Simulationcraft - Feral rotation discussion and script

[aggixx]

Why DPS disproportion on simcraft.org between 5.05 and 5.1? I can't find any difference between them, but the 5.1 is lower.

No spec related reason. You might see casters having gone up because they have 2 Sha Touched weapons and a Elite Regail's dagger where we just have 1 drop, but that's just speculation.

[Sorcerer]

But last time I checked the DPS difference in simcraft main page between 5.05 and 5.1 was ~350-400dps. So something has changed within the feral script, couse I don't reckon any changes to mechanics or script recently, unless mistaken.

5.05 - 119961 dps
5.10 - 115313 dps (there seems to be drop by ~80-150dps within last 2-3 days to some changes, last time I saw was 116xxx k)

[aggixx]

It's a change with the action list not with the sim mechanics. The advanced script we've been simming with for a while only gained DPS from the Dancing Steel change and that's it.

[Jeshu]

In the most current SimC profile, there are the following lines for the "complex" action list:

Code: Select all

/pool_resource,wait=0.1,if=talent.dream_of_cenarius.enabled&combo_points>=5&((energy<50&buff.berserk.down)|(energy<25&buff.berserk.remains>1))&buff.savage_roar.remains-6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=4.5
/pool_resource,wait=0.1,if=talent.dream_of_cenarius.enabled&combo_points>=5&((energy<50&buff.berserk.down)|(energy<25&buff.berserk.remains>1))&buff.savage_roar.remains+6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=6.5
/ferocious_bite,if=talent.dream_of_cenarius.enabled&combo_points>=5&buff.savage_roar.remains-6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=4
/ferocious_bite,if=talent.dream_of_cenarius.enabled&combo_points>=5&buff.savage_roar.remains+6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=6

What's the purpose of the 0.5 second difference in the SR/Rip comparison between the "pool_resource" line and the corresponding "ferocious_bite" line?

[raffy]

The talent spec in the first post script is incorrect as "010102" (which is NV). Does that explain the DPS difference? I'm simulating that script much higher than 115k.

[Leafkiller]

The talent spec in the first post script is incorrect as "010102" (which is NV). Does that explain the DPS difference? I'm simulating that script much higher than 115k.

Last I checked that talent spec is NS+SotF+DoC with 3 unused talent points.

[Leafkiller]

In the most current SimC profile, there are the following lines for the "complex" action list:

Code: Select all

/pool_resource,wait=0.1,if=talent.dream_of_cenarius.enabled&combo_points>=5&((energy<50&buff.berserk.down)|(energy<25&buff.berserk.remains>1))&buff.savage_roar.remains-6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=4.5
/pool_resource,wait=0.1,if=talent.dream_of_cenarius.enabled&combo_points>=5&((energy<50&buff.berserk.down)|(energy<25&buff.berserk.remains>1))&buff.savage_roar.remains+6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=6.5
/ferocious_bite,if=talent.dream_of_cenarius.enabled&combo_points>=5&buff.savage_roar.remains-6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=4
/ferocious_bite,if=talent.dream_of_cenarius.enabled&combo_points>=5&buff.savage_roar.remains+6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=6

What's the purpose of the 0.5 second difference in the SR/Rip comparison between the "pool_resource" line and the corresponding "ferocious_bite" line?

This is one aggixx will need to answer as he developed that code.

[raffy]

The talent spec in the first post script is incorrect as "010102" (which is NV). Does that explain the DPS difference? I'm simulating that script much higher than 115k.

