Jump to content


Photo

Retribution DPS Theorycraft


  • This topic is locked This topic is locked
7310 replies to this topic

#41 Handled

Handled

    Piston Honda

  • Members
  • 103 posts

Posted 17 October 2007 - 05:49 PM

I'm somewhat curious if anyone has started and maintained any type of Retribution Equipment thread as I didn't see it on the first few pages but I have started to question the full AP only gear model and have since considered 2 T5 and 2 T6?

Gear Theorycraft can be very very painful.

#42 flyingtoastr

flyingtoastr

    Bald Bull

  • Members
  • 2,257 posts

Posted 17 October 2007 - 06:43 PM

I'm somewhat curious if anyone has started and maintained any type of Retribution Equipment thread as I didn't see it on the first few pages but I have started to question the full AP only gear model and have since considered 2 T5 and 2 T6?

Gear Theorycraft can be very very painful.


I know the general consensus (especially for BE pallys) is that warrior gear with +haste is far superior than tier sets. Since I have yet to see any accurate information regarding the bonus proc rate for Tier 6, I can't tell whether it would be worth the investment or not. According to WoWhead its a 20% proc rate, but I would much rather see a WWS or combat log than trust it. As for the 2 piece of Tier 5, its effectiveness is entirely dependant on your use of Judgement. If you use Judgement every cooldown it works out to a fair amount of saved mana, but you would be sacrificing a decent amount of stats (especially haste) to use it over non-tier BT gear.

To the best of my knowledge no one has yet tried to keep an up to date equipment list for ret pallys though.

#43 Handled

Handled

    Piston Honda

  • Members
  • 103 posts

Posted 17 October 2007 - 07:19 PM

The problem with haste is it's haste and hit and lacks all forms of crit in max haste gear my crit is so bad I would need amazing other pieces to increase the crit.

#44 flyingtoastr

flyingtoastr

    Bald Bull

  • Members
  • 2,257 posts

Posted 17 October 2007 - 07:38 PM

The problem with haste is it's haste and hit and lacks all forms of crit in max haste gear my crit is so bad I would need amazing other pieces to increase the crit.


This may very well change with the threat reduction, but right now you only want enough crit to keep Vengeance up as close to 100% as possible. After that crit only makes your aggro more spikey while not increasing your DPS as well as AP. With the change to Vengeance and the addition of more crit to the tree (Sanctified Seals) it will mean you will need much less crit from gear to keep that 100% uptime. There is a point where AP starts to give diminished returns for overall DPS as compared to crit, but I'm not sure of the exact numbers. I'm sure a full time retadin could help a bit more there than myself.

#45 Sapp

Sapp

    Bald Bull

  • Members
  • 1,514 posts

Posted 17 October 2007 - 07:43 PM

As a side note on gear since I can't load the calculator in open office...

How do Cataclysm's Edge and the Torch of the Damned compete?

I have fairly free options on both, since the only raiding warrior with anything like strong PvP inclinations dropped off of arenaing recently and respecced Fury, and already has a Stormherald anyway.

Basically, is the -.3 speed on the Cataclysm enough of a deficiency to outweigh the ~60 AP and 8 DPS advantage (and 300 armor penetration)?

I'm trying to decide on their relative value.

I always thought Torch was best as a given but I want to evaluate it in a more complete fashion.

#46 Rho

Rho

    Von Kaiser

  • Members
  • 30 posts

Posted 17 October 2007 - 07:52 PM

I have a question about what kind of group a ret paladin needs to be in to be successful. Currently we have a melee group that looks like this:

DPS Warrior (fury)
DPS Warrior (fury)
Rogue
Rogue
Enhancement Shaman

Presumably if one paladin in our raid goes retribution, he would displace a warrior, and an additional healer would be invited.

However, in the instance that we only have 25, including 2 dps warriors, how should group allocation be handled. Could a paladin make use of a shadow priest group and incorporate consecration into the dps cycle? Should the warrior be put with the hunter/feral group?

#47 Luxury

Luxury

    Von Kaiser

  • Members
  • 62 posts

Posted 17 October 2007 - 08:07 PM

However, in the instance that we only have 25, including 2 dps warriors, how should group allocation be handled. Could a paladin make use of a shadow priest group and incorporate consecration into the dps cycle? Should the warrior be put with the hunter/feral group?


I wouldn't recommend putting the ret pally with the shadow priest as a replacement for WF. Cons spamming won't make up for the lost dps.

#48 Luxury

Luxury

    Von Kaiser

  • Members
  • 62 posts

Posted 17 October 2007 - 08:10 PM

After the haste nerf and the recalculation of PPM % chance on the after-haste weapon speed, it is definitely not worth it to trade crit for haste if you're an alliance paladin, assuming the SoV tests apply to SoC. For SoB pallies it was definitely worth it to trade crit for haste before the haste nerf, and about even after the haste nerf.

