These stats support this argument if you notice the crit vs noncrit JoB through them.
Wow Web Stats
I am getting long chains on non-judgements by these webstats and was experiencing a lot of mana difficulty. If you notice the only non crit JoBs recorded were the initial one and one towards the end of the report that is well over 24 seconds from the previous critical JoB. There are also some larger areas of non-judgement that may have been non-crit JoB which did not land.

If the bug is 100% reproducable, doesnt that warrant a GM ticket and/or Official WoWforum post? This would also seem to fall in line with alot of the anecdotal evidence, my only concern is, if swapping judgements didnt solve the failure problem, as some are reporting, wouldnt that mean there is a deeper problem besides an initial crit lockout debuff?

Ill add my own anecdotal evidence to the fray, tonite in heroics our tank was rather squishy, so i was HoJ every cooldown to make life a little easier, but i kept missing judgements... even more than normal. If HoJ + Crit judgement is to blame, my increased usage and increased experience would seem to support the theory, yet, how exactly would you approach this Vs. a raid boss?

Merovengian, the bug is more or less reproducible and has been acknowledged by Blizzard as an existing issue that they're working on.

When I say 'more or less', we've all had it happen to us, but the lack of a combat log entry or any other screenshot-able material makes it difficult to externally illustrate short of frapsing the event wherein you cast Judgement, lose the mana and incur the cooldown, but nothing happens.

That being said, Blizzard does know that it exists and made light of it in a recent blue post.

So, new Patchnotes are out, i'm especially concerend about:
# Judgement of Wisdom: Now returns a percentage of base mana instead of a percentage of max mana.

Is this still 1%?
Sounds like a hard nerf for me again, first they say Mana Drain gets normalized so its not so hard for low mana Classes to get mana drained, and then they nerf Judgement of Wisdom again, so everyone gets less mana back :/

This was mentioned by blue weeks ago. It's meant to be a buff to small-mana-pool classes such as Enhance Shaman, Ret Paladins. 1% of max mana is basically 1% of base mana for us, while it's a whomping amount more for large mana pool casters.

Instead they've taken some number (who knows what) representative of what they want Wis to return and used that to figure the math in reverse.

Present Example:
For instance if we had (just to make flat numbers - not realistic mana pool sizes) 4k mana we would get back 40 (1%) on Wis right now. Caster with 15k mana would get back 150 (1%) on Wis. This disparity grows as the caster gets more Int (mana) on gear.

Future Example:
In future we and caster both have 4k base mana. If they want us both to get 150 mana then we'd get 3.75% on a proc. Caster sees no difference (even when their gear improves), we get a boost.

It's the inverse of the mana drain - both are buffs to small mana pool classes like Ret.

When the JoW change was first announced I made some calculations on our guild forums about various things and my expactations. I'll quote some of it here for possible basis of discussion:

From the same material I made some calculations how this would affect JoW gains for various classes if the amount would settle at 2% base mana:

I don't know about the rest of you, but i've found JoW to be so little of my overall mana regen that I don't even worry about it being up on a boss at all times. I judge light and all the healadins (we seriously have 3 lol) judge wisdom whenever they want to.

I haven't noticed a serious change in mana regen since we started doing it this way.

Just tested out that theory and I can also confirm it works 100%. Looked through some WWS reports and also seeing the same thing, a streak of nothing but critical judgements from me, but definately not every 7-8 seconds.

This bug didn't pop up till Righteous Vengeance was implemented right? There was another issue with that talent that the DoT is based off damage after mitigation abilities, ie. a crit judgement for 2500 where 2400 was absorbed by a Priest shield, would only tick like the crit was for 100 damage. I posted a bug about it a while ago, but I haven't looked to see if the bug still exists.


