Demonology Thread - Page 47

brantar · 04/01/10, 4:29 AM

Originally Posted by turturin

I view it this way: a decimated soulfire has a DPCT almost as good as immolate. If the spells had equivalent mechanics, immolate would always be your priority (e.g. if both were spammable instant casts, you would always choose immolate). However, they are not equivalent in this situation. Immolate has an effective 24 second cooldown, while a decimation proc allows soulfire to be spammable within a set time period. As this time period is finite when procced from an add death, there is a significant opportunity cost to not casting soulfire while its active.

Lets assume your immolate ran out and needed refresh right as decimation procced. If you decided to refresh immolate, over the course of that decimation proc you will gain 1 immolate initial hit + 3 ticks. You lose the opportunity to cast 1 soulfire (at most haste levels). Since total immolate is only slightly better than 1 soulfire, we can safely assume 1 soulfire >> 3 ticks+ initial immolate dmg.

I go with only soulfires for add-based decimation procs. When decimation range hits the boss (and the uptime for decimation becomes 100% at that point) it again becomes worth it to prioritize immolate and corruption.

I first had the exact same thought as you, namely that I dont loose a whole immolate but only a partial one and that casting soul fire instead might be preferable. However I think you are treating it slightly incorrectly: You shouldnt really look at what has done most damage after the time the soulfire procc ended but instead you should look what has done most damage at the end of the fight(I am aware that this is not entirely possible since an unknown ending time of the fight obviously has an impact. However I will ignore this). To do this you should compare both possible casting sequences until a point where they are basically equal. Let's do an example: Suppose 6 seconds left on decimation, suppose soul fire cast time is 2 seconds, suppose immolate cast time 1 second, corruption cast time 1 second , shadow bolt cast time 2 seconds(these numbers are of course not realistic but that doesnt matter here).
With the soul fire prio cast sequence you will cast: soulfire - soulfire - soulfire - corruption - immolate and than go over into the usual shadow bolt spamming
With the second sequence you will cast: corruption - immolate - soulfire - soufire - and than start with the usual shadow bolt spamming. However observe that the two casting sequences only start to coincide after 8 seconds and not after 6, meaning that you have to compare what did more damage after those 8 seconds: The first casting sequence has 1 extra soul fire, the second has 6 seconds more uptime on corruption and immolate and an extra shadow bolt! It seems to me that the latter is basically always better since soulfire doesnt do that much more dps than shadow bolt to warrant loosing ticks on corruption and immolate.

Jenren22 · 04/01/10, 4:29 AM

I think it's fair to say you can leave Corruption. I would personally keep up Immolate though, as the above poster said, the 4 piece T10 proc I would say is worth every opportunity. I did go with just spamming Soul Fire, but I have now decided to keep up Immolate also. Corruption isn't worth wasting the decimation due its lower base damage, and despite MC having the ability to proc the actual trade-off from using a second to casting corruption and the RNG of getting MC during the few seconds of Decimation is not worth it over a single soul fire. My average soul fire hits for 9.5k, and crits for 22k. My average corruption ticks around 2k per tick. What would you prefer?

brantar · 04/01/10, 9:33 AM

Originally Posted by Jenren22

I think it's fair to say you can leave Corruption. I would personally keep up Immolate though, as the above poster said, the 4 piece T10 proc I would say is worth every opportunity. I did go with just spamming Soul Fire, but I have now decided to keep up Immolate also. Corruption isn't worth wasting the decimation due its lower base damage, and despite MC having the ability to proc the actual trade-off from using a second to casting corruption and the RNG of getting MC during the few seconds of Decimation is not worth it over a single soul fire. My average soul fire hits for 9.5k, and crits for 22k. My average corruption ticks around 2k per tick. What would you prefer?

I seem to have to repeat myself yet a third time: when deciding between casting or not casting corruption during decimation procc it is quite wrong to believe that you have to compare the damage corruption does to the damage soul fire does! What you effectively have to ask is: Does the damage difference between an extra soul fire and an extra shadow bolt warrant me loosing X seconds of corruption uptime(where X is the remaining time of the decimation procc)? The answer to that question in my opinion is basically always "no",i.e. casting corruption even during decimation procc is preferable. It only gets interesting it seems when X gets down to around 2 seconds. But please correct me if you think my logic is flawed.

turturin · 04/01/10, 8:22 PM

Where you're logic is flawed is that you're using made up cast time values, which isn't a problem in itself. Where it creates an issue in your analysis is that you've set the "made up" numbers to favor the DoTs you're arguing for. 1 second dot cast times and 2 second decimated soulfires are both directionally incorrect (in fact your estimate for DoTs should be higher and your Soulfire lower).

