[5.4] Brusalk's Destruction Warlock guide

[Deleted]

While we're at it, did u notice something on rof ember generation ? Kinda feels like it has increased since the patch

This might be true I've some incredible generation sometimes when I'm questing, might be worth testing on training dummies.

[Cocobutters]

I used fel-imp on a few fights where there was target switching (Dark Animus) or where his 3-target dispel was useful (Lei-shen). I'm of the opinion that you should use whatever pet you want to use for the fight if there's no switching, and use imp/fel imp if there's a large amount of target switching within the imps casting range.

The pets are all so close that it's a sub 1% DPS difference to use one pet over the other. I'd much rather be able to help interrupt or dispel people than potentially gain a few hundred thousand damage over the course of a 5 minute fight.


As for Nezyt and Cocoabutters I have 2 things to say:

1) They tend to pad meters unnecessarily while I don't (Prime example being... *cough cough* Megaera).

2) They shouldn't have used Sup/Serv over GoSac. They actually gimped the raid even more by using GoSac over GoSup/Serv because the spread out their talent damage increase from the primary target to the secondary target. (As GoSac is balanced to be competitive with the pet talents on 2 targets, not 1. 1 target GoSac is terrible)

Even with padding I did pretty bad the attempt we killed it )=

[Apero]

I tryed RoF on a lvl 90 dummy. 20 times. Full RoF , Singletarget.
I mostly got 2-4 Emberbits. One time 0, one time 6. Never 3. No matter the amount of crits.
Question is, how you get a decent rain of fire uptime without being oom the whole time? (singletarget)

I have a question about fnb, what about the tooltip? My Mastery increases the power of ember consuming spells by 65% , but fnb increases the dmg by only 57%. Whats the base value of fnb? I always thought 43% , why the low fire and brimstone increase then? I dont really get the math behind it. (could be the hangover ) Can someone explain?

[Deleted]

I've tested RoF too. Got around 0.4 emberbit per tick (200 Emberbits sample) Was 0.20-0.25 last time I tested. Now this could have been luck (nope) etc, BUT this time I always got 2 emberbits at the same time, it was definitely changed in 5.2.

Each tick has like 20% chance to generate 2 emberbits at once.

Edit: browsing Simulationcraft's code, it looks like RoF's ember gain is badly implemented :

virtual void impact( action_state_t* s )
{
warlock_spell_t::impact( s );

if ( result_is_hit( s -> result ) && td( s -> target ) -> dots_immolate -> ticking )
trigger_ember_gain( p(), 0.1, p() -> gains.rain_of_fire, parent_data.effectN( 2 ).percent() );
}

It doesn't need Immolate for some time now, and the gain doesn't seem to be 0.1 anymore. I don't know where the chance (parent_data.effectN( 2 ).percent()) is coming from.

[Harekrsna]

Brusalk, what if any fights did you go Sac on? I went Sac this week on Ji-kun because I figured my pet would despawn when I jumped to platforms (was on egg duty), is this the case or will my pet hangout and keep attacking the boss?

Any other fights you know of where pets don't operate properly and we should go for Sac?

[Googolplex]

Because of pathing issues, 'Sup/Fel Imp' will sometimes be better than 'Sup/Observer' - you might wanna mention that. Excellent guide!

[Brusalk]

I tryed RoF on a lvl 90 dummy. 20 times. Full RoF , Singletarget.
I mostly got 2-4 Emberbits. One time 0, one time 6. Never 3. No matter the amount of crits.
Question is, how you get a decent rain of fire uptime without being oom the whole time? (singletarget)

I have a question about fnb, what about the tooltip? My Mastery increases the power of ember consuming spells by 65% , but fnb increases the dmg by only 57%. Whats the base value of fnb? I always thought 43% , why the low fire and brimstone increase then? I dont really get the math behind it. (could be the hangover ) Can someone explain?

You don't, and you just sit there waiting (or moving) for mana :P

---------- Post added 2013-03-10 at 02:35 PM ----------

I've tested RoF too. Got around 0.4 emberbit per tick (200 Emberbits sample) Was 0.20-0.25 last time I tested. Now this could have been luck (nope) etc, BUT this time I always got 2 emberbits at the same time, it was definitely changed in 5.2.

Each tick has like 20% chance to generate 2 emberbits at once.

Edit: browsing Simulationcraft's code, it looks like RoF's ember gain is badly implemented :

It doesn't need Immolate for some time now, and the gain doesn't seem to be 0.1 anymore. I don't know where the chance (parent_data.effectN( 2 ).percent()) is coming from.

