Register

Pawkets' Feral DPS Guide & FAQ -- Updated for 6.0!

Face-rippin fun.

Moderator: Forum Administrators

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Fri Jul 12, 2013 5:53 am

If both the trinkets you listed are normal versions then you probably don't need to bother with getting Rune, it scales really well with more gear but in normal mode gear it's not significantly ahead of other trinkets (definitely grab the normal one when you can get your hands on it though). Bad Juju is only better than an equal level Renataki's if you have more expertise than you can use.

That's just general advice, though. Would always recommend doing the comparisons on your own character using Catus if you want to know for sure. The Simulation: Distribution option at the bottom of the window {set the iterations to like 5000-1000} will tell you how much DPS a set of gear does, and then you can go from there.
Image

Posts: 16
Joined: Sat Jul 06, 2013 12:26 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Ayleena » Fri Jul 12, 2013 4:27 pm

Thanks for the tips. I'm still running LFR Lei Shen but not rolling for the rune. If it drops, it drops - might try it out. Normal might not ever happen. ToT has been a rough raiding patch for me. 3/12N with a 526 ilevel.

Posts: 28
Joined: Sun Jan 13, 2013 12:11 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Spanja » Wed Jul 31, 2013 9:49 pm

Hello Aggixx, how do you know that crit > agi on your current stuff ? (I see that you gem +320 crit in yellow socket). Do you see it on Catus ?

Revered
User avatar
Posts: 270
Joined: Fri Dec 24, 2010 1:57 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Steakbomb » Thu Aug 01, 2013 12:11 am

Spanja wrote:Hello Aggixx, how do you know that crit > agi on your current stuff ? (I see that you gem +320 crit in yellow socket). Do you see it on Catus ?


He gems that way to get more secondary stats for rune proc and for better Rip/Rake when Rune isn't up.

Im gemmed the same way except I gem haste as in my current gear I get better stats with haste rather than crit.

Gemming this way you gain roughly ~1000 in all your secondaries Mastery/Haste/Crit and lose about 2k Agility but the secondaries are worth it.

Sims about 2k higher than an Agi build.
Image

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Thu Aug 01, 2013 1:59 am

Steakbomb wrote:Sims about 2k higher than an Agi build.

Haven't seen anything close to a figure like this, at least not in T15 gear.

There has since been bugs fixed in Catus that may change the outcome, but I originally decided to go for all crit gems for better overlap with Guardian (no need to regem). The current version of Catus at the time indicated it was about a 200 DPS loss on average, but the variance was showing to be also higher by 2-3k DPS with the lower DPS secondary gem set which is more suitable for ranking as well (since you're really not going to get a strong rank without above average, or at least prominent, RoR uptime anyway).

Here's the up to date damage comparison between secondary versus primaries for my gear (546 equipped, 4p, 543 Weapon & Rena's, 536 Rune):
Code: Select all
PRIMARY
DPS: Median    = 236234
DPS: Mean      = 236751 +/- 97
DPS: Range     = 206972 -- 283628
SECONDARY
DPS: Median    = 236224
DPS: Mean      = 236698 +/- 100
DPS: Range     = 205454 -- 278245
SECONDARY (Crit gems only for maximal Guardian overlap, looser 1:1:1 fit than a fully flexible gem match)
DPS: Median    = 236300
DPS: Mean      = 236675 +/- 100
DPS: Range     = 205179 -- 283270


So basically all of the sets are more or less exactly the same; they're all so close together that they're within the margin of error at 10k iterations. If you have any interest in playing Guardian then you might as well go full crit gems (or as much as your gear can support, mine is crit light so I can afford to do it in every socket), assuming that you have the exact same gear as me of course. If you really want to know if it's worth doing for your gear you should use Catus to sim it yourself, all you need is the Reforge, Regem, and Simulator: Distribution features. The lower item level your gear (and especially your RoR) is, the less worth it it will be.
Image

Revered
User avatar
Posts: 270
Joined: Fri Dec 24, 2010 1:57 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Steakbomb » Thu Aug 01, 2013 3:27 am

Yea I reran the sim in Catus after an update and got some diff numbers. My comp locked up so I would have to redo them to post numbers but I did just redo my gem and reforge to Agi/Haste and it sims a bit higher than I was using.