I don't mind if you don't want to use specific numbers to prove your point, but you need to do so with numbers that at the very least don't shade the bias toward the point you're trying to prove, and if using that argumentation method (as I have at times on these forums), you really should try to ensure that any bias you introduce really would work against your argument (i.e. you need to be much more conservative that you have been).

If you are still able to demonstrate it under those conditions I'd give it a lot more credence. As of now it basically says to me: "biased analysis based on made up numbers shows its bias in its results." (No offense intended of course)

brantar · 04/01/10, 8:56 PM

The argument has little to do with the cast times of the dots. However to convince you lets go back to the example and change the numbers to the opposite extreme: soul fire cast time 1.5 seconds, corruption cast time 1.5 seconds, immolate cast time 1.5 second, shadow bolt cast time 2 seconds. assume 6 seconds on decimation left.
Cast sequence with soul fire prio goes: soulfire-soulfire-soulfire-soulfire-corruption-immolate- shadow bolt spam
Cast sequence with dot prio goes: corruption -immolate - soulfire-soulfire - shadow bolt spam
Both sequences are basically identically after 9 seconds. Sequence 1 has 2 more Soul fires, sequence 2 has 1.5 more shadow bolts and 6 seconds more uptime on both corruption and immolate. Still the latter seems preferable.

turturin · 04/01/10, 9:22 PM

brantar · 04/02/10, 11:43 AM

Interesting, it seems I have underestimated soul fire. I obviously should have checked those numbers myself but was too lazy. Since both examples were chosen with unrealistic casting times and since both extremes give different results the only solution is to redo the calculation with more realistic numbers in order to see where the "break even" point is:

The DPE over DPET ratio for both corruption and Immolate suggests that assuming a cast time of 1.1 seconds for each seems to be reasonable. Let us furthermore assume there are X seconds left on decimation. With the soulfire prio cast sequence we would gain 2.2 seconds extra soulfires, with the corruption and immolate prio cast sequence there would be 2.2 extra seconds worth of shadow bolt cast time and X seconds more uptime on immolate and corruption. The DPET of soul fire is 12748, the DPET of shadow bolt is 9435. Thus the extra damage of casting 2.2 seconds of soulfire to 2.2 seconds of shadow bolt is: (12748-9435)*2.2=7288. Since I am not quite sure how to interpret the 3 different values for tick and crit tick in the simcraft output, I will instead just use the pdps values for corruption and immolate given in the output(the true dps of both would be slightly higher since pdps is equal to true dps of the spell only if uptime would be 100%). Thus the combined dps of immolate and corruption together is bounded below by 1116+933=2049. Thus it will be preferable to cast soulfire if: 7288 - 2049*X > 0 , i.e. X < 3.5 seconds. It should be added that this calculation is neglecting the extra dps gain from potential extra molten core and t10 4pc procss, including those would lower X even further. Thus the moral of the story is: If decimation is basically just about to run out casting 1 maybe 2 soulfires might be preferable, however if there are more than 3 seconds left you seem to be better off casting corruption and immolate first.

Looking at the simcraft output raised another question: Soul fire has DPET equal to 12748. Corruption has DPET equal to 9943. Thus again assuming a corruption cast time of 1.1 seconds it means that we loose (12748-9943)*1.1=3085 damage each time we cast corruption during the last 35% of boss health. Is molten core really that strong to make up for that or am I reading the simcraft data incorrectly?

Warlocomotif · 04/02/10, 12:48 PM

You are reading it correctly and molten core is that good according to results, I however strongly recommend that if you have molten core already up during decimation that you hold off on refreshing corruption.

brantar · 04/03/10, 8:16 PM

Originally Posted by Warlocomotif

You are reading it correctly and molten core is that good according to results, I however strongly recommend that if you have molten core already up during decimation that you hold off on refreshing corruption.

