Simulationcraft - Feral rotation discussion and script

[aggixx Profile]

Alright, Althor just commited a fix for it (<3). The change was a ~775 DPS gain for T14H Incarnation, which is a 94%(?) increase in Incarnation's value. Flapjack also mentioned that the Bottle of Infinite Stars trinket, now that we have the proc information, is a large DPS gain for feral over Relic of Xuen, so I went ahead and implemented that.

Here's some updated talent comparisons, by tier (click for report):



Sidenote: I'm not exactly sure why Heart of the Wild is soooo bad.... something to look into for sure.

Edit: Man, I sure did a stellar job of cropping those graphs to the same dimensions.

[Yriss Profile]

hum, i don't understand, even with probably a little better uptime for agi proc ... This trinket has mastery at stats, and xuen has agi ... How this trinket can be better ? I know mastery is our best secondary stat, but it's not as good as agility, and we loose a lot of agility ...
Perhaps 20sec/45icd could let us buff 3 rakes/min, and not 2 with xuen ? is it enough to compensate the loss of agility ? Or pherpas it's another reason ?

At least xuen can be available for the first raid week ^^
(sorry for my english, i hope it's understandable )

[Terias Profile]

More of a curiosity thing, but has anyone considered using bear mangle/maul in the rotation deadtime?

We have abilities that take us into bear and into cat instantly and off the gcd (and without triggering the gcd), so it's possible we could do it without even losing white swings. Dash (can be glyphed to 2 min) and prowl for cat (prowl potentially being usable when berzerk falls during incarnation), and might of ursoc and growl for bear (obviously only usable on mobs that you can't taunt or won't kill you when you do taunt, like most aoe packs).

For aoe rotations you could get off a thrash/swipe before you're energy capped again too.

[Tinderhoof Profile]

hum, i don't understand, even with probably a little better uptime for agi proc ... This trinket has mastery at stats, and xuen has agi ... How this trinket can be better ? I know mastery is our best secondary stat, but it's not as good as agility, and we loose a lot of agility ...
Perhaps 20sec/45icd could let us buff 3 rakes/min, and not 2 with xuen ? is it enough to compensate the loss of agility ? Or pherpas it's another reason ?

At least xuen can be available for the first raid week ^^
(sorry for my english, i hope it's understandable )

@PAX right now so pardon the typos. My best guess is that due to DOC being so much better that buffing 2 full Rips and 3 Rakes will off set the loss of agility.

[aggixx Profile]

The proc on the Bottle makes up for the passive stats. The Bottle still has a huge chunk of mastery which is a very favorable itemization, but the proc on it is far far better than Xuen's.

Bottle: 2539 agility, 20 second duration, 45 second internal, procs on HIT, 15% proc chance
Xuen: 3027 agility, 15 second duration, 55 second internal, procs on CRIT, 20% proc chance

Bottle's proc trigger and proc chance lead it to have an average time to proc of 4.2 seconds.
Xuen's proc trigger and proc chance lead it to have an average time to proc of 8.0 seconds.

This puts Bottles average proc interval at 49.2 seconds, and Xuen's at 63 seconds, which means Bottle has an uptime of 40.65% and Xuen has an uptime of 23.81%. That is a very large difference.

PS: I noticed that all the difficulties of Bottle have the same proc value... how strange.

[Yriss Profile]

erf ... i didn't notice it was on crit only .... Xuen (str) has 45icd and on hit landing ...
it's still a good trinket, but we won't use it until t15+ ^^

Just checked trinkets @beta 16030 (dungeon journal)
Terror in mist (hero) 1300agi/7796 crit (20sec)
Bottle
- lfr: 956 mastery, 2866proc agi (20sec)
- normal: 1079 mastery, proc 3236 agi
- hero: 1300mastery, proc 3653 agi (20sec)
Perhaps proc agi from xuen has changed too, but i didn't see one :/

