SimulationCraft for Warlocks (4.0.3a numbers at 80 and 85) - Page 5

• Zakalwe · 02/27/09, 3:05 AM

Originally Posted by Morrigan

By the way, I don't think the way SimCraft calculates Conflagration damage is correct. It seems to use the Immolate tick damage after all modifiers, but actual tests show that the Immolate glyph and talents do not increase Conflag damage, which should bring the Decimation/Conflag build down to more reasonable numbers.

Hmm, this is pretty significant if it's true. Based on the wording in the talent tooltip, I implemented the conflagrate damage by dynamically calculating a tick of immolation and multiplying it by five. If I'm understanding you correctly, you're saying I should have just set the base conflagrate damage to the base immolate tick damage times five?

Do you have a source? A link to a description of the tests you mention?

phulshof · 02/27/09, 3:13 AM

How will T8 gear influence the class balance we see in the data now? Do all builds scale equally well with improved stats?

brymortis · 02/27/09, 3:19 AM

Okay I did some brief testing with the 40/31 spec. At under 35% I found that dropping corruption seemed to have boost my dps significantly (Maybe its just how I am casting thats not making corruption work into the rotation). I also gave conflag first priority to maximize the glyph usage.

xaoc. · 02/27/09, 4:11 AM

You're right, I've tried everything and ... here I present to you the currently highest DPS spec: 0/40/31 with Imp, Conflag und Immolate glyphs (yes, the new Imp glyph is ~20 dps higher than the Incinerate glyph)

@ Morrigan

Did you test running MC 1/3,2/3 putting the last point into Demonic tactics or Pyroclasm? (though crit rate would definatly favor tactics with lower gear) Or where you focusing on making sure uptime was higher based on coa/corruption, etc ? Just wondering because I didn't see if you dropped corruption or not.

Orgath · 02/27/09, 4:56 AM

Originally Posted by Zakalwe

If I'm understanding you correctly, you're saying I should have just set the base conflagrate damage to the base immolate tick damage times five?

Do you have a source? A link to a description of the tests you mention?

3.1 Patch notes discussion

Post 311 and down.

Dawei · 02/27/09, 5:05 AM

Originally Posted by silmarieni

Zakalwe,

Can you please test this spec?

http://talent.mmo-champion.com/?warl...0&version=9614

It is your 0/41/30 Felguard/Emberstorm except 3 points were added in suppression and 3 points were taken from emberstorm.
BTW, would this be the optimal way to get the 3 points for suppression or should they be taken from elsewhere?

Thanks a lot,
Sil

Bad idea. Lets do some napkin math to see why. If you are 1% away from the hit cap, then 1% of the time your spells will not land. This is more or less equivalent to missing 1% of you dps. So that means each point put into Suppression is a 1% damage increase until you are hit capped. Compare that to each point in Emberstorm. A single point is a 3% increase in fire spell damage. Lets say fire damage is making up 50% of your damgage and the rest is from your pet and shadow spells(In reality, fire will probably make up much more than 50% of your damage). That means a single point in Emberstorm is a 1.5% dps increase in dps. And that doesn't even account for the Incinerate cast time decrease you get from Emberstorm. Now that was some pretty rough math that is not completely accurate but I believe it would be better to miss out on the hit cap than take points out of Emberstorm. Emberstorm is half the reason that specs even works as well as it does. Like I said, the math was rough so feel free to correct me if I did something dumb.

• Zakalwe · 02/27/09, 5:35 AM

Ok, I finally had a chance to sit down and fix some bugs, adjust some specs/glyphs/rotations, and run a bunch of simulations. The results are reflected in the updated original post.

The major changes were:
* Simplified/fixed Decimation triggering
* Conflagrate changed/nerfed to match PTR testing
* All specs using fire_stone switched to spell_stone

I also made several small adjustments to specs/glyphs/rotations based on suggestions in this thread.