Looks like Blizzard's hot fix to nerf us at 70 weren't exactly well coded. :/

yeah sure i know what you want to say but my problem right now is that there just stands "a percentage" not an exact number..
hopefully ptr will turn out something good

Hey guys, nice work with the thread(s). As explained to those who came looking for me, I've not completely vanished, I've just not had much time for the game or the forums this past month.

Speaking of which, I've gone over what's been posted so far and there are a few suggestions I have for the OP (some corrections and some nitpicks):

-Rotation: I think the description as to why we chose what we did is somewhat vague and doesn't really do justice to the logic/discussions behind it. We didn't choose FCFS due to "fuzzy lines of what to prioritize", but for very specific reasons attained through logic and later proven through modeling.

There are two elements to our "rotation": FCFS and clash resolution

FCFS: Was chosen very simply due to the fact that due to the cooldowns our abilities have, you can't repeat any reliable short term pattern (for example: "1" "1" "1" "3", repeat) that would maximize DPS. The only patterns that repeat are several minutes long (containing several dozen of casts) and therefore not realistic to offer as a "rotation". Therefore FCFS is the only way to go.

Clash resolution: It was argued back and forth what the optimum way to resolve a clash of two abilities coming out of cooldown at the same time, the result was "highest DPS first": Judgement > CS > DS > Consecration > Exorcism > Holy Wrath


-Expertise cap: 26 Expertise is 213.13493 rating ~= 214 expertise rating (not 212). I guess it's worth mentioning that it is assumed that the cap is slightly lower than 6.5%, but higher than 6.25%. However still 6.5% (=214 rating) is regarded as a safe bet, so I'd note the 214 rating number.


-Attack Power: The OP states: "AP is somewhere in the middle of the pack as far as stats go". I understand this is based on the assumption of weighing stats 1:1, however I do think this is very misleading, specifically to novice players looking at AP on their gear. Stat price should be accounted for, which makes AP the 2nd best stat after Str since it always comes at a 2:1 ratio compared to other stats. You can add a note about stat cost for clarification.

-Spellpower: The OP states: "No longer our worst stat, but still pretty bad anyways." What is?

-"In addition, damage seals are no longer normalized, meaning we desire slow weapons once again." Seals never were normalized (unless there's something I'm missing about pre-wotlk SoB).

-Seal of Command: I think it's worthwhile to list one of the changes that it doesn't do double damage on stunned targets anymore, but gains 100% crit instead.

-Consecration Glyph: "Is worth roughly 50 dps" and later stated as "worth ~50-100 DPS if you use consecration as much as you should". Not sure where these number comes from, if it's an assumption of reduced GCD clashes being worth 50-100 DPS then this is anecdotal at best and should not be presented, not even as a ballpark range.

The main benefit (which is not mentioned) is it's possibly the best mana saving Glyph we have, assuming you're spamming Consecration (as you should) and have 5/5 Benediction, it's worth 21.725 mana saved per second or 108.625 mp5.

The side benefit is that it frees up some GCD clashes.

-Hammer of Wrath: I'd add a note about Sanctified Wrath being one of the main reasons why it's so good now, giving HoW an added 50% crit rate, resulting in average 80-90% crit rate on HoW.

-Holy Wrath: If we're listing changes, add a note about it having reduced range, but also reduced cooldown (halved) and acting as an AoE stun now too. It's a fairly powerful "CC" in instances now.

-Glyph of Flash of Light: OP states: "The HoT effect portion of this talent also overwrites the HoT from Glyph of Flash of Light, making that glyph utterly worthless to ret." This is simply not correct, the HoTs don't overwrite each other. The reason this Glyph is almost pointless for ret is that it halves the initial heal which the Sheath HoTs are based on, making them heal for half of what they would. This means it's a slight boost to non crit FoLs and a no gain/to loss to crit FoLs with Sheath spec.

-Yellow Sockets: Not sure pure hit gems should be included here. I would suggest etched (str/hit) gems till hit cap and inscribed (str/crit) afterwards.

