SimulationCraft for Warlocks (4.0.3a numbers at 80 and 85) - Page 141

Freakz0r · 10/16/09, 5:22 PM

Affliction
* Pandemic: This talent now also increases the critical strike damage bonus of the felhunter’s Shadow Bite spell by 100%.

Destruction
* Ruin: This talent now also increases the critical strike damage bonus of the imp’s Firebolt spell by 100%.

Deuel · 10/16/09, 5:22 PM

More changes for destro that will for sure affect sim:

Destruction

* Ruin: This talent now also increases the critical strike damage bonus of the imp’s Firebolt spell by 100%.

Interested to see the numbers now.

Madlax · 10/16/09, 5:35 PM

Quick and dirty notepad maths(you may always correct it with more exact simcraft values) would say:
854 -> 965 DPS from the imp for 3/13/55

Warlocomotif · 10/16/09, 6:06 PM

Originally Posted by Madlax

Quick and dirty notepad maths(you may always correct it with more exact simcraft values) would say:
854 -> 965 DPS from the imp for 3/13/55

I'm getting slightly different numbers.

imp (DPS=867.7) fire_bolt | Count=161.3| 1.9sec  DPE=  1609| 9%  DPET=   868  DPR=   8.9  pDPS= 866  Miss=0.0%  Hit=1390|1179|1489  Crit= 2086| 1768| 2234|31.4%

New avg crit = Old avg crit / 1.5 * 2
New avg crit = 2086 / 1.5 * 2 = 2781
New DPE = ((1-0.314)*1390)+(0.314*2781)=1826
New DPS = Old DPS * (New DPE / Old DPE)
New DPS = 867.7 * (1826/1609) = 984.7

How are you getting your numbers?

[edit] NVM I used the 0/13/58 numbers. Checking if its very different as 3/13/55

imp (DPS=854.3) fire_bolt | Count=161.3| 1.9sec  DPE=  1584| 9%  DPET=   854  DPR=   8.8  pDPS= 853  Miss=2.3%  Hit=1401|1187|1500  Crit= 2102| 1780| 2250|30.7%

New avg crit = Old avg crit / 1.5 * 2
New avg crit = 2102 / 1.5 * 2 = 2803
New DPE = ((1-0.314-0.023)*1401)+(0.314*2803)=1809
New DPS = Old DPS * (New DPE / Old DPE)
New DPS = 854.3 * (1809/1584) = 976

Still getting different numbers

Cipherr · 10/16/09, 7:28 PM

Originally Posted by Warlocomotif

I'm getting slightly different numbers.

imp (DPS=867.7) fire_bolt | Count=161.3| 1.9sec  DPE=  1609| 9%  DPET=   868  DPR=   8.9  pDPS= 866  Miss=0.0%  Hit=1390|1179|1489  Crit= 2086| 1768| 2234|31.4%

New avg crit = Old avg crit / 1.5 * 2
New avg crit = 2086 / 1.5 * 2 = 2781
New DPE = ((1-0.314)*1390)+(0.314*2781)=1826
New DPS = Old DPS * (New DPE / Old DPE)
New DPS = 867.7 * (1826/1609) = 984.7

How are you getting your numbers?

[edit] NVM I used the 0/13/58 numbers. Checking if its very different as 3/13/55

imp (DPS=854.3) fire_bolt | Count=161.3| 1.9sec  DPE=  1584| 9%  DPET=   854  DPR=   8.8  pDPS= 853  Miss=2.3%  Hit=1401|1187|1500  Crit= 2102| 1780| 2250|30.7%

New avg crit = Old avg crit / 1.5 * 2
New avg crit = 2102 / 1.5 * 2 = 2803
New DPE = ((1-0.314-0.023)*1401)+(0.314*2803)=1809
New DPS = Old DPS * (New DPE / Old DPE)
New DPS = 854.3 * (1809/1584) = 976

Still getting different numbers

Thanks for the data. Seems like a great change, that and possibly having the conflag dot scale on crit could provide decent increases in dps for the tree.