That is an interesting result, which was unexpected to me. I quickly did a rough calculation of the damage gain from molten core during execute range and came to the same conclusion. The calculation goes as follows:

I am using the numbers given in the simcraft output here. The DPE of Soul fire is 20485. From this one can extrapolate the base DPE of soul fire X, i.e. the damage it does when it doesnt crit and also when not under the effect of molten core: 20485 = 0.72 * (0.468*X + 0.532*2.03*X)+0.28*(0.318*X+0.682*2.03*X)*1.18 , i.e. X=12215.1 (here it was assumed that the 28% uptime of molten core was uniform, i.e. that it had uptime of 28% for all soulfires). As a first approximation I only calculated the minimal damage gain from molten core: The probability of molten core proccing for at least one of the six corruption ticks is 1-0.88^6=0.54 and this gives at least 3 molten core empowered soul fires. Thus the damage gain from this is bounded below by 0.54*3*(1.18*(0.318*12215.1 +0.682*2.03*12215.1 )-20485)=6567.4. Thus the average damage gain from incorporating corruption during the last 35% of boss health is at least 6567.4-1-3085 =3481.4 damage over the duration of one corruption. With corruption lasting 18/(1.56)=11.5 seconds this means a dps increase of at least 3481.4/11.5 =302.7 dps during execute phase or equally 302.7*0.35=105.9 dps increase over the whole fight.

The idea of not refreshing corruption during decimation if moltern core is up sounds intriguing and intuitevely makes a lot of sense. Sim crafting this altered priority list should be easy, however if I find the time I will see whether the whole thing can be reasonably approximated by hand.

Indaria · 04/06/10, 11:08 AM

Remember that with a corruption lasting 11.5 seconds it will tick at roughly ~1.91 seconds intervals. Since this is longer than your soul fire cast time you should definitely refresh corruption before using up the 3rd charge of soul fire to increase uptime.
This way it will tick just after your soul fire cast is finished, with no chance of a molten core proc overwriting your old molten core.

This assumes you refrain from refreshing corruption during molten core procs as suggested above.

Demi9OD · 04/06/10, 11:39 AM

Originally Posted by Indaria

Remember that with a corruption lasting 11.5 seconds it will tick at roughly ~1.91 seconds intervals. Since this is longer than your soul fire cast time you should definitely refresh corruption before using up the 3rd charge of soul fire to increase uptime.
This way it will tick just after your soul fire cast is finished, with no chance of a molten core proc overwriting your old molten core.

This assumes you refrain from refreshing corruption during molten core procs as suggested above.

Even at a 1 second GCD your soul fire cast time would have to be .91 seconds to prevent a corruption tick before soul fire finished casting.

Poppaperm · 04/07/10, 9:00 AM

On a side note, there is a very decent addon that has started going around that monitors how much your DP is up for so you can quickly look at it in a fight. It also makes it easier to decide if it is worth pulling your FG for teh raid so it lasts the full 45 seconds.

DemoPact Addon

WoW Warlock Forums

This addon was written by Ënmity.

• bartolimu · 04/07/10, 10:51 AM

Ladies and Gentlemen, let this serve as a public service announcement.

Warlock Cataclysm Preview Discussion - READ THE FIRST POST

Let's keep the rest of the Warlock forum as readable as ever, please.

Nerrun · 04/19/10, 8:58 AM

a short question: we have a new lock in the raid, which playes also demo. Is there a disadvantage if 2 demo locks are in the raid? for example that one doesn't benefit correct from demonic Pact?

Rockmonster · 04/19/10, 12:07 PM

Originally Posted by Nerrun

a short question: we have a new lock in the raid, which playes also demo. Is there a disadvantage if 2 demo locks are in the raid? for example that one doesn't benefit correct from demonic Pact?

You should always just have 1 demo lock in raid, the Demonic pact does not stack and you are losing dps by having 2 when one should have gone to another spec.

Unless DP has changed it would take the higher locks spell power unless it was not refreshed and then drop to the lower locks dp.

Because of the changes to DP however and its longer uptime you will probably just have the under achieving dps of the 2nd demo.

ultrajustin · 04/19/10, 6:19 PM

Originally Posted by Rockmonster

You should always just have 1 demo lock in raid, the Demonic pact does not stack and you are losing dps by having 2 when one should have gone to another spec.