It can always change before release

oh, and what about this trinket ? http://www.wowhead.com/item=87079 . I saw twice a similar trinket but in stamina version (trash loot mogushan during normal 25 raid testing), so it should be trash loot too ... I saw nothing on the dungeon journal about these trash loots, so i don't know their stats in this build ... it should probably have 1300 agi, and more haste (lfr trinket has the same value on wowhead)

[aggixx Profile]

Just checked trinkets @beta 16030 (dungeon journal)
Terror in mist (hero) 1300agi/7796 crit (20sec)
Bottle
- lfr: 956 mastery, 2866proc agi (20sec)
- normal: 1079 mastery, proc 3236 agi
- hero: 1300mastery, proc 3653 agi (20sec)
Perhaps proc agi from xuen has changed too, but i didn't see one :/)

Wowhead lists them all as 2539 for some reason. Can't say I'm surprised that they changed it though! I see SimC is already using those numbers, so no big deal.

oh, and what about this trinket ? http://www.wowhead.com/item=87079 . I saw twice a similar trinket but in stamina version (trash loot mogushan during normal 25 raid testing), so it should be trash loot too ... I saw nothing on the dungeon journal about these trash loots, so i don't know their stats in this build ... it should probably have 1300 agi, and more haste (lfr trinket has the same value on wowhead)

When I tried it (replacing Xuen) a while ago it was a huge loss, like 5 or 6k DPS. It's possible this could've changed.

[Leafkiller Profile]

@aggixx - I was looking through the DoC stuff thinking about how to simplify it. I have some ideas that I want to discuss with you next time we can hook up.

I am not sure about one of the NS statements:

Y natures_swiftness,if=!buff.predatory_swiftness.up&talent.dream_of_cenarius.enabled&buff.dream_of_cenarius_damage.down&(target.health.pct<=25|(dot.rip.remains>4&set_bonus.tier14_4pc_melee))&(((target.Fluffy_Pillow.time_to_die>=(8.5+action.healing_touch.gcd)&buff.tigers_fury.up)))

This is right before the Rake code.

Is it still the case that BitW FBs are not resetting the stats snap shots?

[aggixx Profile]

I haven't checked beta, but it was the case on live when we last spoke. What aren't you sure about it?

[Leafkiller Profile]

It is not referencing Rake at all. There might be a better use for NS than that.

[aggixx Profile]

Very possible, but I believe with the (probably limited) things I tried removing that line was a loss.

[aggixx Profile]

So I was perusing the beta class balance analysis thread, and it turns out I'm really glad I did. I stumbled across this post... and preliminary findings are that he is entirely correct. Using rake as your filler anytime that you're not overwriting a stronger rake is a 2.5k DPS gain in the T13H profile with a simple action list tweak.

[Terias Profile]

Didn't want to start my own thread asking this, is it common knowledge that 0 cp savage roar can still proc PS? It's rare, but spamming it for a while just running around I started noticing ps procs.

[Tinderhoof Profile]

I had seen it on the beta, but not for a long time. You should make a post on the beta bug forums.

[aggixx Profile]

So it seems like nobody ever added Force of Nature to the feral action list... So uh.... here's a new comparison. This at level 90, using a shred rotation (click for report):


Sidenotes:

• Heart of the Wild is currently broken in the sim (at least for feral) and is yielding 0 DPS.
• I overwrote the old T4 comparison with the new one so the graph at the top of this page will have changed as well.
• Both the 85 and 90 action lists were optimized with SotF in mind. It's possible that making some tweaks (such as how aggressively the script consumes its combo points) that FoN could pull ahead. There's also probably more optimal usage involving FoN itself.

In other news, brief analysis of T4 at level 85 is that SotF is still king, and Incarnation is really bad. Like, really really bad. Using a rake rotation, Incarnation is a 150 DPS gain over not using any talent at all. That's not directly translatable for how good it is for burst, but the outlook is not bright.

