On the first page of the spreadsheet under Delay. As others said you need to divide your JoB hits by 2 due to self damage being counted and DS cannot be calculated if it's hitting a random amount of multiple targets during a fight.

A lot of this is common sense or really easy to find (the cooldowns are smack at the center of the first page of the spreadsheet, looking at it for 10 seconds you wouldn't have to ask this).

Note: Since you seem to be new here, I'll point you to the forum rules before a mod comes knocking. You don't have to quote every reply by quoting a massive post if you're only replying with one line, neither do you have to answer "every" post directed at you if your answer is the equivalent to "yes I'll try that".
Go try it first, then come back with some info to reduce the thread spam

This is using the First Come First Serve rotation modeling in Rawr.Ret. It has all new 3.1 mechanics modeled, including the new armor mechanics and the changes in yesterday's patch. The situation Zurm was talking about was with full raid buffs, same glyphs as you, and 4pc T7. With BIS 3.0 gear, and if CS is used every 6 seconds I compute it does 825 dps, counting Right Ven but not seal proc. With Judgement being used every 7 seconds I am getting it does 1,237 dps, counting Right Ven.

With a priority cycle of Judge > HoW > CS > DS > Cons > Exo I get the effective CDs of:

• Judge: 7.37
• HoW: 7.29
• CS: 7.37
• DS: 11.48
• Cons: 11.48
• Exo: 16.29

With a priority cycle of CS > HoW > Judge > DS > Cons > Exo I get the effective CDs of:

• Judge: 7.61
• HoW: 6
• CS: 6.46
• DS: 11.20
• Cons: 11.22
• Exo: 16.71

Effective cooldowns are computer by (last_use - first_use) / (total_uses - 1). Note this is with no latency factored in; being able to hit the ability the instant it comes up. That is something I am going to add to the rotation simulator.

So as you can see, you are giving up almost a second off of CS/HoW just to get .3 sec off of Judgement. While this was worth it in 3.0, with all nerfs Judgement has gotten and the buffs that CS got it no longer is.

On that note, the update of Rawr for 3.1 and a general redesign is coming along very well. I think I have all of the bugs worked out now, and should be ready for general use by the time 3.1 hits live. You can test the beta on Codeplex until then.

Ermad - I suspect the fact that you model HoW in cooldowns for the entire fight while I do not model HoW in cooldowns at all causes the discrepancy. I'm also using (last_use - first_use) / (total_uses - 1) with a timeline of 6 hours to minimize deviation.

Both models are technically incorrect, since the ideal is one priority without HoW (mine) and one under 20% with HoW (yours). So based on a rough glance I'd say that CS becomes more important than Judgement under 20%.

I'm also finding that Judge > DS > Cons > Exo > CS gives better cooldowns and DPS than either rotation you listed.

• Judge: 7.50
• DS: 10.00
• Cons: 10.01
• Exo: 15.01
• CS: 7.50

I only simulate HoW for the last 20% of the fight. I have a variable called abilityNext (ie: howNext or judgeNext), that keeps track of when the ability comes off CD. For HoW I just set it to fightLength * (1 - timeUnder20) initially instead of 0, so it only effects rotation for the last 20%. This implementation isn't clear to see because of how effective cooldown is calculated, but it does only simulates HoW for the correct amount of time.

Note: timeUnder20 is .18 in my calculations.

edit: I get Judge > HoW > DS > Cons > Exo > CS being 20dps lower then CS > HoW > Judge > DS > Cons > Exo.

Interesting discussion, no doubt, and it also spawns another question in my head: would it be benificial to your overall DPS to use one set of priorities during the 100>20% portion of the fight and another during the sub-20% portion?

I'm currently playing like this, though I'm not entirely sure about the benefits (for the most part the sub-20% portion of the current fights are too short to make any judgements by, and I'm not proficient enough in the tools to do much real work other than manual modelling)

Either, A) with the gear I was using the Judgement rotation is superior and the gear Ermad uses swung towards CS. I have found that gear can definitely affect a preferred rotation. An excellent example is to throw in crap gear and a great weapon - suddenly CS and DS are highly valued because they also scale with weapon DPS, while items like Judgement and Exorcism do not. Great gear, crap weapon, and items not based on weapon DPS jump.

Or, B) above 20% the Judgement rotation is superior (Ermad could try setting under 20% time to 0 to check) and below 20% CS is so far superior that the small portion within that window pushes 20 DPS ahead of the straight Judgement rotation. If this is true I'd think Judgment above 20% and CS below 20% would be greater than 20 DPS gain (also checkable by setting time under 20% to entire duration).