Last I checked that talent spec is NS+SotF+DoC with 3 unused talent points.

http://code.google.com/p/simulationcraf ... ers#Basics

http://us.battle.net/wow/en/tool/talent ... #UZ!010102

[aggixx]

In the most current SimC profile, there are the following lines for the "complex" action list:

Code: Select all

/pool_resource,wait=0.1,if=talent.dream_of_cenarius.enabled&combo_points>=5&((energy<50&buff.berserk.down)|(energy<25&buff.berserk.remains>1))&buff.savage_roar.remains-6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=4.5
/pool_resource,wait=0.1,if=talent.dream_of_cenarius.enabled&combo_points>=5&((energy<50&buff.berserk.down)|(energy<25&buff.berserk.remains>1))&buff.savage_roar.remains+6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=6.5
/ferocious_bite,if=talent.dream_of_cenarius.enabled&combo_points>=5&buff.savage_roar.remains-6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=4
/ferocious_bite,if=talent.dream_of_cenarius.enabled&combo_points>=5&buff.savage_roar.remains+6>=(dot.rip.remains+(8-(dot.rip.ticks_added*2)))&(dot.rip.remains+(8-(dot.rip.ticks_added*2)))>=6

What's the purpose of the 0.5 second difference in the SR/Rip comparison between the "pool_resource" line and the corresponding "ferocious_bite" line?

The wait=x parameter specifies the number of seconds the line should delay the script if the conditionals evaluate to true. pool_resource is not exactly a "wait while true" but more of a "if true then sleep for x". The threshold therefore needs to be higher for the pool line otherwise it would pool 0.1 seconds past that conditional and then the second line would no longer be true. It doesn't really need to be 0.5 higher but I believe the default value of the wait parameter is 0.5 and that's why I made it 0.5 higher, either that or because it was neater.

[Leafkiller]

The talent spec in the first post script is incorrect as "010102" (which is NV). Does that explain the DPS difference? I'm simulating that script much higher than 115k.

Last I checked that talent spec is NS+SotF+DoC with 3 unused talent points.

http://code.google.com/p/simulationcraf ... ers#Basics

http://us.battle.net/wow/en/tool/talent ... #UZ!010102

Keep in mind that when we specify the talents directly it works different than what you get when a pointer to a talent spec is used, and looks like what you initially posted where "talents=010102" is a DoC build.

[raffy]

After looking at the simc source, it does appear that when talents=integers then 0 = nothing, 1 = first talent, etc., contrary to the documentation which states it matches the armory URL format.

http://code.google.com/p/simulationcraf ... yer.cpp#83

[Leafkiller]

After looking at the simc source, it does appear that when talents=integers then 0 = nothing, 1 = first talent, etc., contrary to the documentation which states it matches the armory URL format.

http://code.google.com/p/simulationcraf ... yer.cpp#83

Yep, I just verified that "talents=http://us.battle.net/wow/en/tool/talent ... r#U!.0.0.1" and "talents=010102" are both DoC. Since I had only been dealing with the local version, I did not notice until your post that the Battle.net talent calculator uses "." for no talent defined and 0-2 for the talents in a given tier. It would not surprise me if this is new to the Cata talents and the documentation you saw used to be correct.

[Leafkiller]

Getting back on topic - specifically to Sorcerer's question, I am seeing the default T14H script in simc at just over 117k dps using 510-5 (I just downloaded it). Using the script posted here, that jumps to just over 122k dps.

[Instaqueues]

A noob question...but just trying to help and I only know enough to be dangerous

With the bug from 4.0 that was talked about recently (in that ursol's vortex thread), it was mentioned since 4.0 Rip gets extended 1 or 2 tickets (3 seconds and not 2 seconds like blizzard thinks). Is this incorporated in the script here? Or is that something beyond control?

[Leafkiller]

On a disrelated note. This post at EJB caught my attention: http://elitistjerks.com/f73/t130906-fer ... ost2227691.

I ran a test on the T14H DoC+SotF script replacing all Shreds with Mangles. I did some quick 10k iteration runs with the following results:

Mangle: 121962.2 dps 57.35 dps error. Damage breakdowns:
Rip: 26.6%
Rake: 25.2%
White: 22.0%
Mangle: 15.2%
FB: 7.8%

Shred: 122074.9 dps 60.18dps error. Damage breakdowns:
Rip 26.2%
Rake: 25.1%
White: 22.0%
Shred: 16.8%
FB: 6.8%

So dropping Shred entirely from your rotation in heroic gear with a DoC/SotF rotation will cost between 100-200 dps, as the extra combo points give you more FBs - especially with the FB energy pooling code included in the rotation.

