That's pretty strange... unless I'm misunderstanding you, that means you can get 4 incinerates (or soul fires) off during each molten core proc?

Yeah been testing it on PTR, and currently you can do 4 hasted incinerates.

Just also realized that the Simcraft meta / ruin specs has 2/3 MC, with current change it might be worth it to give up the point on demonic empowerment for 3/3 MC for 3.3 PTR specs for now like Warloco mentioned.

I'm guessing that'll get fixed - simcraft only does 3, and I'll leave it at that for now.

I actually made that change already, just forgot to edit the talent link in the OP.

Does the DoT part of the new Conflag crit for 200% if you are specced into ruin? I'm currently having problems to log in the PTR and can't do the check myself.

Another interesting thing to note, is the wording on Molten Core indicates that each tick of Corr can proc the buff. If that is the case, might not the Glyph of Quick Decay actually be a higher DPS boost than other glyphs due to the higher molten core uptime?

Been thinking of that too, the downside would be having to recast it more often too, and ofc losing another glyph. Could still be worth it though, especially now when its not a fixed 12 sec duration on the proc, but charges instead.

Only when you have MC up and only 3 incins! after that is should be just normal SB-spam!

Some time in the past 2 months, simulationcraft started modeling FG with 1.7% melee and cleave miss. I just reviewed some of my recent Anub'arak parses and over 2000 attacks I had 0 misses. Over that many attacks I should have had at least one miss at nearly a 2% miss rate with the FG, so I think there is an error in simulationcraft modeling of FG melee/cleave introduced somewhere between r3254 and r3608.

Also, the 0/41/30 spec should be casting curse of doom now that Curse of Agony apparently does not proc molten core, per the description (only activated by corruption). This will probably lower the output of 0/41/30 significantly since it was utilizing a higher proc rate on molten core due to the old mechanics of any shadow damage activating molten core. 0/56/15 will also see a lower proc rate since shadowbolt will not activate it. Since the 0/41/30 profile has a 50% higher proc rate on molten core than 0/56/15, it is safe to assume this is unimplemented in the numbers in the OP.

I'm not sure if it's safe to assume that MC will be better than Demonic Empowerment for a meta spec. When I modeled casting only immolate, corruption, soul fire, and incinerate,molten_core=1 to avoid extraneous molten core procs from shadow bolt, I saw a DPS increase of 50 dps to use 2/3 molten core and 1/1 empowerment in r3608 (I know this is a terrible test, but I'm trying to show the relative values of the talents without re-coding molten core).

With the recent Molten Core changes, have any of the sim wizards taken a look at a 41/30 Imp Build or perhaps a variation that incorporates Conflagrate somehow?. I'm sure the new Quick Decay glyph would help out MC uptime, but I'm unsure with this type of build as to which glyph it would replace: Immolate, Incinerate or Life Tap (or Conflagrate if you went the 40/31 route).

I know with Demo/Destro hybrids we've normally looked at the Felguard, but the Imp's 5% crit with fire spells via Master Demonologist looks tempting for this build (bringing the total to 17% crit with fire spells via talents alone). Immolate, CoD, Corruption, and spam Incinerates till 35% then light up the Soul Fires. The imp can't be dodged or parried and requires a lot less micro management as well. Depending on how much Molten Core uptime you can generate with Quick Decay, you may even be able to get away with using a Firestone for this type of build also.

Swapping imp for felguard will never be higher dps in a 0/41/30 spec. The felguard does 1000 dps more than the imp at t9 gear levels, so you would need to come up with 1000 dps from 5% crit and DE. You would get roughly 200 dps from DE with the new imp crit mechanics, but you can't get 800 dps from 5% crit unless you've got very low crit to start with and you're doing in excess of 16000 dps with spells that can crit. You're much better off using 0/40/31.

One thing worth pointing out that I observed in the PTR patch notes summary:

Warlocks: Shadow Mastery, Pandemic (change reverted), Molten Core, Conflagrate

The pandemic change (Shadow Bite now crits for 100% extra damage)) is now gone. This means Shadow Bite should be critting for only 200% now, down from the 250% it was doing last build. I think the front page numbers aren't taking this into account.

There has been a lot of talk recently about local maxima in haste values. I ran simcraft for my profile (not BiS by a mile, and ignoring my real gear haste) and generated scale factors with an artifical haste ranging from 300 to 800. There are, for me, a couple of humps in the line. (at 410, 460, 540 - then a plateau from 570 to 620. It is a bit bumpy. My simcraft line was

simcraft.exe input=haste.simcraft max_time=315 calculate_scale_factors=1 iterations=15000 threads=2 smooth_rng=0

The profile stub I used was:
I ran this a grand total of once, I'll prolly rerun it tomorrow to see how much variation there is within a 15k pass.

I work better with visuals, so I put the above data into a table. I left out Hit and DPS so we could focus on the 0-2.0 scale factor range.

The black line is the linear progression of haste



I expected to see more of a relation between the haste dips and crit/sp. I believe this is just the point at which haste lines up your spells so imperfectly that Simcraft spends less time casting Conflag and less Immolate uptime.

--------------------

Would there be any merit to altering the Latency calculation to use a random latency value every 10 seconds or so, between two set values? I know my latency fluctuates between 90ms and 120ms at any given point, and this may or may not have an impact on the above chart.

Yeah, Gherkin, a graph would have been more clear but I couldn't be arsed to get an imageshack/photobucket account and figured that the raw data would be more useful. A graph of the actual DPS is probably what you want to be looking at rather than the scale factors. I m not surprised at the SPI/CRIT lines being relatively smooth since they were held constant as I varied the haste. You can think of the haste line in that plot as the derivative of the DPS line so you probably, for what ever your gear level is, want to tweak for more or less +haste around one of the peaks on the RHS.

As for latency, you can think of it as a negative modifier to haste and will have the effect of making a noisy shift to the left of the entire line.

No, because you won't be able to isolate this as latency effects alone. It takes between .15 and .3 seconds on most keyboards to register a click of a button from the moment you press the button to the moment your system registers the depress of the key (which is the "real" trigger for an action). Because you can't effectively model variation in either users keyboard response times or the timing and/or consistency of their button depresses, you're not going to get anything instructive out of the latency fluctuation modeling you suggest.

Did you check to make sure conflag dots can even crit? The most recent patch notes appear to be for a build we don't have yet.

This is the same issue Im having, I did hop on test last night, but at that time it didnt seem these changes were live yet. Conflag dot was unable to crit and was still at 6 seconds not back down to 3 yet.

I've been considering an alternative to scale factors: Generate raw dps data points for a handful of stats in a gnu-plot format. Allow people to use different visualization tools: 3d surface plots, etc

Something like this?



Kinda reinforces the "no real sweet spot" rule of thumb for me. Yeah, there are peaks but they're pretty tiny.

Yeah duffry, exactly like that. But in a forum like this where people are concerned about +/- 10 dps, they might care. Just throwing it out there. I will repeat the runs a couple of times tomorrow to see if there is run over run variability. I will also do a BiS run to see what effect that has (( I predict very little )). I'll include SPI data in the output as well for those who asked.

Because if it crits independently its probably not going to get the massive talented crit chance from our talents that raise the crit chance of conflagrate. Its especially worrisome if it doesnt proc pyroclasm that leads me to believe the dot portion is all on its own in terms of how and what actually effects it. Sounds like the spec would still lose a little dps there, just not as much.

I'm going to guess that the Y-Axis is DPS and the X-Axis is Haste? A graph without labels can be pretty useless.