[mineko Profile]

I really wanted Incarnation to be good, because the cat in armor model looks so cool. Oh well.

[aggixx Profile]

Made a couple changes that I committed earlier:

• Removed feral spirits (shaman symbiosis) from all action lists. Their damage (which is incorrect) has not been addressed yet and at this point it's more accurate to not use them at all.
• I changed Force of Nature usage. The script will precast them at the start of the fight, and then from there on only uses a spare GCD to activate them. This was a 300 or 400 DPS gain in T14H (this should be enough to pass SotF now).
• I added Thrash on OOC procs to the T13H rotation. This was about a 600 DPS gain.

[Leafkiller Profile]

Made a couple changes that I committed earlier:

• I added Thrash on OOC procs to the T13H rotation. This was about a 600 DPS gain.

So...I should update the script to use Thrash in place of AddCombo on OOC procs...

[Tinderhoof Profile]

Does the first hit of Thrash hit hard enough to justify overwriting the bleed from the last thrash?

[Leafkiller Profile]

Does the first hit of Thrash hit hard enough to justify overwriting the bleed from the last thrash?

This is what aggixx modeled:

Code: Select all

    if BuffPresent(CLEARCASTING) and TargetDebuffExpires(THRASHCAT 3 mine=1) Spell(THRASHCAT)
    if BuffPresent(CLEARCASTING) AddCombo()

so I am guessing no...

[Tinderhoof Profile]

Ok, I was just wanting to make sure it wasn't mindless Thrash spam on OOC.

[Sorcerer Profile]

I think I have to ask this here since it doesnt seem we got any attention from CMs regarding class balance thread on US forums, to which I do not have access to bump the 'issue' with mastery/dot scaling.

So can someone please report the issue with rake being main combo builder and mastery actually going ahead or almost wDPS and actually somewhat bump the issue till we get the attention? Not like there is much time left and it looks like they either missed it still or are completly overwhelmed with the issue.

I know it sounds retarded from my side but this really bothers me.

[Tinderhoof Profile]

I don't think it's retarded. The fact that it works goes aginest the design they clearly wanted for Feral. I even think that using Thrash on OOC procs is outside of their design for single target. My guild has been doing a 0% debuff run this week. We have 7/8 so far but the bloods on Deathwing move a bit on the fast side. When we finish up I plan to post my logs and show how much better it ends up being. As I have posted before I am sure they are aware of this issue, and I think it will be fixed by MoP launch. I don't think it will get fixed at 85 because it doesn't really matter right now how much damage each class can put out as long as it's not suddely terrible (like at the 4.01 launch).

[Mihir Profile]

So i herd u liek rake & thrash...

[Sorcerer Profile]

The trash OOC is something they might have overlooked, however it does work nicely overall what could be achieved by such design.

The better the player the better the benefit from it should be, only drawback I can think of is that somewhat spammy situation you are gonna miss OOC so often that the value would diminish, so general cat trash should be tweaked somehow to be more beneficial to use, maybe increase a little direct dmg + scaling. Wish they did keep the ability split so it would be easier to buff/nerf without consequences to bear spec for example.
Heck, they could give cat trash an extra fuction beside what it currently does "increase attack speed by X%" or somethnig along the lines of giving "extra energy regen [per hit] while buff is active on 1 target", so many options that could give a shot tbh.

This would increase value of haste a little bit if for example to not make it clearly inferior if we called such fuction that every white hit while trash is up on a target would generate 0.5-1 energy or sth along the lines.

Even than, going by what above line says, nerfing trash dmg + scaling so it's pointless on single target even with OOC, having extra regen would be beneficial for aoe phases where trash is already mandatory. Nice way to smoothen out our AOE with swipe due to higher energy regen in such phase. That scenario I'd think of considering how RNG OOC can be sometimes, couse this could really make our single target dps vary by insane amount unless we hard casted 50 energy trash every 15sec.