Unless DP has changed it would take the higher locks spell power unless it was not refreshed and then drop to the lower locks dp.

Because of the changes to DP however and its longer uptime you will probably just have the under achieving dps of the 2nd demo.

The DP of the past refreshed despite the spell power of the current one -- edited this post in light of the information in the post below.

Bahlshaab · 04/20/10, 9:21 PM

Originally Posted by Madlax

I'm wondering how multiple DP will be handled now by the game.
Technically they should have removed the "weaker DP overrides stronger DP" now with this change - but can anyone confirm that from the US?

A guildmate recently switched to 56/15, giving us two demonic pact warlocks in our 25dom runs (we run on separate 10dom teams). I've been browsing our logs for solid information on mechanics and came up empty until this interesting Professor Putricide fight. It caught my eye because there was a big downtime for demonic pact, which has rarely happened. Here's my best consolidation of interesting demonic pact mechanics from that fight. For anyone not interested in the details, I will sum-up by saying the mechanics are messed up & in some cases can actually prevent DP from proc'ing correctly. So I concur with others who've already recommended against having two demonic pact warlocks in one raid.

Here's a link to the filtered WoL ICC25 log I used; expires May10.

Note: Not every sequential crit and proc/refresh is shown below, only those that serve as pertinent data points. Entries indicated as "No overwrite" pertain to the most recent proc/refresh listed.

Assumptions:
(1) The two demo locks had raid-buffed spell power within 150 of each other; constant spell power/spirit buffs benefited the two locks comparably and are folded into the 150.

(2) Dynamic spell power buffs in play are only those shown below, therefore the trinket procs are the primary determinants of whose spell power is higher at any given time.

Limitation:
There's no record of what Blizzard thought the two warlocks' respective spell powers were at any given time, and unfortunately those are the only numbers that matter.

Abbreviations of Dynamic Spell Power Buffs:
EP = Elusive Power (via [Abyssal Rune])
DC = Dying Curse (via [Dying Curse])
FS = Frostforged Sage (via [Ashen Band of Endless Destruction])
CP = Cultivated Power (via [Muradin's Spyglass]; stacks less than 10 are indicated)
LT = Glyph of Life Tap

Entry	Event	Demon	EP	DC	FS	CP	LT	Crit Time	DP Proc Time	CD-until	DP-until
1	First Gain DP	Keevegen				X	X	[22:45:24.479]	[22:45:24.816]	45:44	46:09
2	No overwrite	Flaatom		X	X	0	X	[22:45:26.076]	no	off
3	Refresh	Keevegen				X	X	[22:46:00.245]	[22:46:00.971]	46:23	46:48
4	No overwrite	Flaatom				X	X	[22:46:01.860]	no	off
5	No overwrite	F				X	X	[22:46:05.230]	no	off
6	Overwrite	F				X	X	[22:46:06.836]	[22:46:07.413]	46:27	46:52
7	Refresh	F			X	X	X	[22:46:32.103]	[22:46:32.795]	46:52	47:17
8	Gain: Shadowfiend Only	Keevegen				X	X	[22:46:44.386]	[22:46:45.561]	47:05	n/a
9	Crits wasted	K				X	X	5 crits	on CD	47:05
10	No overwrite	K	X			X	X	[22:47:06.856]	no	off
11	No overwrite	K	X		X	X		[22:47:13.297]	no	off
12	DP Fades [22:47:17.778]	Flaatom					n/a	21 non-crit hits	DP FADES
13	Should have proc'd	Keevegen				X	X	[22:47:22.983]	no	off
14	No proc?	K				X	X	[22:47:42.238]	no	off
15	Gain DP	Flaatom				X	X	[22:47:42.238]	[22:47:42.577]	48:02	48:27
16	No overwrite	Keevegen	X			0	X	[22:47:57.055]	no		off
17	No overwrite	K	X			1	X	[22:47:58.774]	no	off
18	No overwrite	K	X			3	X	[22:48:01.096]	no	off
19	Refresh	Flaatom			X	0		[22:48:05.151]	[22:48:05.870]	48:25	48:50
20	No overwrite	Keevegen			X	X	X	[22:48:18.397]	no	off
21	No overwrite	K			X	X	X	[22:48:19.739]	no	off
22	Refresh	Flaatom				X	X	[22:48:30.397]	[22:48:30.775]	48:50	49:35
23	Refresh	F				X	X	[22:48:58.524]	[22:48:59.255]	49:19	50:04
24	No overwrite	Keevegen	X			X	X	[22:48:59.144]	no	off