-Blue Sockets: Again, no point in listing hit/stam gems. For blue sockets if you're going for socket bonus/meta Sovereign (str/stam) gems should always be used. I understand some might argue that the hit cap serves as more than just DPS, but you should be able to get decent enough hit elsewhere.

-Glove enchants: I'd like to suggest the old 15 STR enchant as one of the "cheap" enchants. It's almost exactly the same worth as the 20 hit rating enchant. Additionally you should strike the 15 expertise enchant as it's less than half the worth of the other enchants listed and would be a really poor choice.

-Bracer enchant: Again, the expertise enchant is incredibly weak and a lot more expensive than "Striking". No point in listing it.

-Boots enchant: I think the enchants section needs some ranking. It should be stated that despite Greater Assault being the higher level more expensive enchant, Icewalker is better until hit capped.

-Consumables: Listing more is not always better. Besides the flasks, I'd list the best elixir combo (Str/Mageblood, assuming the changes to JoW will make Int from Mighty thoughts less lucerative) and call it a day, since most people are looking for what's best. Or alternatively rank the consumables.

-Food: "It's not clear to me yet which is the best food": Dragonfin >= Fish Feast > Rest. The first two are roughly equivalent.

-Tier gear: In general, I'd avoid any statements converting "% damage increase" into DPS numbers, even as an average. The reason being is that in most cases when you're trying to compare gear that is very close, the differences are within the +/- 30 DPS range, many times in single digits, in which case the averages given can be very misleading. (If you absolutely must, provide a reference base DPS/gear level).


Anyway, I hope this doesn't come off as criticism, but suggestions for improvement. Thanks to Arikah for the work put in so far

Undocumented changes in 3.0.8:

Paladin
Skills
Protection

Divine Protection The penalty has been removed.
Divine Shield The penalty has been changed so that all damage done is reduced by 50% in place of an attack speed penalty.
Hand of Protection now Cannot be targeted on players who have used Avenging Wrath within the last 30 sec sec.
Avenging Wrath now Cannot be used within 30 sec sec. of being the target of Divine Shield, Divine Protection, or Hand of Protection.

Retribution

Judgement of Wisdom now give each attack a chance to restore 2% of the attacker's base mana. (Old - 1% of the attacker's maximum mana)
Judgement of Light now also procs from ranged attacks.
Seal of Blood now deals 27% of normal weapon damage. (up from 22%)
Seal of the Martyr now deals 27% of normal weapon damage. (up from 22%)

Talents
Retribution

Seal of Command now Gives the Paladin a chance to deal {0.45*Min Weap Damage+0.45*0.23*SpellPower} to {0.45*Max Weap Damage+0.45*0.23*SpellPower} additional Holy damage. (Old - Didn't scale with Spell Power)
Vengeance now stacks up to 5 times. (Up from 3)

Holy

Judgements of the Pure now Increases the damage done by your Seal and Judgement spells by 5/10/15/20/25%, and your Judgement spells increase your casting and melee haste by 3/6/9/12/15% for 1 min.

Protection

Sacred Duty no longer reduces the attack speed penalty on Divine Shield and Divine Protection. Rank 2 now increases total stamina by 8%. (Up from 6%)


Some nice retri buffs there, 6% more damage from Vengeance and they rebuffed martyr some

Clarified a couple things (thanks avitus), added patchnote subsection.

The SoB/SoM are not just tooltip updates; SoB was originally 33%, then 25%, then hotfixed to 22% - it is a straight up buff. The vengeance change actually makes me laugh a little, that's what, the 3rd time they have changed and then later reverted changes to vengeance? I'll take a 6% damage boost happily though... What interests me most (besides the obvious judgement bug fix) is the change to JoW... it looks to be a buff for our mana regen (and protection paladins too), and a nerf to everyone else - napkin math says JoW will be worth 150mp5 now.

