This is the document which just contains some background information and theory on how I developed the sim. It is a bit wordy, but it may help for those of you who want more info.
[B]Background info[/B]: Haste is 325 / 1%. Mastery is 350 / 3%. Crit is 350 / 1%. Versatility is 400 / 1%.
Keep in mind that all secondaries scale multiplicatively off of each other and so once you ‘normalize’ the secondary stat value of one secondary relative to another secondary, the optimal way is to then continue in a 1:1 stat point allocation between those two secondaries. This is the entire basis of what goes into the sim.
Regardless of what is optimal, the exception to this is Mastery, and people stack it very high as an insurance policy to prepare for the one time the raid might dip incredibly low. It doesn’t maximize HPS in the long run, but serves another role.
The main point of these secondary posts are to explain how I came up with the 1.2x values for haste/crit, the reasons why crit is a weaker stat from TW+HS/Floodwaters/Empowered Droplets, and the reasons why haste is actually not weakened by TW+HW/QA/Wavecrash.
[B] Haste [/B]
Tidal Waves, Queen Ascendant, Wavecrash [B] all scale multiplicatively and thus independently [/B] of Haste. This means that having any of these 3 abilities does NOT alter the its stat weight. It is a [B] false notion [/B] that having these 3 abilities makes Haste weaker. Just because these spells may give you fast casts, doesn't mean Haste is all of a sudden a weaker stat because [B] none of these abilities affect the GCD [/B]. If you cast a HS with 20% Queen Ascendant on beta (and 0 Haste), your HS cast time is 1.2 seconds. However, the GCD remains at 1.5 seconds. In essence, this would mean QA/Tidal Waves can at times result in no actual output gain because if you are casting faster than the GCD, you are going to be waiting around for a while until you can cast again. Thus, Haste does not automatically become a weak stat with QA, Tidal Waves, or with them combined.
Haste also happens to double dip on Riptide HoT/HR/HST.
From looking through Beta logs, the Riptide HoT component, HR, HST all comprise at minimum 20% of the output in raids. So this is where I came up with my 1.2x stat weight to haste. If you have these at a higher or lower level, you just have to adjust the modifier in the sim from 1.2 to whatever you want.
[B] Crit and mana regen: [/B]
Side note here before going into crit. I did not take into account mana gains through resurgence for the sim. Crit helps partially in the HPM department through resurgence, but in general, resurgence restores a very low amount of mana.
When you discount the bonus of Tidal Waves on HS, 10% Crit from Floodwaters, Refreshing Currents, and 5% baseline Crit (since all 4 of these would be present whether or not you stacked Crit), the gains are somewhere between pointless, and borderline useless. Sitting at 20% bonus Crit (before the 5% baseline, so 25% total), the resurgence gains attributed to this amount is about 250000-300000 mana in an 6-7 minute fight (based on LFR, normal mode, heroic mode, and mythic mode beta logs). This is about 25% mana.
This is an exorbitant amount of Crit to stack for a very mediocre gain. Remember, a Dreamless Sleep potion gives you about 22% mana. And your baseline regen gives you about 300% mana in this same timeframe. Not to mention, if you don't OoM, you didn’t need the mana in the first place. And lastly, if you simply optimized your output (and chose better spells) to do more healing with the correct secondary stats, you would actually have a higher gain in the long run.
It isn't always about regen, even when you are OoMing. That's the point. Even on beta I can still see shamans who do nothing but CH / HS spam for absurd levels on actual duration fights. You have to get used to using more efficient spells and cast when you actually need to. Having 10% extra mana from stacking something like 20% Crit isn't going to change this.
[B] Crit: [/B]
Unlike Haste which scales independently of all the ‘speed increase’ talents (TW + HW, QA, Wavecrash), all the passives / spells for Crit (Tidal Waves + healing surge, Floodwaters, Empowered Droplets, 5% baseline crit) all scale [B] additively [/B] with Crit.
Again, I was not able to include this completely in the sim (it is partially there) and so it overvalued crit just by a slight amount (about ½ a percent)
But still, for the part that I did incorporate, I estimated us as having 5% baseline crit as well 3% additional baseline crit (on average) from these 3 additive talents. This puts us at about [B] 8% baseline crit [/B]
Now for the positives of Crit - QA.
In the sim, I only considered 3/3 QA. The reason for this was that it isn’t feasible to retain 3 QA relics while also maintaining iLvL on a weapon (yes, I know EoA has a frost/life QA relic, but no one will use an 840 relic over an 860, regardless of what the trait is on the 840). A 15 ilvl boost on beta currently for 1 relic, raises your overall output by about ~1.5%. One level of QA is only worth about ~1.3-1.4% (and this is in the best type of circumstances, much less so if you don’t CH). This does not even consider the other trait on the higher iLvL relic which would make it pull even farther ahead.
Here is what QA actually buffs and doesn’t buff:
1) [B] Using QA in conjunction with HS ALWAYS results in a net 0 gain of HPS. [/B] This is true regardless of the amount of Haste you stack and it is simply because QA is not real Haste, does not lower the GCD, and so having a 1.2 second HS (from 20% QA) does not mean anything but being unable to cast for an additional 0.3 seconds
2) [B] Using QA in conjunction with HW + Tidal Waves ALMOST ALWAYS results in a net 0 gain of HPS. [/B] With 3/3 Tidal Chains, we get 39% faster HW casts under Tidal Waves. This puts our HW cast time at just a bit above 1.5 seconds. Once QA procs, you get the same problem as above in that you cannot cast faster than the GCD.
3) [B] Using QA alone (no Tidal Waves) with HW DOES result in an HPS gain. [/B] This should make sense. But remember that the only spells that can proc QA are Riptide, CH, HW, HS. However, both Riptide/CH already proc Tidal Waves and so they cannot be used to proc QA alone. The only way to get a ‘solo’ QA proc is through HW spam or HS spam. To estimate these both, I have been using an average of casts and healing through various normal / heroic / mythic logs. In these logs, 1/4 of the total casts (in a long, typical fight) come from HW. The resultant QA proc can be used to benefit either a HW or CH which together make up ~30% (on the high end) combined output on a given fight. As for the HS contribution, it makes up 2% of overall casts on a given fight. Same assumptions as above.
4) [B] Proc'ing QA with CH and using the subsequent QA on CH, DOES result in an HPS gain [/B]. This is obviously a big draw towards QA. Remember, QA does not increase the HPS of HS and because you are using CH to proc QA (which spawns Tidal Waves), it will never increase the HPS of HW (see 2). The assumptions I was making here was that you have High Tide and Floodwaters. Looking through logs, CH was about 1/8 of casts in a given fight (This is honestly excessive, but I am trying to put QA in its strongest light) and its output to be approximately 15% (which again is on the high end).
In the above 3) and 4), I made the assumption that EVERY single QA proc was used effectively meaning none were used on HS and none were used on a HW that also was under the effects of Tidal Waves. I also assumed you were able to USE every single QA proc since remember it is only a 5 second buff.
The end result of combining everything in 3+4 (and when you give QA a lot of assumptions and put it in it's basically strongest possible form) is that each point value of Crit is actually about 20% stronger. In other words, Crit is getting a 1.2x stronger weight in my sim. For me personally, it is quite a bit weaker (when wellspring) because I definitely do not have as much output as others from CH on beta. Still, 20% felt like a fair estimate based on the above.