545 equipped, 543 Weapon, Renataki's and 530 Rune

http://us.battle.net/wow/en/character/B ... b/advanced

I also run HotW for a lot of the more difficult fights but run DoC on farm bosses, although I think I am going to start running DoC on H Lei Shen as we don't have the need for my Tranq anymore.
Image

Posts: 28
Joined: Sun Jan 13, 2013 12:11 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Spanja » Thu Aug 01, 2013 5:14 am

Ok thanks, I see. Here's my current gear : http://eu.battle.net/wow/fr/character/k ... a/advanced

Not exactly like yours, but it's just a matter of time ^^
I was wondering if sec. stats were better than agi on my gear too, so I'll sim it.

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Wed Aug 07, 2013 8:26 pm

Some PTR trinket comparisons as of build 17271 (this includes Renataki's, Juju, and Ticking Ebon Detonator being changed to 10 seconds). The character used is full 553 (T16 normal mode) with the 4-set bonus.

WITHOUT Rune of Re-Origination:
http://dl.dropboxusercontent.com/u/1086 ... orune.html

WITH Rune of Re-Origination:
http://dl.dropboxusercontent.com/u/1086 ... sults.html

And a comparison between the highest simming RoR profile (BiS), the highest simming non-RoR profile, and the highest simming non-RoR profile without set bonuses. The basic idea is to show the gap in DPS between the upper and lower end of still fairly ideal gear selection within the same item level restrictions:
Image
Image

Honored
Posts: 107
Joined: Thu Jun 30, 2011 4:05 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby RareBeast » Thu Aug 08, 2013 8:00 pm

So that means that the current valor trinket is better than the normal mode (same ilvl) Rentaki & Juju?

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Thu Aug 08, 2013 8:48 pm

I only simmed heroic 2/2 upgraded T15 trinkets as well as the VP trinket, it's not intended to be a full ranking just an idea of what to expect going into the next tier.
Image

Revered
User avatar
Posts: 325
Joined: Tue Oct 23, 2012 4:19 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Alpheus » Fri Aug 09, 2013 3:55 am

double post
Last edited by Alpheus on Sat Aug 10, 2013 10:31 am, edited 1 time in total.
Image

Revered
User avatar
Posts: 325
Joined: Tue Oct 23, 2012 4:19 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Alpheus » Fri Aug 09, 2013 6:36 am

aggixx wrote:I only simmed heroic 2/2 upgraded T15 trinkets as well as the VP trinket, it's not intended to be a full ranking just an idea of what to expect going into the next tier.


Something seems to have eaten my post, maybe cause I quoted with an image. Anyway -

Are you sure SimC has correctly implemented the cooldown reduction on AoC? I noticed that it was ranked near identical to the multistrike trinket (which has the same agility proc) - assuming that multistrike can't be "perfectly" modeled I would at least assume AoC being quite a bit ahead of it and to be honest, judging by raffy's work, I would assume it to be even higher on dps than TED.
Image

Revered
Posts: 378
Joined: Tue Feb 08, 2011 5:51 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby ShmooDude » Fri Aug 09, 2013 10:38 am

Alpheus wrote:
aggixx wrote:I only simmed heroic 2/2 upgraded T15 trinkets as well as the VP trinket, it's not intended to be a full ranking just an idea of what to expect going into the next tier.


Something seems to have eaten my post, maybe cause I quoted with an image. Anyway -

Are you sure SimC has correctly implemented the cooldown reduction on AoC? I noticed that it was ranked near identical to the multistrike trinket (which has the same agility proc) - assuming that multistrike can't be "perfectly" modeled I would at least assume AoC being quite a bit ahead of it and to be honest, judging by raffy's work, I would assume it to be even higher on dps than TED.


AoC @ 572 in the sim has a usage interval of every 21.59 which sounds about right (20.5 is what I got as the minimum if I did my math right) and berserk is 129.4 which also looks about right.

Looking at the numbers, AoC mostly results in higher Shred numbers (due to more berserk) about the same mangles, better uptime on Rip and more Ferocious Bites.

TED however results in much higher average bleeds (~15k on Rake and ~10k on Rip). The DPS gain from the higher bleeds outweighs the filler/FB damage of AoC.

That said, there may be tweaks that can be made to the AoC rotation to pump it up a little (first thing that comes to mind is clipping Rip if the ratio is >=100 and TF is up so that we always have a TF up, dunno if that'd work well or not, would have to try it obviously).

EDIT: One thing that might change things is the catus run is in T15 gear but the simc run is T16.

