The only crit on proc trinket is UVLS so I don't see why crit being snapshot would change anything.

It's over 2500 Int for 1% crit, although reduced slightly for % Int buffs.

Yes, you are completely right, don't know where I got that 1200 number from.
Still, even a normal mode Cha-Ye's would provide around 3% crit, but I doubt it would change anything in regards to the rotation.

With today's trinket announcements, raid trinket proc rates going up, in no fewer than four different ways: Real Proc Per Minute (RPPM) Trinket Changes - Forums - World of Warcraft

1) All of the Int trinkets from ToT are getting a 5% increase to proc rate.
2) UVLS is getting a 30% increase after the beta nerf, so it will be about 2/3 of what it was initially. Guessing BIS now if you assume perfect use of procs. Also, point (4) below should help UVLS more than other trinkets (edit: maybe not, see below).
3) All RPPM trinkets are getting a new mechanism to "correct" for bad luck (which in reality is just an increase to the long-term proc rate, since you still get all the upside of good luck but some of the downside of bad luck is canceled out). My best estimate is that this amounts to a net 13% increase to the long-term proc rate of all RPPM trinkets.
I did my computation here: https://twitter.com/HamletEJ/status/311953149613195264 . We can see what sims and things wind up saying once people do them.
4) Because, according the new system, you get the non-proc bonus at the start of a fight, we're back to having a very quick first proc in any encounter, followed by normal random procs after that. Should make the opening timer sequence more like it used to be in the past.


Will have to think more about (4) in the context of UVLS. Its proc rate is so low that I'm not sure the early proc will be guaranteed. At 0.325 RPPM now, let's call it 0.5 with haste (generous), 1.5 times the mean proc time would still be 3 minutes. So if you've been fighting within the past 3 minutes you might not get a bonus at all. If, on the other hand, you've been out of combat (or not wearing the trinket?) for the past 10 minutes, you'll be at 11.5 times the usual proc rate, which is fast enough (expected proc 10 seconds) that you can probably hold your CA/Inc for the first proc. This is kind of an awkward result, I'll say more after thinking about it some.

I had a chat with some of my moonkin friends during the Megaera fight last night and they brought up a point that I thought id bring here.

If you don't know the fight, basically what happens is that Megaeras heads heal every time you kill 1 head, there are always 2 heads up.

I usually dot both heads to get more starsurge procks.
Doing the math, it should look something like this:
(I am at roughly 30% crit)
2 global cooldowns to add the dots = 1.11 * wrath(1.8 sec cast time) or 0.77 * starfire(2.6 sec cast).
so the dots would have to proc starsurges to at least make up for such attacks.
During my figths on the fight, my starfires averagely hit for 112614.7 and crit for 215268.9 making the formula:
0.77*((112614.7*70)/100 + (215268.9 * 30)/100) = 110426.43 damage
and my wraths hit for 61967.2 and crit for 134032.3 making the formula:
1.11*((61967.2*70)/100 + (134032.3*30)/100) = 92781.27 damage

with 2 dots up and 12 tics each, with a 30% proc chance and a 30% crit, the value for the dots is:
2*(12*(0.3*0.3*((131775.9*70)/100+(242029.3*30)/100-(1/2)*110426.43)))=236819.60

where 1/2 is 1 global cooldown compared to the 2 globals that I calculated before and 110426 is the average damage of starfire as calculated before. 131775.9 is the average hit of starsurge and 242029.3 is the average crit.

and compared to wrath it is:
2*(12*(0.3*0.3*((131775.9*70)/100+(242029.3*30)/100-(1/2)*92781.27)))=255876.375

236819.60-110426.43=126393.17 <-- During Lunar
255876.375-92761.27=163115.105 <-- During Solar

This is of course if you get every single starsurge inn so the gain might be slightly less. It is probably closer to 100k during lunar and 150k during solar.


If you see an obvious flaw. please let me know.

Keep in mind that I have NOT included the dot damage since it is will be healed anyhow.

It's a lot more complicated than that. You have to add in that the 2 gcds also increase the time in between eclipses, and the different eclipse energi gain you get from casting that extra starsurge. If you use Wrathcalcs you should get a better idea if it would be worth it or not. something along the lines of adding 4 gcds per eclipse cycle and changing the proc chance of shooting stars from 30% to 60%.

