Elitist Jerks
Register
Blogs
Forums


Go Back   Elitist Jerks » Class Mechanics » Warlocks

Closed Thread
 
LinkBack Thread Tools
Old 11/25/10, 8:30 AM   #3016
VoidStar
Piston Honda
 
Human Warlock
 
Draenor (EU)
This discussion was using the level 85 profile and numbers. Fishing for the T10 Proc, especially when we didn't use the Meta glyph, was a very good strategy, but there isn't a comparably powerful and dependable buff worth waiting on at 85 (so far). Clearly, when gear and especially procs change, the ideal priority list list is also likely to change.

Offline
Old 11/25/10, 2:05 PM   #3017
VoidStar
Piston Honda
 
Human Warlock
 
Draenor (EU)
Investigation of Hand of Gul'dan Refreshing Immolate

I've been tinkering with the demonology priority list and talent build in Simcraft as I can see sevaral areas for improvement. One thing I'm investigating is using Fel Flame in cases where Hand of Gul'dan would fail to refresh Immolate, but something odd is going on.

I reduced the action list to the following:

actions+=/snapshot_stats
actions+=/fel_flame,if=dot.immolate.remains>travel_time&dot.immolate.remains<5&!in_flight
actions+=/immolate,if=!ticking

Which produces a run where the character casts immolate once and Fel Flame once every 4.66 seconds for the rest of the fight. This seems like correct behaviour.

If I add HoGd to the end of the list to create this action list:

actions+=/snapshot_stats
actions+=/fel_flame,if=dot.immolate.remains>travel_time&dot.immolate.remains<5&!in_flight
actions+=/immolate,if=!ticking
actions+=/hand_of_guldan

Then I would still expect only one Immolate as I've simply added another mechanism for refreshing it. Instead the sim claims to cast Immolate a little over 7 times. I can't understand why.

I'm not sure if this is a bug with Simcraft, a misuse of the action list, a misreading of the output on my part, or a misunderstanding of the mechanics of the spells involved. Any help would be most appreciated!

EDIT: This is using Simcraft version 403.2 which I downloaded about an hour ago.

Last edited by VoidStar : 11/25/10 at 2:06 PM. Reason: Simcraft version added

Offline
Old 11/25/10, 3:18 PM   #3018
 Zakalwe
The Chairmaker
 
Orc Warlock
 
Balnazzar
Yep, looks like you ran into an interesting bug that was causing the "remains<5" check to fail on any immolate that had been refreshed by HoG. It's fixed in r5723.

Norway Offline
Old 11/25/10, 10:25 PM   #3019
ron9
Von Kaiser
 
Gnome Warlock
 
Arygos
Would it be possible to also model the best DPS rotations for AoE situations? At least show how many enemies it would take to start using abilities like Rain of Fire/Seed of Corruption over the normal DPS rotation.

As it stands now, for destruction, am I correct in saying that you NEVER use Rain of Fire nor Seed of Corruption, but ONLY Bane of Havoc in AoE DPS situations?

Offline
Old 11/26/10, 4:11 AM   #3020
Caltiom
Von Kaiser
 
Human Priest
 
Eredar (EU)
Originally Posted by ron9 View Post
Would it be possible to also model the best DPS rotations for AoE situations? At least show how many enemies it would take to start using abilities like Rain of Fire/Seed of Corruption over the normal DPS rotation.

As it stands now, for destruction, am I correct in saying that you NEVER use Rain of Fire nor Seed of Corruption, but ONLY Bane of Havoc in AoE DPS situations?
I added Rain of Fire, so all 3 AE Spells should be usable if a new version is released.

actions+=/seed_of_corruption
actions+=/hellfire
actions+=/rain_of_fire

add the option target_adds=n and max_time should be set to ~100s.

Keep in mind that the whole Adds/AE thing in simcraft is rudimentary at best. Hopefully next year when everything settles down we can finally finish real multi target support in simcraft (for multidotting, etc.). Multi target damage is theorethically already possible, but the whole dot/debuff system is still a big mess that needs to be redone to fully support it.

