SimulationCraft for Warlocks (Panda Edition) - Page 32

whi · 12/03/12, 3:55 AM

The latter.

For example:

main_hand=loshan_terror_incarnate,id=86328,gems=500int,enchant=jade_spirit,reforge=haste_mastery

results in 98577 dps.

main_hand=loshan_terror_incarnate,id=86328,gems=500int,enchant=jade_spirit,reforge=haste_mastery,upgrade=1

results in 99582 dps.

main_hand=loshan_terror_incarnate,id=86328,gems=500int,enchant=jade_spirit,reforge=haste_mastery,upgrade=2

results in 100957 dps.

Make sure You're using the newest version of simc.

netsach · 12/04/12, 7:02 AM

As i'm not seeing anything realted to this into the code yet, i'll post the information. I guess this doesn't change anything to current sim results, as the dot doesn't seem to be able to generate more ticks via haste, and the damage from GoSac is the same wether it be direct (as currently implemented) or as a dot. Maybe it would have some side effects in case of raid events (immune, distance, etc) but otherwise, that's pretty much for information.

The additional damage Grimoire of Sacrifice grants to Chaos Bolt is now dealt over 3 seconds. This damage over time behaves in a manner similar to a Mage's Ignite, where each new damage over time effect rolls into the previous one. This means that no damage is lost if a new damage over time component is re-applied before an existing one fades.

mannaroth · 12/04/12, 5:37 PM

Something seems to be wrong with the 4 set bonuses on Simcraft 510-3. I've tested on both myself and on another warlock with the 4 set bonus, and actually get a dps LOSS by using it, of around 2k. I tried manually changing the shoulders in the code to non-set shoulders with identical stats, and the profile gained 2k dps.

Expandus · 12/04/12, 9:18 PM

Ye, seems to be bugged with -4 as well. Without 4set dark soul is used 4.3 times and buff gained 4.3 times. With 4set dark soul is used 6 times, but the buff is only gained 3.3 times.

• Zakalwe · 12/05/12, 2:51 AM

Originally Posted by Expandus

Ye, seems to be bugged with -4 as well. Without 4set dark soul is used 4.3 times and buff gained 4.3 times. With 4set dark soul is used 6 times, but the buff is only gained 3.3 times.

Yeah, somebody broke something just before the latest release. A new release with a fix should be out shortly.

Syrophenikan · 12/05/12, 9:35 AM

Anyone else having issues with simcraft 510-5 constantly crashing when you add upgrade=2 attribute to your gear?

When I import my character and run him as is, works fine, but once I added upgrade=2 to my shoulders and relic trinket, it crashes when it's doing stat scaling. The simulation seems to work fine, but I can't scale stats when my profile has upgrade=2 on some gear.

• Zakalwe · 12/05/12, 9:49 AM

We did several releases over the past 48 hours due to various bugs (certain people need to learn to test their changes before committing them, grumble grumble), so please check if you're still having a problem with the latest release. I believe that's 510-5 by now.

EDIT: I swear your post said 510-2 when I read it. If you're still seeing crashes, please open an issue on our googlecode site.

Syrophenikan · 12/05/12, 11:15 AM

Originally Posted by Zakalwe

We did several releases over the past 48 hours due to various bugs (certain people need to learn to test their changes before committing them, grumble grumble), so please check if you're still having a problem with the latest release. I believe that's 510-5 by now.

EDIT: I swear your post said 510-2 when I read it. If you're still seeing crashes, please open an issue on our googlecode site.

You're not seeing things, it did say 510-2, but that was a typo and I edited my post to reflect the correct version I am using. :p I shall submit an issue, thanks.

randa · 12/26/12, 7:22 AM

I see that feral druids are using these commands in their profiles, dot.rip.multiplier and dot.rake.multiplier. Can someone please try to optimize our profiles with similar commands ("dot.agony.multiplier"). to get it more in line with how people play affliction on live. If results are to high for public page, you could link them to me in pm.

Karsteck · 01/03/13, 9:49 AM

Originally Posted by randa

I see that feral druids are using these commands in their profiles, dot.rip.multiplier and dot.rake.multiplier. Can someone please try to optimize our profiles with similar commands ("dot.agony.multiplier"). to get it more in line with how people play affliction on live. If results are to high for public page, you could link them to me in pm.

Other than mastery procs or on-use effects, no aff CDs or raid buffs will give a flat multiplier to our DoT damage. Our CDs give haste, and our procs and pots give int or spellpower. Ferals on the other hand have tiger's fury (% all physical dmg done), and can pick up Nature's Vigil (% all dmg).

What would be useful for aff would be any kind of test on dot dps (most of the time you want something that accounts for hasted or high-crit dots, not just high spellpower or high mastery, so you need tickspeed and crit% as well as dmg per tick.) Then you could write conditions about when to haunt and when to let dots tick out into their refresh window waiting for procs. Since you don't want to haunt unless you have some kind of proc going on (or about to come off ICD) (also vulnerable phases or burn windows, or coming up on execute range you start dumping shards since they're free after 20% on a pure single target fight, esp. if you have KJs cunning so you don't have to interrupt DS for movement.)

