I would also like to add my results for Bryntroll procs when using SoC instead of SoV since most of the test here were done with SoV.

My findings are as follows:
-Meleeing dummies with 2 dummies nearby(allowing SoC bouce) yielded a 3x increase in number of drain life procs(33% to proc drain life on melee hit)
-Standing in the middle of 4 dummies and spamming DS with SoC yield a 10% to proc drain life(No auto attack melee hits were made)

Given the 1st test results i would assume that drain life procs off SoC chain procs. SoC procs off the selected target would not proc drain life. otherwise a 44% proc rate should have been observed(Melee hit + 3x SoC procs would have given me a 4x proc if it did)

I am not sure what the 2nd test results would mean though. If the RV refresh can indeed apply drain life proc i would have expected a much higher number of drain life procs. However i would also like to add that the 2nd test was conducted with a little more than 30 DS and as such may not be accurate. The 1st test however was done with close to 400 melees and should be fairly accurate.

Given these results i would expect the priority system to change with bryntoll equipped. Without doing any math i would expected the new priority to be something like:

CS>DS>Conc>Judge>How>Exo

This would be due to the extremely high proc rate of drain life when casting CS due to SoC Proc chains.

I think he is saying that you don't need to be facing the mob to get it to proc since he got judgement to proc while facing away from the boss.

Retesting with more attacks (around 3000), I can confirm that the proc rate is indeed 11.3%, i ended up with about 11.2%. Not sure how I managed to screw up that it can proc facing away from something, this time I tried it I couldn't get it to proc 58 times in a row. (99.9% chance of proccing at 11.3% chance per hit.)
Basically you guys can ignore my previous post entirely and refer to this one.
And for any PvP enthusiasts out there, the proc cannot occur from SoR hits, but can from JoR hits.

Sorry for digging this back up. Been out of the loop a bit due to Xmas and all.

YEs, I did check the % difference between various rotations. The top 28 something were all within a few % of eachother. The worst ones were around 7% off from best. I unfortunately only have details of the top 10 left, I didn't save the entire log, but Jeh confirmed my findings, which isn't entirely strange, since we did several iterations on our individually developped models to confirm results

So no, I didn't get "massively" different results. The 'bad' conclusion of this is probably that any paladin with 2T10 just randomly mashing buttons without any care of anything will likely put out very similar damage to a paladin following a perfect rotation.

You may have misunderstood my point about CS priority. What I did find is that priotities with CS at the end seem to have more downtime. The lowest DPS rotations I came up with all had CS towards the end. With it's short CD, it's still the best "filler" in the rotation.

Yes, it tracks benefits of hit to melee cap, then to spellcap. After melee cap it's pretty low, but a non-zero upgrade. Even with Bryntroll the value of hit above melee cap is not worth specific gearing (or heaven forbid, gemming). Going over hit cap can still be a transitional upgrade until you can get your personal idea of BiS with hit/exp where you desire.

Sense Undead has a clear mouseover tooltip - it increases all damage by 1% when set to On - regardless of target type selected. Separating Undead/Demon solely for one glyph isn't worth the effort (HW eligible for both, auto-crit on Exo, etc, etc). I include it inside the CoE and Blood Frenzy modifiers, if on. CoE hits all spells, BF hits all physical. Yeah, it may mis-calculate if you don't have buffs selected. Or, just do what I do, leave it disabled. You can then multiply end result by 101% if desired. Glyph of Sense Undead will impact all gear equally, so it's a non-issue for comparison.

Not to mention 1% is going to vanish in RNG fluctuation in real data. Definitely use this Glyph (and make sure you have the correct tracking set!), but for modeling purposes it's pretty moot.

Next version will model in/out AW for Bryntroll - 5% increase, as you expected. I'll try to remember to upload tonight or tomorrow.

Quick question concerning Clcret and rotation. Wasn't sure which thread to pose this question, but I think this is the right one.

If you have your basic rotation set up in clcret and lets say DS pops up as the next ability with Judgment as the following (because DS comes off cooldown slightly quicker than judgment in this example), but my omniCC timers say both have 1 second left, should I wait the extra few fractions of a second to hit Judgment then DS or follow what clcret has to say?

I'm posing this question, because it seems that I have ALOT of downtime between abilities at some points in a fight (when following clcret to the t) and I know our rotation shouldn't have more than a second here and there.

There have been quite a few posts here (with the theory and math explained) about how delaying an ability to use a different ability is always a dps loss, even when it's a small delay.