Offline
Old 11/26/10, 7:24 AM   #3021
VoidStar
Piston Honda
 
Human Warlock
 
Draenor (EU)
In my opinion, the biggest problem with multi-target situations is not that the model only partially supports them, but the difficulty of deciding on something useful to model.

Other than Wrath-era trash, when you're in a situation where there are multiple targets there's usually some reason why some of them must be damaged more quickly than others and without being very fight-specific, Simcraft isn't going to be able to help with that. You also have the problem that adds may be very short-lived (which drastically changes the worth of DoT effects) multiple targets interact strangely with Decimation-like effects and some spells may only be present on one target. It can be difficult to predict what important debuffs are likely to be present on multiple targets. There's also the issue that you can't control what your other party members will do: if they choose to single-target an add down then it'll live only a fraction as long as if they uniformly AoE all targets.

If you're in a situation where you have multiple targets without a priority then it's probably not worth worrying too much and a simple rule of thumb will do.

I guess you could run a simulation where you just spam an AoE and multiply up its DPET for your number of targets (assuming you're under the AoE damage cap) and see where that overtakes the average DPET of whatever spells you feel you can apply to that number of targets, but even that glosses over a lot of the variables.

Offline
Old 11/27/10, 8:19 PM   #3022
Sephirah
Von Kaiser
 
Undead Warlock
 
Draenor (EU)
Does the last version (403-2) miss the option to calculate mastery scaling?
I see "Plot DPS per mastery rating" under Plots, but no "Analyze mastery" under Scaling, so the simulation results give a value of 0.0000.

Offline
Old 11/27/10, 10:59 PM   #3023
whi
Von Kaiser
 
Human Warlock
 
Bronzebeard (EU)
Originally Posted by Sephirah View Post
Does the last version (403-2) miss the option to calculate mastery scaling?
I see "Plot DPS per mastery rating" under Plots, but no "Analyze mastery" under Scaling, so the simulation results give a value of 0.0000.
It does, and as far as I know it's been fixed already, so the option should be there in the next release.

As a workaround you can add this to your overrides tab:
Click Here ← Click Here
calculate_scale_factors=1
scale_only=haste_rating,mastery_rating,crit_rating,intellect,spell_power
scale_intellect=150
scale_spell_power=150
scale_crit_rating=150
scale_haste_rating=150
scale_mastery_rating=150

Poland Offline
Old 11/29/10, 3:02 AM   #3024
 Zakalwe
The Chairmaker
 
Orc Warlock
 
Balnazzar
OP has been updated with the latest results. The main change is that we by default no longer add any special lag penalty after GCD-bound spells, since you can queue up your next spell during the GCD now. Destruction in particular gained a lot from this.

Norway Offline
Old 11/30/10, 10:52 AM   #3025
Rushi
Glass Joe
 
Tauren Death Knight
 
Stormscale
Scaling

Currently the sim has both Affliction specs providing nearly the same DPS and I would like to determine which spec scales best targeting haste as the primary secondary stat and of course int and spell power before secondary stats.
Both Affliction specs favor Haste as the best secondary stat so assuming you target Haste in your gear, doing some simple math with the normalized scale factors shows that Affliction should scale better. If you target Mastery as the second best secondary stat Affliction should always be better than Aff_drain. Except that as I understand it haste has a cap where mastery does not currently.