I've been meaning to have a look at doing this, since I don't fully trust my stat weights with an action list that just dumps shards into haunt as soon as it gets them. I don't know when during dark soul the soulburn:soulswap in the action list tends to happen. Not at the beginning, I'd assume, and that's not ideal. I hope there's at least a way to test whether you have more spellpower than your base.

realCool · 01/12/13, 11:15 AM

So does somebody know if there is some kind of command to compare current ticking dot power to dot power if applied now which includes all stats (spell, haste, crit, mastery, other dmg increse buffs)? If comparing is possible I would try to improve the current action list in order to fit more to the real playstyle

merlinjo · 01/15/13, 3:28 AM

Originally Posted by realCool

So does somebody know if there is some kind of command to compare current ticking dot power to dot power if applied now which includes all stats (spell, haste, crit, mastery, other dmg increse buffs)? If comparing is possible I would try to improve the current action list in order to fit more to the real playstyle

yeah I'd like to know if such a thing is possible and is it possible to do some additional calculations inside profile/action list and store them as a variable so that we can use it for comparison?

jason1975 · 02/10/13, 5:52 PM

FYI - I'm getting a DPS increase of about 10k on the 2 target sim for Destro if you throw in this line after shadowburn (it makes sure Havoc is always used by CB . . . not sure why the original line isn't working):

actions+=/chaos_bolt,if=ember_react&(buff.havoc.stack>=3&buff.havoc.remains>cast_time)

Here's the full action list:

actions=curse_of_the_elements,if=debuff.magic_vulnerability.down
actions+=/use_item,name=shaskin_gloves
actions+=/jade_serpent_potion,if=buff.bloodlust.react|target.health.pct<=20
actions+=/blood_fury
actions+=/dark_soul
actions+=/service_pet,if=talent.grimoire_of_service.enabled
actions+=/grimoire_of_sacrifice,if=talent.grimoire_of_sacrifice.enabled
actions+=/summon_pet,if=talent.grimoire_of_sacrifice.enabled&buff.grimoire_of_sacrifice.down
actions+=/run_action_list,name=aoe,if=active_enemies>3
actions+=/summon_doomguard
actions+=/rain_of_fire,if=!ticking&!in_flight&active_enemies>1
actions+=/havoc,target=2,if=active_enemies>1
actions+=/shadowburn,if=ember_react
actions+=/chaos_bolt,if=ember_react&(buff.havoc.stack>=3&buff.havoc.remains>cast_time)
actions+=/immolate,cycle_targets=1,if=ticks_remain<add_ticks%2&target.time_to_die>=5&miss_react
actions+=/chaos_bolt,if=ember_react&(buff.backdraft.stack<3|level<86)&(burning_ember>3.5|buff.dark_soul.remains>cast_time|buff.skull_banner.remains>cast_time)&(mana.pct<=75|(buff.havoc.stack>=3&buff.havoc.remains>cast_time))
actions+=/conflagrate
actions+=/rain_of_fire,if=!ticking&!in_flight&mana.pct>=50
actions+=/incinerate
actions+=/chaos_bolt,if=mana.pct<20

If I could figure out the line to pool embers during execute so you always did 3 SB's after Havoc (unless target was about to die), I'd do that too. It's a bit beyond me.

Scathbais · 02/12/13, 4:44 PM

Originally Posted by jason1975

]

If I could figure out the line to pool embers during execute so you always did 3 SB's after Havoc (unless target was about to die), I'd do that too. It's a bit beyond me.

[EDIT: sorry I answered the wrong question. what I outlined below is a way to sit on embers if you are about to enter exection phase]

There are two ways to estimate this:

1) use a target.health.pct > X
2) target.time_to_die

I suggest using target health > X where X is larger than 20%. Perhaps 20% plus the the time (in health% if that makes any sense to you) it takes to generate 3 or 4 embers. So, for example, if you could generate 4 embers in the time it takes to reduce target health by 10%, you could put a conditional in that won't cast a chaos bolt if target.health.pct <=30. I would also add logic into this so that you don't unintentionally sit on 4 embers

realCool · 02/15/13, 5:49 AM

As almost all of the new trinkets use the RPPM system I wanted to understand how exactly that system works, as I'm not sure of the mechanics. According to Ghostcrawlser post the formula for the procpercentage using RPPM is the following:

2(ppm) * 1.22(haste) * 1.4(time since last chance to proc) / 60 (sec per min) = 5.693%

So if we take Breath of the Hydra - Item - World of Warcraft as an example. GC posted that the RPPM value for this trinket is 0.5. Also if you start a fight than the time since last proc is capped at 10 seconds. The trinket proccs from dotticks. So asumming we have 20% haste the first dottick should then have a chance to proc the trinket of:

0.5(ppm) * 1.2(haste) * 10(time since last chance to proc) / 60 (sec per min) = 10%

Now if we assume that we get one dottick every second the procchance for each dottick after the first one comes down to:

0.5(ppm) * 1.2(haste) * 1(time since last chance to proc) / 60 (sec per min) = 1%

Is this calculation correct?

That would also mean if we have a dottick every 2 seconds with 0 haste we have a dottick every 1.6 seconds with 25% haste. So then the procchance with 0 haste is:

0.5(ppm) * 1.0(haste) * 2(time since last chance to proc) / 60 (sec per min) = 1.6%

and with 25% haste:

0.5(ppm) * 1.25(haste) * 1.6(time since last chance to proc) / 60 (sec per min) = 1.6%

So in that case the procrate does not scale with haste at all for this kind of trinket in contrast to what GC claimed:

RealPPM procs still scale more frequently with haste

Is there some kind of error in my calculations?