In a perfect priority you will have 1 to 3 free GCD per minute (depending on DP usage). I believe you will find a good portion of what appear to be dead zones are not quite an entire free GCD - you have the 1.5 to complete the GCD of the ability you just completed and up to a second of free time before next ability is available - 2.5 seconds can feel an eternity when we're programmed to mash, mash, mash, however it's not an entire free GCD and using this time elsewhere will delay your next attack.

As mentioned, math regarding delay has been done at least twice - I believe first back in the 3.0 or 3.1 thread. In general, even casting Exorcism right now (our lowest priority) is superior to waiting 0.1 seconds and casting Judgement. This is often the difference between a bad Ret and a good Ret - a good Ret presses their buttons fast - doesn't even much matter what buttons, as long as it's rapidly after CD. The difference between good and great is procs, AW usage, movement, and the fine details (such as precision gearing).

Approximately 1 or 2 times in 100 pressing the immediate ability is a net loss due to harmonics (I'll again credit Neraya for first identifying and naming this situation). Basically what happens is by using ability X it causes a cascade through all the other cooldowns - Y is now 1.5 seconds later and Z is 3 seconds later. When they next come off cooldown it pushes other abilities later due to priorities, end result - less abilities used overall or more delays between abilities. There is no feasible way to identify this as it occurs during combat unless you're a math savant able to carry all your cooldowns in your head constantly and know the exact timeline for the next few minutes, or are otherwise precognitive. Lost DPS from this occurrence is relatively small and easily disappears into RNG fluctuation.

Bottom line - hit something. If A is off cooldown now and B is off in 0.1 seconds, either hit A now or hit B in 0.1 seconds - don't waste time thinking further and hit either 0.2 seconds from now. The benefit of clcret and why it helps make at least a good Ret is planning one ability ahead and being able to hit as shortly after a cooldown completes as possible - any completed cooldown (reference discussions above that most priority sequences are within ~5% total DPS of each other).

Righteous Vengeance jumps in its % of damage done position on the Twin Val'kyr encounter relative to a more standard tank and spank encounter like Deathbringer. I remember hearing that this occurs because of double dipping. What exactly is the mechanic behind the double dipping? The +100% damage buff first applies to the critical strike that determines the size of the RV ticks, and then the +100% damage buff applies a second time to the actual ticks?

Could this mean that 2P T9 / 2P 10 is superior to 4P T10 on an encounter like Blood Queen if you are one of the early vampires? For those who are not familiar with the encounter, when you become a vampire you gain a +100% damage buff that does not wear off.

Considering that some people are able to get their T10 2pc bonus today, how much of a DPS increase from additional Bryntroll procs can we expect from the (average) additional DS use? Further testing data about the bonus in general is also welcome.

Exemplar's spreadsheet has the most accurate model of Bryntroll out there.

With my current gear Bryntroll's proc should provide 752 DPS. If I simply flip on the 2P T10 switch without changing any other variable, then the proc should provide 806 DPS. So 54 DPS from the increase in Divine Storm usage and decrease of Consecration and Exorcism usage, or more likely a ~7% increase depending on an individual's level of hit, haste, and crit.

Gloves and shoulders so you can get 2piece asap, simply because they cost 60 badges each instead of 95 for the legs/head/chest.

There's no hard and fast set of order, no. The order in which you get the pieces is determined by the gear you have at the time. You don't want to break the hit cap (or pass the expertise cap) and you want to get the bonuses as fast as possible.

Yes, this is exactly it. RV double-dips from the "aura" effect (dark/light) AND the Empowered effect. With the right cooldowns, procs, and luck, people can get some extremely signficiant RV ticks (I've seen upwards of 40k).

That being said, and relating to your second point, this is when the stars align. While on progression (meaning you aren't likely to have full 277 gear) then I'd say that 2/2 is probably the best setup for this fight, especially if you have 258 pieces for your 2 pieces of T9. To avoid confusion, however, I want to stress this is only the case for this encounter (as I believe it is the only encounter with such a modifier) and for typical gearing we still want 4pc T10. The best pieces to swap for this probably would be helm, gloves, or legs, as these have the same stat distrubtion accross T9 and T10.

Does Rawr currently model the double dip effect of RV on +% all damage modifiers? Adding a "% Dmg=1" entry (+100% all damage) to an item seems to merely double the DPS contribution of 2P T9 rather than quadruple.

Likewise, Sanctified Retribution and Crusade/Sense Undead seem like they're not exhibiting a double dip effect on 2P T9's DPS contribution in Rawr, as they are providing a 3%, 6%, and 1% DPS increase respectively. Vengeance may fall into this category as well but I cannot test it.