(Haste_Aff) - (Haste_AffDrain) compared to (Mastery_Aff) - (Mastery_Affdrain)

Affliction gets +0.0736 more DPS per point of Haste
Aff_Drain gets + 0.0662 more DPS per point of Mastery

My understanding of Haste is a bit cloudy but at some point, perhaps not reachable at T11 but with raid buffs and procs I believe it is, shadowbolt will reach a maximum casting speed @ 50% haste from my understanding. Will this cause the Scale factor for haste in the Affliction profile to fall below the scale factor for haste in the Aff_Drain profile? With Eradication and Imp Soul Fire does that mean that @ 15% haste + 20% eradication + 15% imp soul fire shadowbolt is being casted as fast as possible and therefor the Scale Factor for haste is lower at that point? If this is the case how much lower is the scale factor and how does that affect the scaling of Affliction v Aff_drain? Another consideration is at that point I am assuming Glyph of UA is going to be useless and we would want to switch to Glyph of Corruption?

Finally as an aside with Aff_drain when nightfall procs do you immediately cast the instant shadowbolt in the middle of a Drain Life or wait for the Drain Life to complete and then cast the instant shadowbolt?

Offline
Old 11/30/10, 12:08 PM   #3026
~Thalia~
Von Kaiser
 
Gnome Warlock
 
Mar�carge de Zangar (EU)
Gcd caps at 50% haste, being reduced to 1 sec.

But for a 2.5 sec cast to reach the gcd (theorical haste hard cap for SB), you'll need 150% haste which is of course impossible to reach. It would be 4918.5 haste rating at level 80.

Edited for clarity regarding wasniahC 's post below

Last edited by ~Thalia~ : 12/02/10 at 10:15 AM.

Offline
Old 12/01/10, 6:05 PM   #3027
wasniahC
Von Kaiser
 
Undead Warlock
 
Twisting Nether (EU)
Originally Posted by ~Thalia~ View Post
Gcd caps at 50% haste, being reduced to 1 sec.

But for a 2.5 sec cast to reach the gcd (theorical haste hard cap), you'll need 150% haste which is of course impossible to reach. It would be 4918.5 haste rating at level 80.
Even then, it will continue to boost the DPE of dots, so haste will continue to provide dps - Likely not the highest DPS, at that point, but a hard cap doesn't really exist as such any more, with the change for haste with dots

Offline
Old 12/01/10, 7:10 PM   #3028
Enigmafury
Glass Joe
 
Human Warlock
 
Разувий (EU)
I noticed a comment on wowhead about Dark Intent
Dark Intent - Spell - World of Warcraft
And the 3 ones below it...so if we aim for max personal dps(not rdps) it would seem that stacking DI is a good idea and it will boost our dps.

Also hear a rumor that if you had like 5 locks in a raid and all of them drop a DI on 1 lock, that lock would gain 15%(or 15,92%) haste, is it true?

Offline
Old 12/01/10, 10:20 PM   #3029
tresextas
Glass Joe
 
Undead Warlock
 
Baelgun
Originally Posted by Zakalwe View Post
It's modeled as if the buff were constantly up and fully stacked to 3. I could also have modeled it dynamically by having the warlock actors target each other with the spell, but I wanted to avoid the effects that would have on DPS, particularly during scale factor generation.
Not to nitpick, but isn't this an unrealistic scenario that the buff will be up 100% of the time with all 3 stacks? This is concerning dark intent, the above comment helped me remember this.

Last edited by tresextas : 12/01/10 at 10:36 PM.

Offline
Old 12/02/10, 1:06 AM   #3030
Saufsoldat
Von Kaiser
 
Undead Warlock
 
Madmortem (EU)
Originally Posted by tresextas View Post
Not to nitpick, but isn't this an unrealistic scenario that the buff will be up 100% of the time with all 3 stacks? This is concerning dark intent, the above comment helped me remember this.
It depends on your raid composition, but if you have at least one warlock, resto druid, fire mage, boomkin, unholy dk, or spriest you will have close to 100% uptime. Other classes/specs can probably also provide you with enough dots/hots to grant a decent uptime, those are just the first ones that come to mind.

Since it's been a while that the beta was out we cannot say with 100% certainty if two warlocks giving each other dark intent will stack when cata goes live.

Offline
Closed Thread

Go Back   Elitist Jerks » Class Mechanics » Warlocks

Thread Tools