5.2 ww trinkets/meta considerations

[Deleted]

so lets take a look at the bluepost again.

Meta:

Capacitive Primal Diamond – 15.00 base RealPPM on landing melee/ranged abilities and swings. No ICD. At 5 stacks, fires Lightning Strike, which deals [280 + 75% AP] Nature damage. That base proc rate is multiplied by an additional coefficient by spec:

• Assassination: 1.535
• Combat: 0.99
• Subtlety: 0.98
• Feral: 1.934
• Windwalker: 1.084
• Beast Mastery: 1.604
• Marksmanship: 1.594
• Survival: 1.449
• Enhancement: 1.093
• Retribution: 1.923
• Frost DK (2H): 1.309
• Frost DK (DW): 1.572
• Unholy: 1.34
• Arms: 1.771
• Fury (TG): 1.784
• Fury (SMF): 1.951

Why is the PPM different by spec considering that realPPM is supposed to give every class/spec roughly the same proc rate already?
Three reasons. A) RealPPM procs still scale more frequently with haste, which many specs have drastically more or less of. B) The proc that it gives you in this case scales directly with AP, which many specs have drastically more or less of. C) It’s not terribly uncommon for A or B (but not usually both) to occur on a trinket, and we can swallow that much variance.

as far as I can see, we will get a relatively low coefficient.
I don't play every single melee class, but I don't think we got a lot more ap than other classes, do we?
I believe they put such a low coefficient because they think we always go full into haste (which most of the casual players actually do, its much easier to play) but I tested a lot on the ptr and my thoughts are that

• haste (4-5k) > mastery > crit
• haste (4-5k) > mastery > haste > crit

these 2 are the only possibilities.

this, in my opinion, would make the meta much weaker for a windwalker compared to most other classes.



Trinkets:

Vicious Talisman of the Shado-Pan Assault – 15% chance on landing harmful abilities and spells, and melee/ranged abilities and swings. 105 sec ICD.
(~20% uptime)

Renataki's Soul Charm – 0.56 RealPPM on landing harmful abilities and spells, and melee/ranged abilities and swings. 22 sec ICD.


Talisman of Bloodlust – 3.00 RealPPM on landing harmful abilities and spells, and melee/ranged abilities and swings. No ICD.

Bad Juju – 0.50 RealPPM on landing harmful abilities and spells, and melee/ranged abilities and swings. No ICD. Voodoo Gnomes are mostly for flavor; they deal ~200 damage before armor per hit.

Rune of Re-Origination – 0.46 RealPPM on landing harmful abilities and spells, and melee/ranged abilities and swings. 22 sec ICD


I see no point to use Talisman of Bloodlust + Rune of Re-Origination, because you will probably never want to sacrifice crit+mastery to boost your haste even more. you will very likely always have 5 stacks of the hastebuff during bossfights with no downtime.

which 2 would you prefer?
I think Talisman of Bloodlust and Bad Juju will work good together. probably my choice.

[WonderPoop]

I dont even know why they are putting in the Run of Re-Origination with that proc. 0 mastery and 0 crit for 6-7k more haste? Why would anyone want this?

[Areohotah]

because haste wont have to be your highest stat next patch with the ilvl increases. sure we have to get another 1k haste or so but it wont make a big difference especially with the 4pc

[Callimonk]

Talisman of Bloodlust + Rune sound like a totally terrible choice.

It's interesting to me that Blizz gives us so many things to encourage us to drop haste - energy spheres, 15% extra regen, etc - and expect us to have a higher amount of haste.

I foresee Bad Juju and Rune being good choices, or Bad Juju + Talisman. Obv as discussed in another thread, Rune will need to be cancelled out/removed with TeB being used immediately. Depending on how it works, Renataki's would be a good next choice.

[Totaltotemic]

I couldn't care less about how strong the meta is specifically for WW compared to other classes, it's better than the one we currently use and that's all that really matters.


That said, of the 5 trinkets in 5.2 we can eliminate 2 that won't be useful: Rune of Re-Origination (which is god-awful, no spec wants it in the game because lowering stats doesn't work when you already gear for as little of "bad" stats as possible) and Talisman of Bloodlust because we simply cannot use haste procs, Bloodlust/Heroism being the primary example because we Energy cap too easily.