Or, C) Ermad or I (or both) have math errors.

Most likely a mixture of all three.

I think in 3.1 I'll try the rotation my spreadsheet spits out on a Patchwerk-like situation, then try the rotation Ermad suggested - see which gets me personally cleaner effective cooldowns. I would recommend a similar approach to anyone else - take our modeling tools with a grain of salt.

While your overall conclusion is correct, I'm quite certain that judgement has a multiplier in it for weapon damage, so weapon damage DOES directly affect judgement (just less so than CS or DS). Edit: As a clarification, by Judgement I meant Judgement of Blood/Martyr.

You can find these coefficients on wowwiki:
Seal of Command - WoWWiki - Your guide to the World of Warcraft -> Judging takes 24% of Weapon damage
Seal of Blood - WoWWiki - Your guide to the World of Warcraft -> Judging takes 36% of Weapon damage
The other judgements do not depend on weapon's damage.

As of today, Crusader strike prioritization seems to be absolute.

Judging is losing a lot of damage, whereas the seal itself is more than doubling. On my tests even before the change today on PTR, I was finding Crusader Strike's new talents are making it scale far better... which lead to higher dps by using it 1st over judgements. (no 4pc)

Right, since CS procs Seals (and Blood's Seal will hit harder) and due to the PvP Libram, first with CS may be best.

It is good Redcape is waiting on redoing his sheet, since they keep changing things (this change is great though).

The buff the libram gives is 10 sec in 3.1, so it will have 100% uptime with any rotation.

This link to the Wowhead tool contains Redcape's updated stat weightings for 3.1: Items - World of Warcraft

Weightings taken from: The Retribution Paladin Thread (Wrath/3.0)

Items - World of Warcraft

is a better weighting (I just added a number for weapon dps).

I'm seeking a little clarification on the math behind this post...

How is the PvP Glove bonus being calculated, in relation of the other multipliers from talents? Because the glove bonus adds directly to the base multiplier before it's multiplied out by the talent modifiers.

(1.1+.05)*talents = CS damage

It's multiplicative, not additive. 110% of a normal strike with a 5% increase. (1.1 (include other modifiers here) X 1.05 (glove bonus) = 1.155).

For my calculations I assumed the CS bonus was purely multiplicative with all other bonuses. If that isn't the case I can change my model, but it is going to affect the final value by only miniscule amounts so it isn't a huge deal. At the absolute most it would change it from 35 to 34 or 36 or somesuch. If anyone has hard data suggesting how it does multiply out though I would be happy to change my sheet to reflect that.

I'm not expert at excel (I know enough to muddle about your spreadsheet and understand some of the simpler formulas), but I didn't see any point where you account for SotM/SoB having to double-dip in hit/expertise (roll for swing/strike, and another for the seal hitting). I'm likely mistaken, but could you double check that you are?

I ask because I noticed in my comparisons last night that the DPS differences between the Rawr and the spreadsheet were only significantly different when you are NOT at hit/exp cap, otherwise they were very close.

That was because in the latest version of Rawr I was triple counting hit/exp for Seal procs instead of double counting. I fixed this bug, and hopefully 2.2b6 will be out soon for people to test.

So, we're going to see RAWRs weights much more in line with Redcape's then? I mean, that's what I'm reading this as anyway.

Ermad or Zurm linked the new beta version of Rawr item weights in the Ret thread. The numbers were very close except for Hit/Expertise (Recape values hit more and Rawr values Expertise more). At least the other numbers were nearly the same.

Personally I think Redcape is closer to the mark.

Agreed. Str is still best stat. I still wouldn't gem hit (unless your Judgement misses like mad and you have no JotW mana) or expertise.

Hit logically should be better than expertise. Exorcism, Consecration, and Holy Wrath cannot, to my knowledge, be dodge/parried/blocked. Therefore if even a small portion of our damage comes from any of these sources (believe me, it does) then increasing hit increases their damage, increasing expertise would not. Anything increased by expertise is also increased by hit.

Therefore it makes sense to me Str>Hit>Exp but the margin between all three has narrowed significantly in 3.1.

I think you misread what he was saying:

It's still str > hit > exp, just redcape has hit at 1.69 and expertise at 1.19, if memory serves. We both value str exactly the same at 1.70.