The other thing to note is that while TED's average is higher, its delta is also way higher. AoC is much more consistent in dps. AoC is 276-355 in simc and 234-292 in catus. TED however is 273-375 in simc and 183-303 in catus. Catus seems to have a lower delta for AoC and a higher one for TED. Dunno if that's the gear difference (set bonuses could certainly make a difference) or something else going on. Will be easier to compare when we get like to like sims in both.

Exalted
User avatar
Posts: 1650
Joined: Mon May 24, 2010 10:21 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Tinderhoof » Fri Aug 09, 2013 2:15 pm

I think TED is being inflated more by the Sims then it should be as the computer can make much better use of the proc than a human can. For bleeds to get the bigger benefit you need to refresh them REALLY fast. The window of usefulness is very small as you quickly lose the size of the agi proc. Now the extra crit will help some for combo points, but ya I am guessing you are going to see wild swings in output from fight to fight.

I can see this being a trinket you use on a farm fight to parse later in the tier. During progression it really could go either way depending on how lucky you get with the timing on the proc.

Revered
Posts: 224
Joined: Tue Jun 28, 2011 10:41 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Dysheki » Fri Aug 09, 2013 2:29 pm

I agree with Tinderhoof. I think the versatility AoC/4T16 gives will be beneficial for progression fights to get things down quickly that need to die ASAP (our weakness).

For parsing or for single target fights TED would come out on top but the variance will be huge because you might get those fights where you have the 5 CPs ready to go for a lot of procs but others you're fighting uphill to get a rip going.

This is all assumption based on absolutely no math done by me. I'm a practical person and I let people smarter than me figure out that stuff =P
Image

Exalted
User avatar
Posts: 890
Joined: Tue Oct 23, 2012 7:15 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby raffy » Fri Aug 09, 2013 8:15 pm

IMO, you'll need really good reflexes for TED (and luck); I keep simming the 10sec Soul Charm over 10sec TED, especially as I increase GCD latency and/or reaction time. It's very difficult to have 5cp available right when TED procs, although AoC and 4pT16 make this more likely. Possibly some of this is also due to action list issues, ie. the lack of "Oh, we have a fading buff, so keep refreshing stuff." I've been hesitant about diverging the Catus list from the Simc action list because Aggixx (and others) have been very good at keeping it up to date. If you miss the first few seconds of the proc, Soul Charm is better (and it has higher RPPM) but has Exp whereas TED has Mastery. I also think the Multistrike trinket is viable, but I need more in-game testing to verify what procs it.

Exalted
User avatar
Posts: 503
Joined: Wed Jun 29, 2011 4:49 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Stenhaldi » Fri Aug 09, 2013 9:17 pm

Yeah, the fact that Renataki's is simming below TED and significantly below re-origination makes me suspect that the action list may not be handling Renataki's very well.

Comparing Renataki's to 553 TED, the TED proc grants about 18% more average agility over its duration (21380*.95/17150) but Renataki's has a 20% larger proc rate. So comparing average agility, TED is already behind -- and then considering that you can almost always land a rip with high-stack Renataki's and almost never with high-stack TED, this should push Renataki's easily ahead.

Comparing Renataki's to re-origination, the Renataki's proc is only slightly weaker for bleeds (not scaling as well, obviously), is better for non-bleeds, and only has a 10% smaller proc rate. The difference between the two should not be very large. (Re-origination does grant better passive stats, but also costs some damage passively by constraining secondary stats to 1:1:1.)

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Fri Aug 09, 2013 11:31 pm

Tinderhoof wrote:I think TED is being inflated more by the Sims then it should be as the computer can make much better use of the proc than a human can.


While it's true that it is inflated some amount in SimC at the moment, it isn't because it's a computer and not a human, it's because SimC currently doesn't support "reactions" to stat changes. Supporting such a thing would be quite complicated, if raffy can do an analysis of it via Catus easier than that would be a better plan.

Renataki's isn't handled optimally in the action list because there's no logic incentivizing waiting for the stacks to accumulate before doing a rip clip.
Image

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Sat Aug 10, 2013 6:42 pm

I've updated my trinket results after confirming all of the (relevant) trinkets are using a 10 second ICD. As far as I know the only one that was not using such an ICD in the previous results was Haromm's.
Image

Revered
User avatar
Posts: 325
Joined: Tue Oct 23, 2012 4:19 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Alpheus » Sat Aug 10, 2013 6:56 pm

aggixx wrote:I've updated my trinket results after confirming all of the (relevant) trinkets are using a 10 second ICD. As far as I know the only one that was not using such an ICD in the previous results was Haromm's.