Take-home messages:
(1) All cases of "No Overwrite" show that crits outside of CD do *not* necessarily overwrite

(2) Entries 2, 16-18, and 24 show that a crit plus a (projected) higher spell power value at the time of crit does *not* necessarily overwrite

(3) Entry 6 shows that overwrites can happen under apparently identical conditions to when they previously did not happen

(4) Entry 8 shows that a proc was wasted on a temporary summoned minion, failing to proc the buff on the rest of the raid (no overwrite) and putting the demon needlessly on CD. This is a mechanic unique to the two-DP-lock situation, i.e. a single DP-lock would either proc for the whole raid or no one.

(5) Entries 10-12 show the worst problem with the failure to overwrite. Due to RNG the demon providing the buff failed to crit in 21 hits, during which time two crits by the other demon failed to overwrite. Result: demonic pact faded. Arguably this could be considered more of an absence of benefit of having two DP locks, as DP would have faded if there was no second lock at all.

(6) Entry 13 shows that even when DP has faded and a crit occurs, DP does not necessarily proc. Result: demonic pact was down for a full 24 seconds *even though one of the demons was off CD and critted*. This is reminiscent of the old Prayer of Spirit bug, but I do not think that's the problem here as I've seen a marked lack of clear evidence of that bug since the patch.

(7) An awesome question raised by my guildmate Lokrick: could the internal cool down be triggering even when a proc/overwrite does not take effect? An 'attempted' but unsuccessful proc would not show up in the logs. All I have right now is evidence that overwrites do not necessarily occur even when there is no possibility of being on CD: Entry 2 for Flaatom (it was his first crit of the fight so he was not on CD), and Entries 10 & 11 show Keevegen's two crits after coming off of CD (from proc'ing on the Shadowfiend). There is no overwrite.

Reeshet · 05/04/10, 1:21 PM

Originally Posted by Bahlshaab

13 Should have proc'd Keevegen X X [22:47:22.983] no off

<snip>

(6) Entry 13 shows that even when DP has faded and a crit occurs, DP does not necessarily proc. Result: demonic pact was down for a full 24 seconds *even though one of the demons was off CD and critted*. This is reminiscent of the old Prayer of Spirit bug, but I do not think that's the problem here as I've seen a marked lack of clear evidence of that bug since the patch.

When you're doing all this in depth analysis of the logs, you would be well served to consider other encounter mechanics before assuming this is a bug that no one but you managed to observe.

[22:47:19.660] Professor Putricide begins to cast Tear Gas
[22:47:22.511] Professor Putricide Haunt Kryptkeeper +1832 (O: 1839)
[22:47:24.195] Vyndea afflicted by Tear Gas from Professor Putricide
[22:47:24.195] Rafian afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Zut afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Sorei afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Ganogra afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Razzully afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Locia afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Cheetahsturq afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Amalgahìde afflicted by Tear Gas from Professor Putricide
[22:47:24.270] Kryptkeeper afflicted by Tear Gas from Professor Putricide

Although the crit at [22:47:22.983] is 1.212 seconds before Tear Gas afflicts the raid, if you check the logs, you'll see it's quite common for there to be a 1-ish second delay between when the demon crits and when the raid buff is refreshed.

Tear gas fades at [22:47:40.331] and lines 14 & 15 happen at *exactly* the same time at [22:47:42.238] which considering he's coming back from the table on side of room is probably pretty much immediately after he's attackable.

As far as the buffs not being overwritten, without tracking exactly how much spellpower was in the Demonic Pacts, I don't think you can make any valid determinations on if it should have proc'd or not. The warlocks DON'T have identical spell power (per your assumption #1) and you DON'T know what Blizzard thinks their spellpower was at the time of each crit.