• Zakalwe · 02/27/09, 5:45 AM

Some observations:

It turns out 0/41/30 beats 0/40/31 by a little bit - this is because the Felguard does 400 more DPS than the Imp, and (when correctly implemented) Conflagrate can't quite catch up.

Deep destruction seems to still be sub-par, sadly. I haven't played destro since BC, so if anyone can spot any way to improve the rotation, please let me know. Should any particular spell priority be changed during Backdraft?

If it turns out that Decimation rotations involving 1xFiller-1xSF are viable in a raid situation, there's still room for significant improvement in the demo specs.

Warlocomotif · 02/27/09, 5:51 AM

Originally Posted by Morrigan

You're right, I've tried everything and ... here I present to you the currently highest DPS spec: 0/40/31 with Imp, Conflag und Immolate glyphs (yes, the new Imp glyph is ~20 dps higher than the Incinerate glyph)

I'd venture to say that glyph of incinerate will beat glyph of immolate. You have roughly 4 times more incinerate damage than immolate damage, and even if you disregard the damage loss to the initial damage (4x5%=20%) that would be a pretty even match up.

Still disregarding the damage loss to the direct damage:
12% / 1.2=10%, 47% * 1.05 = 49.35, making incinerate (before glyphs) nearly 5 times better. This is with disregfard for the DD damage loss so it would be slightly closer, but im pretty sure glyph of incinerate would beat immolate.

[edit]

If it turns out that Decimation rotations involving 1xFiller-1xSF are viable in a raid situation, there's still room for significant improvement in the demo specs.

I'm mostly worried about fights like Grobbulus, they'd make weaving SB->SF quite iffy.

• Zakalwe · 02/27/09, 6:16 AM

I went ahead and generated some scale factors, they're included in the original post now. Nothing too surprising, though they're a bit more even between the various specs than I'd have guessed.

Warlocomotif · 02/27/09, 6:22 AM

Originally Posted by Zakalwe

I went ahead and generated some scale factors, they're included in the original post now. Nothing too surprising, though they're a bit more even between the various specs than I'd have guessed.

Scale factors for meta/ruin vs FG/ES are a bit dissapointing, but I suppose meta/ruin will scale well in terms of raid dps to make up for that.

Yuckie · 02/27/09, 6:35 AM

Originally Posted by Zakalwe

* Eradication is assumed to have an internal 30 second cooldown. (Unverified)

I've read through most of this thread, and did not see it mentioned. If it was brought up already and just hasn't been updated then please forgive the repost.

On the test server (02/27/09 @ 3:24am PST) I cast corruption on 4 lvl 60 dummies (this should have no bearing on proc rates) and kept this cycle running for 5 recasts.

4 Targets, 5 Recasts, 6 Ticks per casts, 120 total ticks.

Over the course of this Eradication proc'd 10 times, each time refreshing the duration on the buff. I never lost Eradication during this test.

So currently Eradication has no internal cooldown (I had the buff refresh from 27 seconds back to full). This should play heavily in Afflictions favor if your limiting it to a 30 second internal.

Liania · 02/27/09, 6:36 AM

Would like to see some test results on this specc: http://talent.mmo-champion.com/?warl...0&version=9626

Assuming you're hit capped without suppression. Spells to use would be immolate/incinerate and conflagrate, maybe corruption aswell for 2setbonus. Not 100% sure which glyphs would be the best, but I assume it would be Conflagrate/Imp/Incinerate.

Thank you.

• Zakalwe · 02/27/09, 6:54 AM

Originally Posted by Yuckie

So currently Eradication has no internal cooldown (I had the buff refresh from 27 seconds back to full). This should play heavily in Afflictions favor if your limiting it to a 30 second internal.

Thanks for testing this. I went ahead and removed the internal cooldown, but when doing that I also discovered a bug - eradication was being triggered at a 10% chance (like on Live) rather than 6%. So I fixed that too, and when looking at the results it turns out that those two changes exactly outweighed each other.

• Zakalwe · 02/27/09, 7:03 AM

I've tested 0/22/49, it doesn't beat 5/13/53.