Warlocomotif · 10/16/09, 9:48 PM

For those wondering: I always thought Shadow Bite already critted for 200% (since its a melee attack). Apparently the authors of Simulationcraft thought the same thing, as the attack is already doing 200% damage in the data from the first page. As such, don't expect to see the DPS from the affliction simulation to go up past what it is now.

On an unrelated note, it'd probably be a good idea to change the doomguard spec to the fel hunter using spec. I know that for simulation purposes it makes no difference at all- but a lot of less clever people use the first page of this thread as their "what should I spec?". They'll end up speccing for the improved succubus because the front page's Doomguard profile uses the succubus spec- despite the fact that they should spec for the improved fel hunter since the doomguard doesn't have 100% uptime.

Duckey · 10/17/09, 12:13 AM

Originally Posted by Warlocomotif

For those wondering: I always thought Shadow Bite already critted for 200% (since its a melee attack). Apparently the authors of Simulationcraft thought the same thing, as the attack is already doing 200% damage in the data from the first page. As such, don't expect to see the DPS from the affliction simulation to go up past what it is now.

This thought piqued my interest, so I logged on to live to check. I have been sitting at a dummy for the last 20 minutes with only my felhunter attacking, and it appears that Shadow Bite IS already critting for 200%. Here are the numbers I have partway through:

Type / Min / Avg / Max
Hit / 209 / 272 /323
Crit / 460 / 543 /640

It seems as if afflictions latest "buff" is something that is already on live.

Edit: Should have taken the time to check out the PTR's before posting this.

I am currently on my Warlock on the PTR, and these are the results I have obtained from 15 minutes of testing:

Type / Min / Avg / Max
Hit / 237 / 325 /381
Crit / 601 / 809 /943

So it's now critting around 2.5. Definitely interesting.

Warlocomotif · 10/17/09, 12:17 PM

Originally Posted by Duckey

This thought piqued my interest, so I logged on to live to check. I have been sitting at a dummy for the last 20 minutes with only my felhunter attacking, and it appears that Shadow Bite IS already critting for 200%. Here are the numbers I have partway through:

Type / Min / Avg / Max
Hit / 209 / 272 /323
Crit / 460 / 543 /640

It seems as if afflictions latest "buff" is something that is already on live.

Edit: Should have taken the time to check out the PTR's before posting this.

I am currently on my Warlock on the PTR, and these are the results I have obtained from 15 minutes of testing:

Type / Min / Avg / Max
Hit / 237 / 325 /381
Crit / 601 / 809 /943

So it's now critting around 2.5. Definitely interesting.

Napkin math:

New avg crit = Old avg crit / 2 * 2.5
New avg crit = 1912 / 2 * 2.5 = 2390
New DPE = ((1-0.212-0.054-0.027)*956)+(0.212*2390)=1182.57
New pDPS = Old pDPS * (New DPE / Old DPE)
New pDPS = 505 * (1182.57/1081) = 552.45

So a 47.45 DPS increase for affliction.

Lorre · 10/19/09, 1:45 PM

I just wanted to note that though simulationcraft indicates that the BiS trinket is unaffected by Felguard MD and Meta, it actually is affected.

I recorded a 3122 hit of the 258 version of the trinket with metamorphosis up and felguard active and CoE on the mob.

The damage range of the trinket 1959 to 2275 which modified by CoE is 2213 to 2570.

Adding metamorphosis 20% multiplicitively makes the range 2655 to 3084, which still doesn't account for that hit, but adding metamorphosis and FG additively and then multiplying by CoE does with the final damage range of 2766 to 3212.

This means on average with 36% uptime on metamorphosis the trinket proc is undervalued by 5%+(20%*.36) = 12% which raises the partial dps of the proc from 193 to 216 or 23 dps against my simulation gear set.

dedmonwakeen · 10/19/09, 8:13 PM

Originally Posted by Lorre

I just wanted to note that though simulationcraft indicates that the BiS trinket is unaffected by Felguard MD and Meta, it actually is affected.

I recorded a 3122 hit of the 258 version of the trinket with metamorphosis up and felguard active and CoE on the mob.

The damage range of the trinket 1959 to 2275 which modified by CoE is 2213 to 2570.

Adding metamorphosis 20% multiplicitively makes the range 2655 to 3084, which still doesn't account for that hit, but adding metamorphosis and FG additively and then multiplying by CoE does with the final damage range of 2766 to 3212.