Holy tits man. Simcraft is completely out of date on rain of fire. It's worse than that.

They made it so it can't crit :O

rain_of_fire_t( warlock_t* p ) :
    warlock_spell_t( "rain_of_fire", p, ( p -> specialization() == WARLOCK_DESTRUCTION ) ? p -> find_spell( 104232 ) : ( p -> specialization() == WARLOCK_AFFLICTION ) ? p -> find_spell( 5740 ) : spell_data_t::not_found() )
  {
    dot_behavior = DOT_CLIP;
    may_miss = false;
    may_crit = false;
    channeled = ( p -> spec.aftermath -> ok() ) ? false : true;
    tick_zero = ( p -> spec.aftermath -> ok() ) ? false : true;

    tick_action = new rain_of_fire_tick_t( p, data() );
  }

As for parent_data, that is the reference to the rain_of_fire_t struct which contains the spell data for rain of fire. rain_of_fire_tick_t is the struct that contains the tick and impact event. rain_of_fire_tick_t is the "child" of rain_of_fire_t, and parent data references the "actual" rain of fire.

pd.effectN( 2 ).trigger() I'm guessing is the tigger chance to gain an ember.

---------- Post added 2013-03-10 at 02:39 PM ----------

Brusalk, what if any fights did you go Sac on? I went Sac this week on Ji-kun because I figured my pet would despawn when I jumped to platforms (was on egg duty), is this the case or will my pet hangout and keep attacking the boss?

Any other fights you know of where pets don't operate properly and we should go for Sac?

On ji-kun I was Demo and was on the boss full time.

I'm planning on going Sac for horridon, primordious, and another boss I can't remember the name of, if the heroic mechanics aren't that different from normals.

Basically if I would spend a large amount of the time casting ember consumers over generators, I would use Sac over a pet option.

---------- Post added 2013-03-10 at 02:57 PM ----------

I've tested RoF too. Got around 0.4 emberbit per tick (200 Emberbits sample) Was 0.20-0.25 last time I tested. Now this could have been luck (nope) etc, BUT this time I always got 2 emberbits at the same time, it was definitely changed in 5.2.

Each tick has like 20% chance to generate 2 emberbits at once.

Edit: browsing Simulationcraft's code, it looks like RoF's ember gain is badly implemented :

It doesn't need Immolate for some time now, and the gain doesn't seem to be 0.1 anymore. I don't know where the chance (parent_data.effectN( 2 ).percent()) is coming from.

It looks like simc still has the proc chance on embers such that it gives RoF a 25% chance to generate embers (had 118 resource gains out of 473 total ticks), which each tick only having a chance to generate 1 emberbit.

Conflag/Incinerate is generating around .13 ember per cast average which is around expected for a 35% crit rate (as reported by simc)

Immolate's ember gain proc chance is around the crit % reported as expected.

One thing I'm not sure on, is that the incinerate t15 4pc bonus is reporting an average gain of 0.13 per cast. I'm not sure that simc is modeling the 4pc accurately.

[Deleted]

As for parent_data, that is the reference to the rain_of_fire_t struct which contains the spell data for rain of fire. rain_of_fire_tick_t is the struct that contains the tick and impact event. rain_of_fire_tick_t is the "child" of rain_of_fire_t, and parent data references the "actual" rain of fire.

pd.effectN( 2 ).trigger() I'm guessing is the tigger chance to gain an ember.

I figured that, but I don't know what field to look at in sc_spell_data.inc :s (not too fond of C++ )

On another note Destro is amazing on Lei Shen, looking forward to Heroic progress on him (and screw the fury warriors -_-)

[Brusalk]

I figured that, but I don't know what field to look at in sc_spell_data.inc :s (not too fond of C++ )

On another note Destro is amazing on Lei Shen, looking forward to Heroic progress on him (and screw the fury warriors -_-)

As far as I know the spell data stuff is black magic.

[Gobuchul]

If RoF ember gain mechanics changed in 5.2, it is (as far as I know) an undocumented change. I'd be happy to update the sim accordingly, but I need to know what exactly the new mechanic is. Feel free to PM me the results of any testing.

And RoF is obviously still critting in the sim - may_crit doesn't mean what you think it means.

[Brusalk]

If RoF ember gain mechanics changed in 5.2, it is (as far as I know) an undocumented change. I'd be happy to update the sim accordingly, but I need to know what exactly the new mechanic is. Feel free to PM me the results of any testing.

And RoF is obviously still critting in the sim - may_crit doesn't mean what you think it means.