Bahlshaab · 05/06/10, 5:24 PM

Demonic Pact Proc Times:
The typical proc time in the copious logs (of mine) I've combed through is 0.5-0.7 second. I have not seen proc times of 1 full second or higher (others' latency may vary). Therefore I doubt that a >1.2sec proc time is a plausible explanation, excepting an extraordinary coincidence. It's an interesting conjecture though, and I appreciate being reminded of fight mechanics I need to watch for in my log analyses (such as tear gas). Another thought it brings to mind is the question of latency and whether the time stamps always represent the timing of events server-side; perhaps some one knowledgeable would comment.

Here is another case of DP not proc'ing (I have a handful of these), this time in a Valithria fight, where there are no stun/phase-change effects that would prevent a proc, that I know of. As before, the existing DP faded and two subsequent crits did not proc a new DP. Flaatom had not proc'd a DP yet in the fight and so could not have been on CD.

WoL ICC 25 Apr18
[23:52:52.863] Keevegen's Demonic Pact fades
[23:52:53.220] Flaatom's Demonic Pact fades
[23:52:53.744] Flaatom crits Gluttonous Abomination *2867*
[23:52:58.871] Flaatom Cleave Gluttonous Abomination *3488*
[23:52:59.210] Keevegen crits Gluttonous Abomination *2067*
[23:52:59.687] Keevegen gains Demonic Pact from Keevegen
[23:53:00.039] Flaatom gains Demonic Pact from Keevegen

I haven't yet spotted a non-proc in a one-DP-lock raid; I'm still looking.

Spell Power & DP Proc Mechanics:
Expected DP procs based on spell power are quite straightforward to determine -- that was the point of supplying the SP data in my original post. At any given time, total spell power comes from gear + constant buffs + dynamic buffs.

Here's how:

1. Determine the difference in raid-buffed spell power* values between the two warlocks. I did this using SimulationCraft:
Warlock K raid-buffed spell power (including glyph_life_tap) = 3942
Warlock F raid-buffed spell power (including glyph_life_tap) = 4038

*Raid-buffed spell power here means gear + all constant buffs (including pets, food, flask, those from optimal_raid, and in my simulations I apply glyph_of_life_tap before the snapshot is taken).

Note that the raid-buffed values account for all multiplicative effects on SP, unless I am overlooking something. The dynamic buffs are all additive, and this is why the subsequent calculations can be based on the *difference* between the two base SP values [4038 - 3942 = 96] rather than requiring the values themselves. Examples of both approaches are shown in the steps below. In my original post I gave a difference of 150 SP between the two locks to be extra conservative (the larger the difference, the larger the potential impact on whose final SP is greater).

2. Confirm that constant buffs relevant to SP that are listed in the log match those used by SimC and are applied to both warlocks (e.g. Totem of Wrath and Flametongue are disabled in my simulations).

3. Add in dynamic spell power buffs. According to the log, the following dynamic buffs were active:
(a) at 22:45:24:816 (time of DP proc in Entry 1 from table above)
Warlock K: 3942 (incl LT) + 180 cultivated power = 4122 SP

(b) at 22:45:26:076 (time of crit by Demon F in Entry 2 from the table above):
Warlock F: 4038 (incl LT) + 285 frostforged sage + 765 dying curse = 5088 SP

Again, note that dynamic SP buffs are additive. This means that instead of using the actual raid-buffed SP values, the relative values of zero and 96 could be used as the base values to which dynamic buffs are added. This is because we do not need to know the actual SP numbers, we only need to know whose final SP is greater.

For example (see Entries 1 & 2 from original table):
Warlock K: 0 (base) + 180 cultivated power = 180 SP
Warlock F: 96 (base) + 285 frostforged sage + 765 dying curse = 1146 SP

Note that (1146 - 180) = 966 = (5088 - 4122).

For the cases highlighted in the original post (Entries 2, 16-18, 24) the outcome (i.e. the determination of whose SP is greater at the time) is independent of which warlock starts with zero versus 96 because either way the difference of 96 is very small compared to the trinket procs of 765 SP for one warlock or 590 SP for the other warlock. For the same reason, using the approximated 150 base value provided in the original post leads to the same outcome of whose SP is greater for the highlighted cases.

4. Draw a conclusion. In this case, Warlock F was at a substantially higher spell power at the time of crit and so DP should have been overwritten *if higher spell power necessarily wins*.