This means on average with 36% uptime on metamorphosis the trinket proc is undervalued by 5%+(20%*.36) = 12% which raises the partial dps of the proc from 193 to 216 or 23 dps against my simulation gear set.

Interesting. It seems there is a class-agnostic "damage multiplier". I'll have to change the architecture to support that.

dedmonwakeen · 10/20/09, 8:38 AM

We recently put out a simple GUI wrapper around the SimulationCraft engine. Since Warlocks are probably the heaviest users, I'm requesting that you guys poke it and prod it a bit. It is a very light-weight wrapper. My intent is still to keep as much of the core functionality in the wrapper itself, which does limit the functionality somewhat.

Please don't clutter the thread with defects, open an Issue on the site instead. Thanks!

Warlocomotif · 10/20/09, 9:24 AM

As soon as I saw that a GUI was added I downloaded it ASAP (Because I think a GUI for Simcraft would be really cool). I went to the folder I was downloading to and right-click->extrac here. Got an error- turns out it was still downloading.

I really like having a GUI for Simulationcraft (and I will edit feedback about the GUI in here), but I think that 22 megabytes extra for just a gui is perhaps a bit much.

[edit]
The first thing I'd like to see is a choice between Demonic Pact and totem of Wrath when "spellpower"is ticked in the options tab.
Second, It would be nice of you could import more than 1 armory at a time. For example having subtabs for every imported profile under the "Simulate" tab.

dedmonwakeen · 10/20/09, 12:54 PM

Originally Posted by Warlocomotif

I really like having a GUI for Simulationcraft (and I will edit feedback about the GUI in here), but I think that 22 megabytes extra for just a gui is perhaps a bit much.

The first thing I'd like to see is a choice between Demonic Pact and totem of Wrath when "spellpower"is ticked in the options tab.
Second, It would be nice of you could import more than 1 armory at a time. For example having subtabs for every imported profile under the "Simulate" tab.

The increase in the download is due to the Qt DLLs, specifically QtWebKit. 22MB for internet browser isn't too bad.

Of course, if I sat down and implemented a real "paper-doll" interface this DLL would no longer be required. However..... I've tried very very hard to limit the scope of the GUI and keep the functionality in the engine in case other people wish to wrap it with a different interface.

Regarding DP:
If you are simulating a Demo Warlock with DP, then it will auto-override if it produces a bigger bonus. If you are simulating another class/spec and want it to benefit from DP you have two choices: (1) Create a Demo Warlock profile warlock_dp.simc and add it to Overrides or Simulate, or (2) add default_enchant_spell_power=some_value to Override to just use a static bonus.

I'm warming up to the idea of having sub-tabs under Simulate. Have you tried using the "<" and ">" buttons? They will cycle through all the profiles you have imported which is almost the same functionality.

Shayo · 10/20/09, 4:49 PM

Is there currently any way to use simcraft to simulate specific timing of an on use trinket? I'm asking specifically for cases where you want to test dps gains of a trinket with criteria like "On conflag cast, use talisman of volatile power?"

Also, my testing of the trinket doesn't seem to return the right results, so I'm not sure I'm doing it right.

First, I removed trinket1 entirely from the warlock_t8_00_13_58.simcraft profile. (simcraft-r3080-windows)

   7683 100.0%  Raid
   7688  100.0%  Warlock_T8_00_13_58

I then replaced trinket1 in the warlock_t8_00_13_58.simcraft file with this line:
trinket1=talisman_of_volatile_power_heroic,stats=126crit,equip=onspellhit_64haste_8st ack_20dur_120cd
to accurately model the trinket, but this resulted in a 90 dps change where I would have expected ~113 based on scale factors for just the crit.

   7775 100.0%  Raid
   7778  100.0%  Warlock_T8_00_13_58

I checked http://simulationcraft.googlecode.co.../Examples.simc, but I didn't see any specific examples for how to fine tune usage of on_use trinkets.

Warlocomotif · 10/20/09, 6:18 PM

Did you add a /use line to the action list? The value shouldn't go below the starts= value, but to get the use I think you will have to add it to the actionlist.