Thoughts on Different Talents (and Stat Weights)

[Twintop]

UPDATE 2012-08-09: Put new charts at the front, left the original post at the end.

A few updated/new charts for you all to gnaw on some more, from this post:

http://warcraft.twintop-tahoe.com/?p=191










----------------------------------------------------
8/3/2012
A few updated/new charts for you all to gnaw on some more, from this post:


http://warcraft.twintop-tahoe.com/?p=150












----------------------------------------------------

Last night I posted new blog post where I tackle the topic of which talents are best for different encounter types (and the stat weights associated).

A few charts for you all to gnaw on:







Highlights:

- Shadow Word: Insanity is awful -- if you spec in to it for PvE you're a masochist (and not the good kind).
- Mind Bender is the best Level 45 talent -- period.
- Twist of Fate is the best Level 75 talent in most cases.
- Talent choices only cause your DPS to vary by ~4% in all fight types except for those with heavy movement.
- For Heavy Movement, the best Level 75 Talent is Divine Insight.
- Stat priority is: Int > Spell Power > Spirit = Hit > Haste > Crit > Mastery

If you're interested in my long-winded analysis, check out the full post: http://warcraft.twintop-tahoe.com/?p=117

[Deleted]

interesting post thanks for sharing.

This really makes me a happy priest as that's why my wild guesswork was telling me anyway and in general I find those talents to be the most intuitive / fun.

Also nice to see that most of the talents are fairly close in performance, enough so that changing during farm to keep things fresh is viable.

any word on the lvl90 talents?

[Kaesebrezen]

any word on the lvl90 talents?

The only one using rotational on a single-target fight is Halo at 25-30 yards.

Also nice to see that most of the talents are fairly close in performance

The talents in the lvl 45 row are a dps increase +0% up to 3%. 0% is SW:I, FDCL is at a mere 1.5%, and Mindbender wins at 3%. Same for the lvl 75 talents.

They're close in performance because their impact on the performance is minimal.

[Flaubertlives]

Really nice job. Interesting (and a shame) to see that DI is now the worst choice on anything apart from heavy movement (or muti DoTing one can assume).

Am glad that Mind Bender is above FDCL, it just seems counter intuitive to have to Spike randomly.

[Crudelitas]

i expected that DI would be higher than ToF in any situation. i didn't crunch any numbers ofc. i'm not a math person it was just a guess.

[Kaesebrezen]

i expected that DI would be higher than ToF in any situation.

I think it was until the DP nerf.

[Jigawatts]

It is nice to see that the buff to mastery was very helpful. It still is our worst stat (and one of the stats has to eventually come in last), but at least we won't be running from it as if it was the plague.

Overall I am a little underwhelmed by the DPS increase of all of our tiers that do anything to increase DPS. It is such a small increase (0% to 3%) that it almost seems uninteresting. I was hoping for a more substantial 5% for any talent that increases DPS. The positive is that at least they are close, although in the end there is still a clear winner in most situations (especially with the Mind Bender tier), we will see what future patches bring

[Arlee]

Hey Twin, would you mind possibly explaining this graph real quick. It seems like an average of the different results for each stat but your values in the All line seem to not show any relations (Mastery at 0.4096 on the first two lines but at 0.4423 on the last for example).

[Twintop]

Hey Twin, would you mind possibly explaining this graph real quick. It seems like an average of the different results for each stat but your values in the All line seem to not show any relations (Mastery at 0.4096 on the first two lines but at 0.4423 on the last for example).

Sure thing, Arlee. From the blog post:

Lets take a closer look at a few of these, specifically the highest DPS for each of the 5 fight types, the two highest DPS for each of the fight types, and all 45 sims averaged together

Best Per Fight = The #1 highest DPS talent combinations for each fight type, averaged together. This would be MB+TOF for Helterskelter/LightMovement/Patchwerk/Ultraxion and MB+DI for HeavyMovement.
Best Two Per Fight = The #1 and #2 highest DPS talent combinations for each fight type, averaged together. This would be all of the values from the last grouping, plus the next highest DPS talent combos
All = All (#1 through #9) DPS talent combinations for each fight type, averaged together.

Because the top 2 talent combinations are fairly close, and will have more of a difference thanks to player style than mathematical superiority, I'm using the second line (Best Two Per Fight) for my Stat weights for things like the BiS list and stat priorities.

[Flaubertlives]

I'm assuming by "best per fight" and "best two per fight" he's just including 1/9 and 2/9 respectively of the talent combinations. And all is an average of all 9.