Either way you can always dot the other head during movement since the extra starsurge proc chance would beat the 30k instant damage from a SnF/MF.

On the other hand, getting additional Starsurge procs reduces time between Eclipses, so you also need to take that into account. At some point proc overlap is also a relevant issue, as mentioned by stuppp.

How do you guys feel about the [Volatile Talisman of the Shado-Pan Assault]?

So far I've been using [Relic of Yu'lon], and have been very satisfied with it's performance, however I've noticed that the balance druids in the top guilds (well, at least some of them) are using the volatile talisman. Now the thing is, they've already switched off their RoY for the heroic EoT, and now they've switched hc EoT for the volatile talisman. So my point is that from what I've seen on armory, it should be like this:

[Volatile Talisman of the Shado-Pan Assault] > Heroic [Essence of Terror] > [Relic of Yu'lon]

Also, I've checked simcraft, and it suggests a ~2500 dps increase for the talisman over the RoY. It's all nice and everything, but I bought the talisman to check it out at our pre-raid LFR run, and I wasn't really impressed. I kinda felt like I was doing less burst and less overall dmg as well.

This morning I thought I'd give it another try, and went for a dummy. I've did 2 6-min tests (pop cooldowns 2 times, and stop when they come up for the 3rd), and 1 9-min test (same deal, +1 cd round) with the RoY and the same with the talisman. The two were quite the same in average (87,567 with RoY, and 87,203 with the talisman).

Now of course I understand that my sample rate was very low, but still it confirmed what I felt in the LFR tests. Two things that I've found quite interesting/unexpected:
- even tho the 2 trinkets produced nearly the same results, it was the RoY that ended up ahead
- the talisman ending up second suggests that the priority mentioned above is wrong, either by just the talisman being the last on the list, or the yulon being the ahead both of them

And again, I didn't forget the fact that my sample rate was very low, and the fact that I didn't have the same buffs, that I would have had in a regular raid environment. Actually, that's exactly why I'm asking your opinion (moar samplz!). I don't really understand why some of the top druids are using the trinket over a 2/2 heroic Essence of Terror, as (to me) it seems even worse than the Relic of Yu'lon.

It should be pretty clear that VTotSPA is better than EoT, at least if you discount GCD-cap and haste breakpoints. It has more Int, and more average haste (bigger proc, only half as long, but more than twice as often).

Optimistically, RoY has about 400 more average Int than the talisman, but the talisman averages something over 1700 more Haste. Haste loses a lot of its appeal once it puts your instants at the GCD cap, or when you cycle so quickly that you never leave NG. Still, 4 Haste usually looks better than 1 Int on paper.

Large, frequent, haste procs do have some play-style issues. They make timing casts, (and decisions about how far to run during a MF GCD) something you have to pay more attention to. If you already have plenty to worry about, you may not get the most use out of the haste proc.

I took a look into the effects of the new long-interval proc rate bonus in order to convince myself of Hamlet's math. The results of my sim are consistent with Hamlet's derivation of a 13.1% increase in overall proc frequency.

On a character with 20% haste, I modeled 0.5 rppm trinkets with and without the new scaling proc rate after long proc-free intervals. I checked for a proc every second and tracked the time between procs for each trinket over about 11,000 hours:




I repeated this simulation 1000 times and the increase in proc frequency converged to 12.95%. The new system is quite effective in eliminating extremely long proc-free intervals. The long tail of the exponential distribution is pushed to the left, allowing for shorter proc-free intervals and more procs.

There does seem to be some parameter-based variation in how much the proc frequency increases: lower haste and rppm values gave slightly smaller increases (maybe due to an artifact my implementation?), but I couldn't find any reasonable values that resulted in increases outside of 11% to 13%.

MATLAB/Octave code is available on request if anyone wants to play around modeling a specific trinket with specific haste values.

FoN changed in new PTR data. Now summons one Treant at a time, and has 3 charges: 5.3 PTR: Build 16767 - Spell Changes and More - Wowhead News

Only question is whether you'll have to use 3 times as many GCD's or they'll change that somehow. Should be live on PTR in a little bit.

Also a new glyph for Moonkin utility spells granting energy. Should be handy pretty much anytime you're using them, since we don't have other important glyphs anyway.