That's the thing, this "double dip" effect doesn't apply to everything (and I believe Rawr is modeled correctly). It was explicitly removed from our previous talent/effect modifiers (Vengeance, crusade, etc) a while back (I can't remember exactly when, but 3.1 seems to ring a bell).

The Twins thing is more of a fluke than anything else, which actually has me wondering if the vampire effect will even have the same issue. I'm sure Blizzard is watching parses for the fight, so it's really a matter of how much they care (since ease of coding would dictate reusing the twins effect code).

I'll firmly agree with the general basis that 2+2 is a desired transition gearset until 4 T10 is obtainable - on Queen or otherwise. Regarding whether it will be 4T10 on Queen in specific...

I just did some napkin math involving my spreadsheet. I calculated how much DPS I would perform normal, then how much with every ability doubled (which is not the same as doubling bottom-line DPS, thank you RV double-dip), both with 2+2 and 4T10. Where X is % of fight normal and 1-X is % as vampire, I found the balance point (X) was 95% in my gear. Therefore if you are a vampire for 5% or more of the fight 2+2 was superior.

Again this is napkin math with my current gear. Results may vary. As long as you can guarantee (or average) 5%+ of that fight as a Vampire, the 2+2 appears superior. If you lose the RNG lottery (don't get picked), you will fall a bit short of what 4T10 could do (100-150DPS, probably). If you win the lottery (get at least 5% vamp time), you'll exceed 4T10 (by around 300ish DPS, I think). It's not huge either way.

2+2 should generally exceed, but not by an enormous margin. I'm not personally sanguine about carrying and "downgrading" to it for a specific fight. Others may choose to do so, especially if chaining external cooldowns to try to register on some best list.

Regarding 2 piece T10's impact to Bryntroll - I get from 5-8% depending on how DS resets (different seeds). Glutton's manipulation of the spreadsheet is correct (and I applaud his skill and facility), but I don't think he ran multiple seeds for multiple random samples of autoattacks resetting. This is with the Bryntroll AW tracking I implemented but haven't yet made available (forgot to upload last night).

Any class with a double-dipping ability (Deep Wounds, Ignite, Poisons, etc) could argue the same. I would actually argue Mages (Arcane, of course) may be superior. We scale well with % modifiers because we have several already and the chain can grow out of reason - Arcane Blast and Vampire status could do the same. If your mutilate Rogues are performing (i.e. topping meters) they may be second best. And so on.

I would say in general your best DPS (multiple) would logically be your best selections for Vampirism. Obviously this could relatively easily include one of us, as Ret can dish out significant damage. Any such argument should be made on a guild by guild basis. A good Beast Mastery Hunter (i.e. large % damage from pet, which would remain unbuffed by Vampirism) could be superior to a bad Ret, so I'd decline a blanket statement regarding Ret suitability. In general, we should not be last picked for dodgeball, definitely not.

Even if you're in the group selected as +4 (8 total active), you should be a Vampire for more than 5% of the fight, thus making mixed Tiers (2+2) potentially superior. Whether you go that route once you can use 4T10 is, again, personal choice. The boost from the mixed set isn't huge, while Vampirism overall is, naturally, enormous.

Mages are superior on Valkyrs and probably will be on Blood Queen as well, but that doesn't change the fact that we scale better with damage "boosts" than the other classes you listed.

Obviously your best DPS should get it if they're leaps and bounds beyond the rest of the raid, but if everybody is relatively close in performance then the case for giving it to a ret paladin has a lot more validity.

Mages are superior on the Valkyr predominantly because they have so many different cooldowns they can sync with the Empowered buff, as well as an extremely high dps/high dpm rotation they can use for a short duration of time, allowing them to output a simply staggering amount of damage for a very short period of time.

Since the buff on the Blood Queen will not be limited in duration (as well as the fight requiring a non-trivial amount of movement), it is unlikely that mages will benefit any more than any other class that doesn't have double-dipping mechanics.

Likewise part of our favour on Twins is cleave - DS and Consecrate are constantly hitting both. That DS can crit. Yes, it will be at 50% damage rather than 200%, but cleave does count.

More importantly - I have remembered and new version of spreadsheet (including updated Bryntroll modeling) is now available.

So I picked up the tier 10 tonight expecting it to just work on white melee hits. It definitely procs off pretty much anything. Here is a screenshot to show you a small sample.



Here is a recount showing with Seal of Corruption. It works the same way. Also notice how RV is second on damage. This bonus is pretty amazing.

SoC behaves differently than SoV. Each hit from a white hit cleave will act as a 40% chance to refresh DS. If you're consistently hitting three targets, then you have a 21.6% chance to *not* proc every white swing.