Constructive feedback, such as specific suggestions as to what contributions to SP may be missing from, or incorrectly accounted for, in the above calculations would be great.

5. Toss in some what-ifs (multiplicative calculations require actual SP values).
(a) What if DP was erroneously included in Warlock K's spell power? Then:
4122 + [10%(4122)=412] = 4534, still less then Warlock F.

(b) What about the 5% bug that boosts the first application of DP? (disclaimer: I have not paid careful attention to the details of this; I'm throwing in an extra 5% anyway, and applying it *after the initial 10%* just to get the largest number possible). Then:
4534 + [5%(4534)=227] = 4761, still less than Warlock F.

As noted in the original post, not knowing the SP values that Blizzard is using prevents us from knowing with certainty how the proc mechanic is actually working. It does *not* prevent us from determining whose SP value should be highest at any given time and then seeing if there are apparent inconsistencies with the proc's in-game.

Warlocomotif · 05/06/10, 7:38 PM

It would proabably be worth while to have 2 (grouped) demo locks mash on an exodar dummy for few a while. That kind of parse would be much, much easier to read and go through.

On another subject; this thread has been outdated since 3.1.2, and is almost entirely out of date right now. I have attempted to contact the OP but he doesn't seem to be around anymore. I will most likely write a replacement for it, and would like if anyone with suggestions would send them to me.

Oxylos · 05/14/10, 7:51 PM

Decided to switch to demonology today and read through the thread and wanted to try to consolidate the information since the OP is outdated and theres been some recent arguments.

Spec/Glyphs: Talent Calculator - World of Warcraft With a possible switch from IDT into Suppression if necessary.

Spell Rotations

Normal:
1. Life Tap [Life Tap Down]
2. Shadow Bolt [ISB Down]
3. CoD
4. Immolation Aura [Meta Up and in range]
5. Immolate [Immolate Down]
6. Corruption [Corruption Down]
7. Incinerate [Molten Core Up]
8. Shadow Bolt

Execute:
1. Life Tap [Life Tap Down]
2. Shadow Bolt [ISB Down]
3. Immolate [Immolate Down]
4. Corruption [Corruption AND Molten Core Down]
5. Soul Fire

Gear choices: 4pt10 is great. Failing that, 2pt9 is useful.
Question: Spirit vs. Crit on equivalent pieces? e.g. Crushing Coldwraith Normal vs. Circle of Ossus?

Sorry for the non-contribution, just wanted to make sure I'm (and perhaps anyone else coming to this thread like I was) are Doing It Right as of today! Any corrections greatly appreciated, thanks.

Warlocomotif · 05/14/10, 8:19 PM

That's pretty much correct. For personal DPS you would give minor priority to crit, for raid DPS you would give significant priority to spirit. As for immolation aura, it's usually a good idea to try fish for some kind of buff for it (ideally 4xT10, but trinket procs and alike are also very good)- so long as you cast it before there's less than 15s left on metamorphosis everything is fine.

sckeener · 05/20/10, 12:04 PM

Originally Posted by Oxylos

Execute:
1. Life Tap [Life Tap Down]
2. Shadow Bolt [ISB Down]
3. Immolate [Immolate Down]
4. Corruption [Corruption AND Molten Core Down]
5. Soul Fire

during the <35% period on the target, you are only keeping Immolate up for the 4t10 proc and corruption should only be cast if you are having to move during that period. If you do not have 4t10, then you can skip 3 & 4 during the <35%.

as for #2, I'm unsure if that is an improvement for personal dps. It obviously is for raid dps.

Tinava · 05/20/10, 1:13 PM

Originally Posted by sckeener

during the <35% period on the target, you are only keeping Immolate up for the 4t10 proc and corruption should only be cast if you are having to move during that period. If you do not have 4t10, then you can skip 3 & 4 during the <35%.

Are you sure on that one (don't cast corruption unless moving). Molten Core gives you a significant boost to your soul fire at under 35 percent, and I'd think you would want to get that in as much as possible.

nuibank · 05/20/10, 2:11 PM

Immolate is always used because it's one of the higher DPCT spells demonology has. The T10 4piece is just a bonus. Corruption should be used to get molten core procs, even under 35%.