[TC] Project Windsong

[Akraen]

I'm going to embark on a final hurrah of calculamatrons before Warlords hits on Windsong. I'm curious if anyone has any working math or anything interesting/helpful to share before I figure this out.

I've seen some pretty nice results and uptimes in some very anecdotal and unscientific experimentation over the past week-- at least, some very high on-pull burst where having 1500 haste, mastery, and crit together (4500 secondary stats most definitely > 1650 intellect) really seemed to be nice, especially capturing +6 sec on each. The question is going to be whether or not the scattered RNG of permutations of combinations and non-combinations and probabilities of it being within a single or multi-target, or multi-target cleave (or even AoE) result in lower contributions than an evenly distributed intellect proc.

With snapshotting now being clarified as a non-issue, this strengthens the idea of a distributed gain for a proc like Windsong as you do not necessarily need it to be stacked so that it can be snapshotted, and only the cumulative gains of stacked buffs over non-stacked buffs becomes a point of question.

I'll be upfront, I'm not sure-- but I'm curious enough to say it's worth figuring out, especially as item levels get higher and boss fights get shorter. The shorter fights get, the greater certainty that a larger percentage of a given fight will be within pull conditions of all 3 secondaries being increased by 1500.


Funny and completely unrelated fact on the topic: You can proc Windsong outside combat, remove PBoI, re-equip PBoI before pull, and gain Windsong +145.5 amplified contribution to your Mastery and Haste. If you're min/maxing to that obscene level of hilarity, you deserve whatever kill you ever get.

[Methusula]

Ooh, and then there's double amp...just wish there was a way to proc the int on prismatic prison of pride.

[Dutchmagoz]

Your last part akraen, unequipping/equipping pboi before combat puts it on full cooldown, so you won't have it for 2 minutes. Just a headsup for people who may end up disappointed with their PBOI proc after testing that.

[Kver]

Not that you need to jinx it for PBOI proc to be awful.

[nathyiel]

Akraen is back !! Yeah !!!

My through is that Windsong can potentially be better than Jade. For this :
first, we need to take that secondary stat = intel. We will rectifying it later.
So Jade is 1650 and Windsong is 1500. So Windsong uptime should be 1.1 times that of Jade.
But Windsong can have an uptime bigger than 1 when jade is capped at 1.

So in short, if the sum of the uptime of 3 proc is bigger than 1.1, then Windsong is theorically better than Jade.

Now, if we take that the factor between intel/secondary stat is around 1.4. Then the ratio should be around 1.6.

Quick conclusion : if the average up-time of any of the 3 buff is bigger or equal to 0.55 (1.6/3=0.533) then Windsong is more interesting.

[Komma]

So Jade is 1650 and Windsong is 1500. So Windsong uptime should be 1.1 times that of Jade.

???

I don't understand this.

[nathyiel]

1500 * 1.1 = 1650
In short, if jade have 50% uptime, then Windsong need 55% uptime to be as good as Jade. (if intel = secondary stat)

[Komma]

1500 * 1.1 = 1650
In short, if jade have 50% uptime, then Windsong need 55% uptime to be as good as Jade. (if intel = secondary stat)

As you mentioned, this assumption doesn't make much sense. And that doesn't even being to describe the difficulty in evaluation; each secondary is of different value in different situations. Haste sending you below the GCD, especially if it's sending frostbolt below GCD, is worth less. Crit above shatter cap is worth less.

There's another factor that's often overlooked with procs - the fact that they favor fights without 100% uptime. Guilds that take down immerseus quickly, have downtime in between Galakras mob packs, opt to run away on Iron Juggernaut P2, or stop attacking during Nazgrim defensive stance will all benefit from RPPM procs that are powerful but have relatively low proc rate. In an imaginary boss fight where you only attack the boss for 10% of the time, the RPPM proc that has the largest effect wins regardless of uptime.

[nathyiel]

It's a TC discussion, not always realistic.
I have keep it to up-time because it can be compared easier. But if it's for fight with downtime and burst phase, the windsong win here because of what Akraen explain in the OP.

And haste below GCD isn't really a problem when bomb snap-shooting at huge haste is so much important

[Akraen]

We can just break the problem into parts and solve the uptime component first, listing our assumptions. It's perfectly ok to cut out one unknown at a time and allow the image to focus.

I suggest we solve uptimes -> burst haste windows -> other procs additions/deductions -> amplification -> value of each permutation's effect on all subcategories, recognizing that uptimes are the easiest point to eliminate the idea.