Yeah... After I posted that I realized that I was looking at the direct damage and not the "dot" tick damage for the crit %.

[Sinj]

One final comment and I'll shut up for now

For fun I decided to see just how much of a difference a full-mastery reforge would have on aoe/single-target for my own character.

So I simmed it first at my 1:1:1 (4800 mastery:4800 haste:4800 crit) and on single-target I got 123,250 DPS

Then I simmed my single-target with full-out mastery (7518 mastery:4472 haste:4182 crit) and I got 121,851 DPS
(EDIT: I could probably tune it even more so i got even more mastery, but this should suffice as the results will show - the point I'm about to make does not depend on whether or not I reforged perfectly; only that there's a noticeable difference.)

_______________

Then I simmed it for aoe (8 targets):

1:1:1 (4800 mastery:4800 haste:4800 crit) = 497,341 DPS

full-out mastery (7518 mastery:4472 haste:4182 crit) = 517,997 DPS

Conclusion:
I lost 1,500 DPS on singletarget DPS from going from a balanced reforge to a full-out mastery. However, I gained 20k DPS in my AoE sims from doing so. And this is not taking into account that mastery might be undervalued even in the AoE-sim, since the sim can't emulate adds entering the fight and dying from shadowburn+havoc etc.

What I take from this is that, considering the encounters we'll meet in 5.2 where adds/multitargets are plentiful, maybe we should focus more on mastery, even though the single-target sims tell us not to? Maybe to play it safe go for a 3:3:5 distribution (haste:crit:mastery respectively).

Any thoughts?


EDIT2:

Oh and one final thing: Mastery was way better than even Intellect in the AoE sim (Int: 19.08, Mastery: 16.78). Doesn't really relate to the post above, but figured I'd mention it

[Brusalk]

This might be true I've some incredible generation sometimes when I'm questing, might be worth testing on training dummies.

We just did Sha of Fear 10H and on our last wave of 5 adds I couldn't spam shadowburn enough with the ember generation on 6 targets. It's never been so strong before...


Something's changed FOR SURE

[Gobuchul]

One thing I'm not sure on, is that the incinerate t15 4pc bonus is reporting an average gain of 0.13 per cast. I'm not sure that simc is modeling the 4pc accurately.

That's not per cast, that's per activation of the set bonus. Looks correct to me - 0.13 is the expected average gain if you have 30% crit chance.

[Deleted]

Looks like this was enough for Destro to catch up with Aff on Simcraft with T14H gear (+4k dps)

[Count Zero]

Funny, I read these posts after goofing off in some t14 cleanup tonight as Destro and having as-good or better results on bosses (I came in after Protectors in Terrace) than I did as Afflic before.

[Deleted]

I've done ToES and HoF as Destro this week, and this was much more fun than Affliction... which was already more fun than Assassination. Therefore: happy me.

[kaamila]

brusalk i noticed youre reforging all mastery into crit and keeping haste unless using it get to hit cap.are you finding this a higher dps gain than haste being the go-to stat? also is there a reason for the int/mastery gem in boots if ur reforging all mastery off anyways? following the way you do it, i see myself having 2100 more crit, 2453 more haste, and 2978 LESS mastery. im thinking i should get rid of that haste for more mastery?

[Deleted]

Single target fights: Crit >= Haste > Mastery[/B] (Haste becomes better than crit at higher ilvls)
Two-target fights: Haste > Crit >(=) Mastery[/B]
AoE fights : Mastery >>> Haste > Crit

Thanks for these results, which confirm what my common-sense was hinting : the more ember-spells you used, the better mastery is, and "the more spells you cast the more ember you get".

If I ever have a lot of spare time in the coming months I will probably make a new app that calculates useful DPS on this encounter like for Protectors

That would be a "good thing" for the heroic progress. Besides, i guess that 95% of the current protector could be reused for Megaera ...

I don't know where the chance (parent_data.effectN( 2 ).percent()) is coming from.

This is spellid 104232 (RoF for destro) and the value is 25. Besides, the gain is 0.2 embers (cf line 3462 in sc_warlock.cpp)

[Deleted]

That would be a "good thing" for the heroic progress. Besides, i guess that 95% of the current protector could be reused for Megaera ...

Not so sure about that, my Protectors app was pretty crappy and didn't work in most cases Better start from stratch.

This is spellid 104232 (RoF for destro) and the value is 25. Besides, the gain is 0.2 embers (cf line 3462 in sc_warlock.cpp)

Yes it was changed last night after we exposed the problem to Zakalwe (https://code.google.com/p/simulation...sc_warlock.cpp)