Mastery is higher in "all" due to 2 of the talents, FDCL and DI, scaling heavily with mastery. Whereas the "best" has mastery lower due to the lack of FDCL in them (as it's beaten by mindbender in all cases)

EDIT:

Twin beat me to it anyway.

[Twintop]

I'm assuming by "best per fight" and "best two per fight" he's just including 1/9 and 2/9 respectively of the talent combinations. And all is an average of all 9.

Mastery is higher in "all" due to 2 of the talents, FDCL and DI, scaling heavily with mastery. Whereas the "best" has mastery lower due to the lack of FDCL in them (as it's beaten by mindbender in all cases)

EDIT:

Twin beat me to it anyway.

Emphasis mine because this is spot-on. This is also the reason why in the last line Haste is the worst, because SWI drags the value of Haste in to a culvert and does inappropriate things to it.

[Arlee]

Emphasis mine because this is spot-on. This is also the reason why in the last line Haste is the worst, because SWI drags the value of Haste in to a culvert and does inappropriate things to it.

OIC, thanks My brain is slightly fuzzy due to pain meds for my severely sprained wrist I was just looked at it and being really confused. Didn't even notice the blog post link >< Thanks for the explanation though

[Kaesebrezen]

Mastery is higher in "all" due to 2 of the talents, FDCL

I dare to say: By the end of MoP, FDCL will be either a DPS loss or worse than it's now.

Explanation: Mastery and Haste increase the damage (per execute/time) of Mindflay, while Mindspike damage only scales with Spellpower. At a certain haste/mastery point, Mindflay will always be stronger than FDCL, and that point will be reached (using the stats of the QA 460 itemlvl items) by the last raiding tier.

Mindspikes greater SP coefficient might be enough to offset some of thoses effects, but in the end this will result in Mindbender getting better over the course of the addon, while FDCL will stay the same.

[Innoruuk]

Thanks for your work on this, Twintop. It was a very interesting read and Im glad to have found your blog! I do have a question for you or whomever.

With Twist of Fate performing so well on all of these fights (outside of heavy movement) and all of the sims being a single target, is it safe to assume that any fights with adds will only further increase the value of Twist of Fate? By way of increased uptime through killing adds. Or do you predict fights with adds is where Divine Insight will outshine ToF due to increased procs from multi-dotting?

Also, I have seen it asked in other threads but not definitively answered, has it been tested if Vampiric Embrace heals on a <20% health ally trigger the effect? If so I assume this and smart use of (Dark Binding glyphed) Prayer of Mending and/or Renew will further increase the value of Twist of Fate?

Thanks again for the time put in to this!

[Deleted]

Thanks for sharing, TT.

Awesome testing, cheers

[Twintop]

With Twist of Fate performing so well on all of these fights (outside of heavy movement) and all of the sims being a single target, is it safe to assume that any fights with adds will only further increase the value of Twist of Fate? By way of increased uptime through killing adds. Or do you predict fights with adds is where Divine Insight will outshine ToF due to increased procs from multi-dotting?

Excellent question! There are a lot of different use cases to think about:

If the adds are short lived (e.g., Rhyolith, Ragnaros in FL, Magmaw) ToF will probably be better because you will have more time to put the increased damage towards a full rotation on the boss.

If you have adds at phase transitions right before a burst of damage is needed (e.g., Spine of Deathwing, getting Nefarian in BWD as close to the next 10% point before phase 3), ToF is probably good as well because the extra damage gives you some much needed burst right out the gate.

If you have adds that are long lived and are offtanked (e.g., Golemagg in MC, phase 1 Nefarian in BWD when Onyxia is up, Madness of Deathwing), DI will have double or even triple the chance to proc (because you'll have DoTs up on two or three targets). I would think that two targets would be optimal, as any more than that would cause the casting rotation seem very disjointed and make it difficult to keep all DoTs up.

Really it comes down to looking at individual fight mechanics and seeing which talent can abuse the fight mechanics the best (just like Sorrowsong was used on Heroic Ragnaros).

Also, I have seen it asked in other threads but not definitively answered, has it been tested if Vampiric Embrace heals on a <20% health ally trigger the effect?

This is an interesting question. While I haven't tested it, I do know that, traditionally, heals from Vampiric Embrace and Devouring Plague's on hit or DoT damage (in Cataclysm) has not proc'd trinkets/enchants that require 'on healing'. Now that VE is a raid cooldown, I would expect it to act like direct heals instead of passive secondary heals, but it might not.