The Shado-Pan Assault trinket and Bad Juju are budgeted the same except one has Hit and one has Mastery as the passive, both of which can be reforged if needed. They'll be even with the exception that the VP trinket does not have a Heroic version and one is a more random pattern while the other is much more predictable.

Renataki's Soul Charm has the most potential for DPS because we can see the proc, pump up TEB, and unleash TEB once it hits 15 seconds left and it'll gain more of a benefit during that time frame than the other trinkets.

[Callimonk]

Most of my math is showing SPA trinket to not be much of an upgrade in comparison to - say - H Elegon trinket. H Elegon has a higher uptime, if I recall, and this one is showing to be a very small Agi increase. Plus, the passive Mastery on Elegon will be more useful to us than the passive Hit. What I'll probably look into is whether or not Terror would be replaced before Elegon - while crit remains strong for us, Mastery seems like (in practice) it will be stronger.

So, I think for someone without H Elegon, SPA will be useful. With Elegon, the difference will be nearly negligible.

Bad Juju, on the other hand, should be pretty easy to attain, given that it drops from the first boss. Yah, the boss has a very large loot table, but point being that it'll be a boss that will be on farm much faster.

With Renatiki stacking, and having an incredibly low ICD, I've changed my mind about it. Even with passive Expertise, that one looks like it could be amazing.

[Deleted]

The Shado-Pan Assault trinket and Bad Juju are budgeted the same except one has Hit and one has Mastery as the passive, both of which can be reforged if needed. They'll be even with the exception that the VP trinket does not have a Heroic version and one is a more random pattern while the other is much more predictable.

idk, shado-pan trinket will be static at around ~20% uptime. In theory Bad Juju could have a 100% uptime, because there is no ICD. sure the RealPPM are low, but you higher the chance of a proc by stacking haste and thats what the Talisman gives you - a lot of haste with a very high uptime.

[Totaltotemic]

In theory Bad Juju could have a 100% uptime, because there is no ICD. sure the RealPPM are low, but you higher the chance of a proc by stacking haste and thats what the Talisman gives you - a lot of haste with a very high uptime.

That's..... not how RPPM works. Yes, haste does increase the frequency of the proc, but only by the amount of haste you have. To realistically get to the average proc rate giving 100% uptime, you'd need 600% haste. Now, will our average level haste make it to 18% uptime instead of, say, 16%? Sure, but not any more than that.

[Hinalover]

Well I just tested on the PTR that, at least based on the tooltip, the Tigereye Brew buff does NOT update it's proc value due to different Mastery buffs. What I did was I went and got the Dominators Deadeye Badge since it was an easy trinket to get that procs (or in this case On Use) Mastery. I then stacked to 20 on the PTR, activated the badge and took a screenshot:

https://www.dropbox.com/s/24efrulb12...413_174943.jpg

I then canceled the Mastery buff and took a second screenshot:

https://www.dropbox.com/s/110exa1zg4...413_174946.jpg

As you can see both screenshots shows the same value on the Tigerseye Brew buff.

[Deleted]

That's..... not how RPPM works. Yes, haste does increase the frequency of the proc, but only by the amount of haste you have. To realistically get to the average proc rate giving 100% uptime, you'd need 600% haste. Now, will our average level haste make it to 18% uptime instead of, say, 16%? Sure, but not any more than that.

correct me if im wrong. the rppm trinkets getting a higher proc chance every second, up to the point where it procs. thats what i read a few weeks ago.
more haste will just give a litte boost to that chance, because you auto attack more often and there is no cd.

[Totaltotemic]

correct me if im wrong. the rppm trinkets getting a higher proc chance every second, up to the point where it procs. thats what i read a few weeks ago.
more haste will just give a litte boost to that chance, because you auto attack more often and there is no cd.

Right, and the way the RPPM formula calculates the proc chance at any given time uses a time value of 1 second, but that 1 second is reduced by however much haste you have. 20% haste should result in 20% more (as in 30% to 36%, not 30% to 50%) uptime. The 600% haste number comes about because the trinket with its base RPPM only has a 16.7% uptime.