Patch excitement aside, can we please get a few more large data samples on judgement hit cap (7% vs 8% vs 9%) and accurate data sampling on berserking? They're really irking me :|

This is the only thing I disagree with. If you have two clashing cooldowns, you should not be picking the "highest DPS first", but rather the highest "DPC" (damage per cast) first, since the opportunity cost of each ability, ignoring mana concerns, is the same 1.5s GCD.

Can you give an example where you feel going for DPC changes things over going for highest DPS?

The cast time (cost) is the same, as you say, one global cooldown, so it seems the only difference is which will do the most DPS using that global.

Losing 1.5 seconds of DPS time for a higher DPS ability has to equate to more overall DPS lost than losing 1.5 seconds of DPS time for a lower DPS ability, it would seem.

Its really about getting more casts of the higher DPS ability in over a sustained period of time. Avitus does a pretty good job of breaking it down here.

Agreed!

A single 10minute combat log on a test dummy would satisfy me regarding Berserking. We don't need perfect numbers, but having evidence (as opposed to anecdotes - no one has posted recount/log data) it chain-procs and how frequently it seems to occur lets us throw out a ballpark figure. Is it worse than Massacre (1 PPM or 100AP - unlikely), is it marginally better than massacre (45 sec internal cooldown, figure 5 sec to proc or 120 AP), or is it insanely better (no cooldown, can chainproc, high proc rate ~200-300AP)?

We know Judgement is special (uses ranged attack mechanics, rather than melee or spell). Having a test of even just 1000 (a small sample, but larger than most test performed) could again give us a starting point. I suspect either A) matches your hit cap (5, 6, 7, 8, 9, whatever it may be) or B) matches old 9% cap regardless. Someone with a 5% cap running a couple hundred 0 hit gear judgements could get us easy data (since the variance is huge between 5 and 9). If melee and Judgement don't line up then we know there's a difference.

It depends somewhat on the situation. For example, Divine Storm would be a better choice over Judgement if, say, one is fighting four mobs and has the [Venture Co. Libram of Retribution] equipped.

If that is what Karakas was referring to with his DPC preference, then I apologize for questioning it. I obviously and completely agree that what spell generates the most DPS is situational and changes depending on the nature of the fight or state of the fight.

I did a 10 minute test a few pages back on a friends warrior. The Retribution Paladin Thread (Wrath/3.0) No combat log but it shows the times and number of procs I saw. Certainly seemed to get RNG'd a bit at the end though.

I'll do another test this afternoon when I can and post the results again here.

The DPS cycle argument is based around single mobs fights, it has already been mentioned in the OP to prioritize DS and Consecrate for mult-mob fights.

The problem most people are mentioning is that if you hit CS after JoB every time, you end up having CS basically being on an 8 second CD for most of the fight due to GCD and judgement CD. Has anyone tested putting a DS in before the CS in the priority judgement? With two peice T7 the dps loss would be reduced and it would take CS out of sync with the judgement for a short time again.

If no one has I will try to get some testing of the three different dps cycles done.

I think it was just a semantics issue, since DPS for Crusader Strike can either be calculated as damage / cooldown or damage / time lost on other attacks (i.e. GCD).

Ok well I just did two 10 minute tests and the difference between the two tests was huge in the number of procs... 15 on the first test compared to 29 on the second (I kept going an extra 10 seconds on the second test and got 3 more procs but these were not included). This was using roughly the same rotation on a TG Warrior of WW, BT and instant cast slams. As I said in my most before the warrior only has Berserker enchanted on the MH. I'm very interested to see what would happen with dual enchants.

Clearly even on the low end of 15 procs in 10 minutes that's quite a bit better than massacre and 29 just completely destroys it. The enchant repeatedly renewed the buff and would sometimes keep the buff up for upwards of 40 seconds at a time. Now I know these tests aren't as rigorous as most would want them to be and my rotation wasn't perfect most likely but for me this completely seals which enchant my Betrayer of Humanity will be getting once I pick one up.

Edited title for spelling.