at this time.

The 3 minute CD is 180 seconds. My shadowfiend does ~24k Damage over 15 Seconds. I'm at a 271 avg ilvl with about 36% crit. Every 1 crit reduces the CD by 10 seconds, so we can add about .056% of the shadowfiend damage to a MF crit, 1333 damage. To make numbers easier I will use 1000 casts.

1002 MF Ticks (334 casts) will have 360 crit and 642 non-crit. While my average tick is 2084, my avg crit is 4300.

(2084*642 + 360*4300) + (360*1333) = 3365808 Damage / 334*1.7 = 5927.81 DPS

1000 MBs will have 360 crit and 640 non-crit. Average hit is 4055, crit is 8513 (NO SHADOW ORBS)

4055*640 + 8513*360 = 5659880 Damage / 1000*1 = 5659.88 DPS

With no shadow orbs procing, yes if you factor in the Sin and Punishment talent, MF is better than MB on CD. However the chance of no shadow orbs procing @ end game gear's haste levels is very low. So let's assume each MB has at least 1 orb on it. After some math I learned that my crits were about 210% of the normal, and the 33% increase is on the initial damage then the crit is multiplied.

4055*1.33*640 + 4055*1.33*2.1*360 = 7528837.40 Damage / 1000*1 = 7528.84 DPS.

So arguably, wait till at least 1 Orb is up to MB, but most likely by making that extra effort to wait and look you will lose dps anyway.

EDIT: some abbreviations expanded.

So I think we can say with relative surety it is akin to clipping mind flay for DoT applications post 4pc t10. A theoretical dps increase, but in practical terms a dps loss with its high chance of error.

MB on CD regardless of orbs, and at worst you sacrifice a POSSIBLE 200-300 dps increase that comes with more risk than is probably worth it.

dont forget that we get a permanent 5% haste now while in shadow form. at 1243, 26 below the 8 tick mark, i'm getting 8 ticks while in shadow form.

Those ratings assume 3/3 Darkness and the 5% haste buff are there. I did test the 10% and 30% marks though, this is what I got:

670 haste rating - 7 ticks
660 haste rating - 6 ticks

69 haste rating - 6 ticks
42 haste rating - 5 ticks


My concern is testing the 50% (1269 rating with Mind Quickening and Darkness) mark. I don't have enough haste on my gear even with 20 haste in every socket to do this myself. I was hoping somebody with higher haste could help me out there and confirm that you get 8 ticks before 1269 haste rating.

In Heroic 10man tonight, I discovered that Mind Flay no longer procs any stacks of Instability from Unchained Magic on Sindragosa. I was able to continue spamming MF while I had the debuff with no ill effects. Has anyone else noticed this? I assume it's a bug, but I don't know how widespread it is. I asked our raid's warlock to test Drain Soul and Drain Life, which both applied an Instability stack on cast.

I noticed this last Wednesday night in H 25 ICC. Any other spell works as usual, but I could flay to my hearts content with no stacking effect. It may be related to flay no longer proccing DFO.

I can reforge my double haste trinks back to haste and be around 1500, what did you need tested exactly?

With shadowform and darkness, I tested 1256 haste (38.30% tooltip, 49.57% with talents) and had 7 ticks, and 1278 haste (38.98% tooltip, 50.3% talented) with 8 ticks. I couldn't narrow it down any closer without regemming to lesser quality gems, but that's a pretty narrow gap that seems to agree with 50% haste (1269 rating) being the cutoff point.

Basically what Itsumi just did in the above post. The cutoff point for VT's 8th tick, with 3/3 Darkness and Mind Quickening was 1269 according to the math I did. I just couldn't test it for myself in game due to my own lack of haste rating.

i am new to shadow priest so dont full understand how the tick calculation works for dots.
also can we abreviate MB with orb procs to MB0, MB1, etc, where the number is the amount of orbs present on cast?

as stated there is a dps increase from sin and punishment, and from the dps gain via MB due to orb procs.

the probability that a 1 orb will proc during a MF should be:
p= 1(3*1)*(0.18*1)*(1-0.18) ^(3-1) = ~36%
the probability 2 orbs will proc is ~8%
and 3 orbs is 0.5%

using your numbers and assuming 0 orbs are present on cast MF should be ~723dps more than you calculated (based only on dps from orb proc chance)
~700dps for 1 orb present on cast
~486dps for 2 orbs present on cast
and 0dps for 3 orbs present

not only this but the goal is maximum dps so there are further calculations needed. casting MB before 3 orbs have proced is a dps decrease for MB (i.e. MB is ~4000dps and MB3 is ~8000dps, using your numbers). the question is then does delaying until MB3 outweight the dps lose from using MF when MB is higher dps than MF.


Shadowy Apparition
I think the basic concensus on the rotation may be slightly off due to the increase chance you can have of procing a shadowy apparition when moving while casting instants.
The actual dps calculations for this are too complicated for me so i'll stick to an attempt and proc chance calculations.

Over 18 sec my SW: P ticks 8 times (tick every 2.25 secs). I have assumed over 24 sec period (DP duration) it will only proc 10 times (rounding down).
I am using a 24sec period assuming DP will be refreshed within that time and there is a 100% uptime on SW: P
GCD is 1.5 sec (?), or 66% the time of a SW: P tick.

chance shadowy apparition will not proc
not moving at all:
0.88^10 = ~28%
moving for 1 GCD (i.e. DP refresh):
(0.88^9)*(0.28*0.66 + 0.88*0.33) = 0.88^9*0.4752 = ~15%
moving for 2 GCD in a row (1 SW: P tick definitely occurs withing the movement):
0.88^9*0.28^1 = ~8%
moving for 1GCD seperate and 2GCD's in a row with 24 sec period:
0.88^8*0.28*0.475 = 4.7%
(edit: changed the proc chances to correct values)

I personally have been playing around with rotation (moving on every instant) - SW: P, DP, VT, etc
then SW: D only when i refresh DP. not sure this is actually a dps increase.
is the improved DP hit greater than SW: D at more than 25%hp?

really i hope this math isnt too horrible can someone more skilled do it?