The non-Rune sims are using a generic trinket to stay relevant (compared to leaving the slot empty). As we all know how much of a difference Renataki makes with Fune (hell, we saw how much of a difference synced procs from the VP trinket made), would it be possible to pair them up with the most likely candidate (ie. TED/AoC). Perhaps Raffy will make another trinket table in T16 BiS gear and we'll know more but until then I believe posting the non-Rune table might be a bit misleading.
Image

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Sat Aug 10, 2013 8:44 pm

The problem with "arbitrarily" selecting one trinket to be the one that goes in the second slot is, how do you then value that trinket? Using two of the same trinket on one character can cause some strange consequences (like using two AoCs and having a super short TF CD).
Image

Revered
Posts: 378
Joined: Tue Feb 08, 2011 5:51 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby ShmooDude » Sat Aug 10, 2013 10:00 pm

The only "correct" way to do it is to compute all possible trinket combinations like raffy does. The problem with this is there's so many combinations possible now with 15 potential ilvls for every trinket (or 5 if you assume 2/2 upgrade) that it would take forever to compute them all (my understanding is simcraft takes a lot longer to do this kind of thing than catus). Even when raffy does do it he generally takes a small subset at a time (around the same ilvl) and sticks to those.

As far as using actual trinkets, perhaps choose an ilvl of each trinket and then do a list using that as the 2nd and then every other trinket. You've already got a rune comparison, just add (probably normal) TED, HT, AoC and 5.4 Mastery to the potential "base" trinkets. Of course this is a lot more permutations so it'd take quite a bit longer. Alternatively, simply do what raffy does and compute all possible 2 trinket combinations of an ilvl.

Whether this would change the results or not, I don't know. Though the rune vs not rune rankings are decently different (AoC is way better w/ rune than w/o, couple other swaps in the orders as well)

Exalted
User avatar
Posts: 1331
Joined: Fri Nov 25, 2011 7:49 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby aggixx » Sun Aug 11, 2013 2:19 am

Honestly I don't think it really matters all that much. The non-rune rankings are more of a courtesy to the players without it than anything else; you only need a normal mode one and then you don't even care about the non-rune rankings anyway. And I don't suspect a placeholder trinket is really all that inaccurate, out of all the combinations of non-rune trinkets probably only a very very small minority of them have any meaningful synergy whatsoever.
Image

Revered
User avatar
Posts: 325
Joined: Tue Oct 23, 2012 4:19 am

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby Alpheus » Sun Aug 11, 2013 7:07 am

On PTR I'm noticing a lot of Rune procs now that the ICD is reduced. Literally every pull goes something like this: Rune + Blades, Rune, Rune, back-to-back. I understand that it might be due to the bad luck streak prevention (at least the first proc) and then just luck from the others, but I had 4 pulls in the past 15 minutes and every one of them was exactly like this. Getting a Rune proc after a previous Rune really serves no purpose since both Rip and rake will be up with sufficient uptime.
Image

Revered
Posts: 378
Joined: Tue Feb 08, 2011 5:51 pm

Re: Talent & Trinket Analysis (w/ Stat Weights!)

Postby ShmooDude » Sun Aug 11, 2013 7:46 am

Alpheus wrote:On PTR I'm noticing a lot of RoRo procs now that the ICD is reduced. Literally every pull goes something like this: RoRo + Blades, RoRo, RoRo, back-to-back. I understand that it might be due to the bad luck streak prevention (at least the first proc) and then just luck from the others, but I had 4 pulls in the past 15 minutes and every one of them was exactly like this. Getting a RoRo proc after a previous RoRo really serves no purpose since both Rip and rake will be up with sufficient uptime.


Maybe for DoC, but HotW largely spams Rake during rune for more DPS (to the point where rake is our number 1 damage done where as its Rip for DoC).

Also, it might (almost defiantly) be worth trying to clip a rip early at the tail end of a rune proc to keep a strong bleed up. This wasn't worth simming before because with the ICD basically ensured that "back to back" rune procs would result in an overwrite anyhow (with 22 seconds being about the Rip duration anyhow).

PreviousNext

Return to Kitty DPS

Who is online

Users browsing this forum: Elfreako, Google Adsense [Bot], Untamable and 11 guests