Running a similar test with HotW gives the following dps comparisons (there is no FB energy pooling in the current simc HotW rotation):
Mangle: 115520.4
Shred: 116382.3

Note that FB is worth less in the HotW rotation (7.2% vs 7.8% or 6.2% vs 6.8%) and as a result the extra combo points from Mangle are not as beneficial.

There was a followup post on EJB: http://elitistjerks.com/f73/t130906-fer ... ost2231391
Unfortunately the poster did not specify what talent spec he tested and whether or not he was using the rotation posted here or the default simc rotation. He also suggested that an hybrid approach might be best with Mangle as fillers except during Berserk.

Note: these numbers assume you are standing behind the boss and are not subject to parries.

[Tinderhoof]

Something else to note. With the current Rip extend bug I am guessing that Mangle is going to be the better filler and it being able to extend Rip more often, and do FB more often.

[Leafkiller]

Something else to note. With the current Rip extend bug I am guessing that Mangle is going to be the better filler and it being able to extend Rip more often, and do FB more often.

I thought about that the other day, but today I was thinking about all the times the boss turns around and we get an error message on our shreds vs. parries we might get in that situation using Mangle. Certainly not getting error messages when not behind the mob makes it easier to execute a rotation...

[raffy]

A noob question...but just trying to help and I only know enough to be dangerous

With the bug from 4.0 that was talked about recently (in that ursol's vortex thread), it was mentioned since 4.0 Rip gets extended 1 or 2 tickets (3 seconds and not 2 seconds like blizzard thinks). Is this incorporated in the script here? Or is that something beyond control?

Simc implements it as 1, 1, 2.
https://code.google.com/searchframe#plx ... com&l=1682

But as my tests showed yesterday, you can make apply Rip, extend it "3" times, and get 13 ticks in-game (+5 from 3 extends).
http://fluiddruid.net/forum/viewtopic.p ... =25#p14858

In my simulator, I implement it as RandomInteger[{1,2}] extra ticks.

My sim shows Ursol's Vortex as DPS neutral if Rip extend was actually 1, 1, 1.

[aggixx]

More specifically, in SimC it's 1, 1, 2, 1, 1, 1, 1, 1.... In reality 1, 2, 1, 2, 1, 2, 1, 2 is probably more accurate. If you extend 3 times you would gain 4 ticks on your Rip with either model, but when you start extending the rip 10 or 20 times it gets pretty far off.

[Leafkiller]

Given how close Mangle dps is to Shred, at least in the DoC/SotF profile, I was thinking about times you might want to use Mangle over Shred. One that came to mind is either when Rip is down, or Rip is expiring soon and you are short on combo points/energy.

It is something I plan to look into - although the difference may be pretty small.

[aggixx]

Would be quite interesting to see some mangle weaving coming out to be a DPS gain, but I could see it being possible. Sounds a lot like Fire Mages in 4.0 Scorch weaving to keep from going OOM.

[Etapicx]

I tried with the Ursol's Vortex bug and with my other account (I used a Shaman to PoP BL for me) I had some lucky OoC procs and i was able to extend Rip with mangle to last for Berserk to come up again there i just needed to apply it. Generally i used Shreds on the OoC procs but other then that just mangle.

[Kroníc]

Hello, first time posting here. With the recent PTR 5.2 change to Nature's Vigil, where does this put it? Behind DoC/HoTW?

[Tinderhoof]

Not sure about it on it's own. What it does do is make Incarnation even worse for PVE then it already is.

[Etapicx]

Force of Nature still be bad even thought the change?