SimulationCraft for Warlocks (Panda Edition) - Page 9

Jaofindel · 06/30/12, 7:57 AM

Also, when using Grimoire of sacrifice, you may want to soulburn a pet to get 2 stacks again(after refreshing dots and using haunt) to get 8~10 seconds of EXTREME Drain Soul damage.
i'm hitting 650k tick crits with the gear buff for heroic dungeons at lv 90

• Zakalwe · 06/30/12, 9:49 AM

Originally Posted by Jaofindel

Also, when using Grimoire of sacrifice, you may want to soulburn a pet to get 2 stacks again(after refreshing dots and using haunt) to get 8~10 seconds of EXTREME Drain Soul damage.
i'm hitting 650k tick crits with the gear buff for heroic dungeons at lv 90

Yep, this turned out to be about a 400 DPS gain.

In other news, the aoe profiles now use optimal aoe talents instead of just using whatever talents the profile is set up with for single target. This means destruction should catch up to affliction once the reports update.

zinnin · 06/30/12, 12:22 PM

Is affliction casting only Drain soul over MG because your modeling the nightfall bug in game currently, or because it is actually simming higher now?

• Zakalwe · 06/30/12, 12:26 PM

It's simming higher now, and I'm not modeling any nightfall bug. But it's probably only higher because they forgot to nerf it, like Ghostcrawler said in this comment that they were planning to do.

Karsteck · 06/30/12, 10:29 PM

Originally Posted by Zakalwe

In other news, the aoe profiles now use optimal aoe talents instead of just using whatever talents the profile is set up with for single target. This means destruction should catch up to affliction once the reports update.

There's something odd / buggy going on. The talent URL in the destro AoE profile is
Talent Specification - Game Guide - World of Warcraft.
(grimoire of supremacy), but the profile has uptime on sacrifice. (and pretty much no pet dmg, since it does get sacced at the start, so it's not affecting the results.) I'm guessing the URL output is wrong?

Also supremacy pets don't have their renamed abilities in the output. E.g. 2-target sims use a supremacy felhunter, but it's a "felhunter" doing "shadow bite", not an "observer" doing w/e it is. Presumably this is just cosmetic, and supremacy is just modelled as +20% pet dmg or w/e it is, but it makes it you click the talent link instead of just looking at the sim output to see what pet talent was chosen. (well, I guess you can tell from there not being a service:felhunter, and the pet not getting sacced, but still).

• Zakalwe · 07/01/12, 5:53 AM

I'm using the same profiles as for single target, but with a new talent override option that depends on the number of targets available - the URL just isn't updating to reflect that. In the future I'll make sure we either update the URL or otherwise indicate more prominently that the override option is in effect. Note, though, that the talents section further down in the report does show the correct talents.

As for the supremacy pets, as you say, that's purely cosmetic. I may get around to implementing them properly at some point, but it definitely doesn't have priority.

ultrajustin · 07/01/12, 1:59 PM

Has the imp scaling been updated in the last few tweaks?

My normal imp is doing 15% more dps than any other pet (as affliction) on the boss dummy with no debuffs/buffs up.
e.g. using recount letting them attack for a few minutes each, lv 90, 12.6k spell power
imp: 5.8k
felhunter: 4.2k

• Zakalwe · 07/01/12, 3:23 PM

Imp scaling is defined in the spell data, and yes, it did update when we updated our spell data to the latest patch. The felhunter is still a couple hundred DPS better at level 85.

It's certainly plausible that the imp overtakes the felhunter at 90 - the firebolt scales with 100% of spell power while the shadow bite scales with 58.9% of spell power.

Karsteck · 07/02/12, 5:42 PM

recent (edit: sorry, old news I guess) beta build warlock changes:

hellfire (while moving) instead of rain of fire (can maybe skip harvest life if siphon life will do more healing over the course of the fight than harvest life over hellfire). Or if you need to aoe while moving.
activating immo aura costs no fury, just fury over time.
leaving meta cancels immo aura. So probably need to choose something fury-efficient for meta uptime, like void ray, not harvest life.

There is still a ranged autoattack in meta, even though touch of chaos is what replaces shadow bolt. Haven't tested the mechanics.

Hellfire still doesn't stack will of unbinding.

• Zakalwe · 07/02/12, 5:53 PM

The latter two are not new in the latest beta patch, and simc already models it that way. Hellfire is not worth using because harvest life still generates 10 fury per target.

I put the ranged autoattack back in a while ago - I named it "melee" as per Ghostcrawler's comment.

netsach · 07/08/12, 3:31 AM

I'm trying to understand how to interpret some new things i've seen in the actions lists, and i'd need help :
- ticks_remain<add_ticks%2 (for agony)
- remains<=action.drain_soul.new_tick_time*2 (for all affli dots)
What do both of these mean and why these conditions ?

Thanks in advance.

• Zakalwe · 07/08/12, 3:37 AM

Isn't that the other way around - the first one is used for all dots and the second one just for agony?

Anyway, add_ticks is the amount of ticks that would be added at your current haste level, while % is the division operator, so roughly speaking ticks_remain<add_ticks%2 evaluates to true if you have less than half of the dot remaining. This is the point at which you can recast a dot while having the everlasting affliction glyph without wasting any duration.

And new_tick_time is the time per tick if the spell were recast at your current haste level. This is just to ensure that agony doesn't drop off at all, because if you let it drop off it starts again from zero stacks.

netsach · 07/08/12, 4:29 AM

Yep, it's the other way around, i made a mistake. I didn't know the "added duration" through everlasting affliction had a duration cap. Thanks for the heads-up !

turturin · 07/09/12, 2:26 PM

Does glyph of dark soul not provide a dps increase for affliction?

With glyph:
27% haste for 20 secs every 2 mins = 27%/6 = 4.5
+ 3% passive haste except when dark soul is active so 3%*5/6 = 2.5

average haste = 7%

Without glyph:
30% haste for 20 secs every 2 mins = 30/6 = 5%

average haste = 5%

I get that cooldown stacking should make the situation without glyph better than the 5% average, but is it really that much better?

I would also think that the the 3% passive increase would be particularly effective at entry level 90 raid gear, since haste will be relatively much lower than it is at 85 in heroic DS gear.

zinnin · 07/09/12, 4:47 PM

You don't get the passive effect when Dark Soul is on cooldown. So you are getting 3% average haste because if you are running the glyph you will not want to pop Dark Soul except for the last 20 seconds of a fight.