If so I assume this and smart use of (Dark Binding glyphed) Prayer of Mending and/or Renew will further increase the value of Twist of Fate?

If you get a Renew to tick on a target below 20% (which isn't going to be very often since it takes 3sec, minus some with haste, to get the first tick) or PoM pops on a target that goes below 20%, then yes you will experience the ToF bonus. This isn't going to happen very often, though! Think of it like this:

If ToF gives 15% more damage for 10 seconds, you'll have ~7 GCDs to make use of the increased damage. How much DPS could you do in the 1 GCD it would take to cast Renew or PoM on a target? Will the increased 15% damage outweigh the loss of that 1 GCD it took to cast? There are likely situations where this could occur (3orb DP getting close to being ready, reapplying DoTs, Hero/Lust on the way, etc.) but with the uncertainty that you'll get PoM or a Renew tick to hit a target below 20% HP is pretty low, especially if you have decent healers in your group.

I wouldn't worry about it unless the stars align in a fight.

[h3lladvocate]

I dare to say: By the end of MoP, FDCL will be either a DPS loss or worse than it's now.

Explanation: Mastery and Haste increase the damage (per execute/time) of Mindflay, while Mindspike damage only scales with Spellpower. At a certain haste/mastery point, Mindflay will always be stronger than FDCL, and that point will be reached (using the stats of the QA 460 itemlvl items) by the last raiding tier.

Mindspikes greater SP coefficient might be enough to offset some of thoses effects, but in the end this will result in Mindbender getting better over the course of the addon, while FDCL will stay the same.

Wha??? Do you have math to show this? Could post on beta feedback, and maybe get a buff to FDCL?

[Flaubertlives]

Wha??? Do you have math to show this? Could post on beta feedback, and maybe get a buff to FDCL?

I'd not thought about it before, but logically, it makes sense. MS does scale with haste to the extent of scaling with the GCD, but that's not a whole lot.

[Deleted]

can you get ToF to proc with PW: S on a target below 20% health?
If so fighting for that weakened soul on the tank against other spriests and the disc priest is gonna be fuuuun.

[Kaesebrezen]

Wha??? Do you have math to show this? Could post on beta feedback, and maybe get a buff to FDCL?

Let's keep it simple

Add "&buff.bloodlust.react=0" to the mindspike line in the profile, you will no longer cast mindspike during bloodlust, yet the dps doesn't change. We now can conclude, that using the FDCL buff during Bloodlust is not a DPS increase. Mindbender on the other hand does get 1 or 2 additional attacks of when affected by bloodlust.

Next is the fact, that any increase in mastery increases the basically increases the damage of Mindflay (we have a procrate of 23% when raidbuffed on mastery)
The DPS gain from FDCL is a result of MSp dealing more damage than Mindflay. Mastery makes that gain smaller. For Mindbender, this is not the case.
Yes, Mindspike has a higher SP coefficient, but that is not enough to compete against mastery/haste for Mindflay.

Further, siming lvl 85 endgear gearsets with Mindspike and comparing it to not using mindspike has FDCL as a 0,75% dps increase. That's the same result my sims inthe Q&A itemlvl 560 set gave me. FDCL geats weaker the better your gear is, Mindbender gets stronger the better your gear is.

and maybe get a buff to FDCL?

FDCL is A LOT worse than Mindbender. It's in the same boat as SW:I. Both are close to no DPS increase, given by their mechanic. Even worse is, that buffing their damage to compte with Mindbender also does NOT work.
Both SW:I and Mindspike (with the 20% from the talent) are already relativly hard hitting abilities, you can't buff the damage further.
SW:I even got a damage nerf because it was hitting for too much, and the killing blow to FDCL was the Mindflay buff to compensate for the DP nerf.

Both talents need a complete redesign (honestly, i'd say scrap that whole talent tier, both for healing and shadow priests and give it another theme).

What i recently was thinking off (a raid support tier much like the shaman lvl 75 tier):
- Talent 1: Vampiric Dominance
- Talent 2: Buff to the raid cooldowns (DH, Barrier and VE), reducing cooldown and something like the Barrier Glyph for Barrier, a damage reduction buff for DH, a smart-heal function for VE. This would require a swap for the lvl 75 and 45 rows.
- Talent 3: Cascade - remove the damage, it's already an either/or spell compared to halo/DS, and give us another lvl 90 talent.