If we work through all those stages without a loss to Windsong, then I think the process validates the uptime argument within enough reason to establish confidence in the theory.

[Akraen]

Your last part akraen, unequipping/equipping pboi before combat puts it on full cooldown, so you won't have it for 2 minutes. Just a headsup for people who may end up disappointed with their PBOI proc after testing that.

Oh yeah that's why it's just a fun little fact, not something worth doing-- but it'd be funny if someone thought it was.

[Talontoe]

Your last part akraen, unequipping/equipping pboi before combat puts it on full cooldown, so you won't have it for 2 minutes. Just a headsup for people who may end up disappointed with their PBOI proc after testing that.

Oh yeah that's why it's just a fun little fact, not something worth doing-- but it'd be funny if someone thought it was.

Your evil
He giveth with one hand and taketh with the other

[Mageski]

Ah memories of the old broken T9 set bonus that let you gain about 1.5% more Crit by buffing molten armor with your 2p on and then switching back to your T10 gear and effectively getting 3 set bonuses.

[Akraen]

I saw some amazing results in my raid last night with Windsong. Shattered all my prior records, but that's mostly due to my guild improving so I won't just credit it to Windsong.

Uptimes fluctuated a lot. Pull was mostly reliable (more reliable than PBoI), and I found like magic they proc'd together when I wanted them to for Alter Time.

My gut is saying we should all be using Windsong. Still wrapping my head around a good method to prove it.


http://www.warcraftlogs.com/reports/CMp ... amage-done" target="_blank
http://www.warcraftlogs.com/reports/CMp ... amage-done" target="_blank
http://www.warcraftlogs.com/reports/CMp ... e&fight=29" target="_blank

[nathyiel]

What is saying SimC in these case ?

[Trelane]

Simcraft showed Jade Spirit as 1.0% higher overall DPS than Windsong in my Frost gear. Both have the same proc rates (2.2 rppm according to Simcraft tables). Simulated uptimes were:

Jade Spirit: 42.7%
Windsong Haste/Crit/Mastery: 16.3% (x3 = 48.9%)

Presumably Windsong gets higher total uptime because overlapping procs doesn't hurt it nearly as much as Jade Spirit. I can't see how it could be superior though, since secondary stats would need to have comparable value to Int.

[Charmander!]

While my guild's kill times are considerably faster than yours, unless int is valued equal or very close to a secondary stat, jade spirit appears better at a glance. My uptime on jade spirit on every boss (excluding immerseus, galakras and spoils) is in the 48-60% range, your cumulative uptime on windsong seldom exceeded my jade spirit uptime on a fight to fight basis (as a percentage), not counting fights with major downtime like galakras or spoils. While 2 secondaries are worth more than 1 int, the gap between them is not wide enough to warrant using a non-jade spirit enchant.

I thought about using windsong months back but dismissed it then as its uptime on the procs I would want (mastery, crit) wasn't high enough to justify it as the excess haste from a haste proc went to waste if meta was up at the same time, or during bloodlust where I'd much rather have an int proc.

In addition, int affects everything I would be casting during the proc. Every cast during a proc benefits from it - getting a crit proc alongside other trinket procs during orb where I'm spamming lances, or getting a haste proc during meta, or a mastery proc during the aforementioned lance spam, would be a loss as not all of my casts are receiving a benefit from the secondary stat, whereas all of them would be benefitting from raw int.

[Akraen]

Yeah due to uptimes I'm sliding more in favor of what Simcraft is suggesting. Edge to Jade Spirit still, but not by much. I could see RNG swings either way on this one but a majority of the time in JS's favor. My Malkorok parse is a good example of why-- absurdly low uptimes and they didn't proc all together.

I think Windsong is more "fun" though - at least from a proc-watcher's POV.

I am curious now if Arcane and Fire would ever consider them. Probably not Fire, but Arcane's secondary scaling is pretty nice. More mana regen from the haste, more mana adept, and the crit certainly isn't a bad thing.

[nathyiel]

If the number was reversed, it could have make Windsong all lot more on part with Jade Spirit.
I speak of if Windsong was 1650 stat and Jade was 1500 int.

In early level, Windsong is a lot better because stat scaling is a lot more interesting than in end gear.
It could be a good option for leveling to 100 with all the scroll of that can be craft just for 5 or 10 early level in enchant.