[Deleted]

From what I have been playing around with on the PTR stat priority would be something like:

Haste (5-6k) > Mastery > Crit - More haste will be required than on live as we will not be getting as many combo breakers. Also crit will become weaker meaning that we will only want to use FoF when low on energy, as we did during early MoP.

I would go with Bad Juju and Talisman of Bloodlust in terms of trinkets. If Talisman has a 100% uptime during combat and a negligible ramp up time, it allows us to cut a chunk of haste and gear more into Mastery/Crit. Hit and Exp won't be an issue and in most cases will be quite easy to cap from a few items.

[Requital]

It's really hard to be positive from the trinket stand point I could see different trinkets useful for different fights. As it sits Rune of Re-Origination would be crit for me not haste but next patch it would likely be Mastery over anything else. I could see a tank getting interesting use out of one DPS not so sure about I would take one to play with it but I don't feel like I would put priority on it.

Bad Juju - Seems to be a given to be one of the BiS trinkets at least from what I can tell w/o being able to really test it.

Talisman of Bloodlust - I don't think I would want to use this, For a Haste proc to be useful you would have to have 100% up time with no downtime. This reminds me of the ArP trinket DBW in Icc if you are pushing to a certain level of a stat then a trinket gives more of the that stat that isn't consistent you end up with too much or too little haste. That is where I would classify the Talisman. I think I would play with it but not put priority on it.

Renataki's Soul Charm - While the expertise is probably wasted good thing you can reforge a chunk of it away but it's not too bad at all.

Vicious Talisman of the Shado-Pan Assault - crappy uptime not a huge fan but the uptime is an indication of the buff size so it's not that terrible.

I'm likely to push for the Bad Juju and hang on to my Terror in the Mists as long as humanly possible. Hopefully by the time I have to replace it I have a better idea of where my stats will be and what will be better.

[Deleted]

After having done a little more testing, I'm going to change my opinion on which trinkets will be BiS.

Bad Juju & Bottle of Infinite Stars. Both have static mastery and agi procs. Infinite stars double upgraded is less than 300 agi short of heroic Bad Juju.

Talisman may be decent prior to getting your 2-set but as soon as you get your 2-set all the haste you will need will be close to the 4k mark. After having done a draft BiS list for 5.2 the kind of stats we will be reaching mastery wise will be verging on 16k with slightly more than 40% crit. I am very much looking forward to this tier considering how strong mastery is proving to be.

[Deleted]

Windwalker: 1.084

as far as I can see, we will get a relatively low coefficient.

I guess the damage will be increased by Tiger Stance, just like the Elemental Force enchant. And possibly Tigereye Brew too.

Additionnally, this gem could close the gap between 2H and dual wield, since RPPM proc chances do include the 40% attack speed bonus from Way of the Monk.

[Crevan]

If Talisman has a 100% uptime during combat and a negligible ramp up time

It will have nowhere near 100% uptime. With around 5.5k haste and 2.9 attacks per second, the average time between procs will be ~15 seconds. And with a proc duration of 10 seconds, you'd have to be very lucky to go above 1 stack often. The proc on the normal version of this trinket, with the same 5.5k haste and 2.9 a/s, averages to about 1450 static haste rating in a 10 minute fight (accounting for heroism and the haste from trinket proc itself).

[Deleted]

2.9 attacks per second

Why would the number of attacks per second matter ?

[Crevan]

Why would the number of attacks per second matter ?

Because of the way RPPM works. The formula for the proc probability on any given "attack" accounts for the time since "last chance to proc". With 2.9 attacks (or chances to proc) per second, the time since "last chance to proc" is 1/2.9 on average.

[Deleted]

Because of the way RPPM works

It's how it works, but remember why it works thus. The aim is to have the same number of procs regardless of the distribution of attacks over time.
Over a T time, is you have n attacks where t(i) is the time at which the attack number i occurs, then the average number of procs is :
sum[ (rppm/60) * (1+haste) * (t(i)-t(i-1)) , i=1..n] = (rppm/60) * (1+haste) * sum[ (t(i)-t(i-1)) , i=1..n] = (rppm/60) * (1+haste) * T
The number of procs does not depend on n, or on the t(i) values. It does not depend on attack speed.

[Crevan]

Good point. However, it does not change the conclusion: the uptime will be much less than 100%, and most of the time when the buff is up, it will be at 1 stack.