#49 Arkhubar

Arkhubar

    Glass Joe

  • Members
  • 7 posts

Posted 17 October 2007 - 08:12 PM

It was suggested to me I repost here. Have at it! ^_^

Ok. The below are some theories I've come up with and had limited ability (or none at all) to test over the last few months. I know there are errors. I need help finding them. Shoot holes in the theories if you can, because I can't test these theories well on my own at this time. But maybe this will give some people some different ways to think about optimizing Retadin DPS. Also, I already know places where I need help adjusting the numbers for 2.3 (or in general), I'll outline them all and hopefully people will be kind enough to supply them for me. If someone after seeing this wants to try and lend me a hand with testing let me know, I <3 you long time.

Theory 1: Spell Rotation

I've noticed quite a few paladins use a rotation of the following kind based on their WWS logs: Seal of Blood/Command, Judgement, Crusader Strike. Some exorcism thrown in (1-2), some consecrate thrown in (nothing major). It also seems like they do not use very many if any mana potions.

Another common trend is people tend to like to go BS for the weapons. Therefore I propose the following:

The superior DPS rotation to use is Consecrate (Rank 4), Hammer of Wrath, Seal of Blood/Command, Judgement, and Crusader Strike. If exorcism is available, that takes priority over Consecrate in terms of mana (but if you can fire both off then do it).

You can achieve this rotation easily if instead of leaning BS you take alchemy and use an alchemist's stone, even if you don't get a judgement of wisdom up on the mob from another paladin (but you should). Some numbers:


14.875 mana per second for Seal of Command
26.666 mana per second for Exorcism (2.3 COST: IDENTICAL?)
91.666 mana per second for Hammer of Wrath (2.3 COST: 73.333, - 18.333)
23.6 mana per second for Crusader Strike (2.3 COST: 39.333, + 15.667)
15.5 mana per second for Judgement
48.75 mana per second for Consecration rank 4
82.5 mana per second for consecration rank 6


Obviously these numbers will be off for 2.3. If someone would be kind enough to supply me the new mana costs for 2.3 (and if someone would let me know if the mana cost on CS has changed), I can adjust all these numbers. Assume HoW, Exorcism go down and CS goes up for now. This means that the standard rotation most rets use goes up in mana cost and the new one will not be as expensive as before.

Quote:
Rotation 1:
Seal of Command, Judgement, Crusader Strike: 53.975 (2.3 COST: 69.642)

Rotation 2:
Seal of Command, Judgement, Crusader Strike, Consecration Rank 4: 102.725 (2.3 COST: 118.392)

Rotation 3:
Seal of command, Judgement, Crusader Strike, Consecration Rank 4, Hammer of Wrath: 194.391 (2.3 COST: 191.725)

Rotation 4:
Seal of Command, Judgement, Crusader Strike, Exorcism, Consecration Rank 4: 129.391 (2.3 COST: 145.057)

Rotation 5:
Seal of Command, Judgement, Crusader Strike, Exorcism, Consecration, Hammer of Wrath: 221.057 (2.3 COST: 218.391)


At this point I'll note I'm intentionally not including blood. I'm not going to bother modeling it, just assume automatically it is far more mana efficient (lower mana cost and mana back from others healing you). Doesn't really change the point. But if you notice, the rotation I'm proposing goes anywhere from 2x-4x as much mana per second drain on the paladin as the standard rotation. However, I swear its doable. Here's why:

Super Mana Potion: 20
Fel Mana Potion: 26.6666667
Super Mana Potion w/ Alchemist's Stone: 28
Fel Mana Potion w/ Alchemist's Stone: 37.3333333


values for chain chugging pots w and without alchemist's stone. If you notice fel manas w/ alchemist almost cover consecrate rank 4 by themselves. Now use major magebloods, make sure your fellow dins get wisdom up there for you to refresh, etc, and its very easy to keep up the mana. Even without wisdom, if I have blood I can pretty much perma sustain it without trying. I don't recall if I can do it no wisdom and command without it being a fight where I take damage.

The theory is as such:

Unless you are threat capped already,if you can generate more mana and then use it, then you generate more dps than you would by leaning BS and not using alchemist's stone (and if you don't chain chug and you aren't at threat cap then you just are playing sub optimally). For me, Consecrate rank 4 is worth 88 dps on a single target, and if you can hit multiple targets it gets even better. The BS weapons IMO aren't superior enough to justify passing up on 15 all stats and the extra mana, and I don't know if a trinket exists that will provide more dps (I mean that - maybe one does, I just don't know). The rotation I proposed above is a consequence of this theory: Its roughly what I am comfortable throwing at a mob in a regular rotation without bottoming out. It might be possible to do better, I'd need some time to sit at a mob using multiple ranks of consecrate etc.

Theory 2: Spell Damage maybe doesn't suck *that* much

When you use blood, judgement of blood, CS, and white hits as your only dps, yes. Spell damage sucks. But notice I have multiple things in that rotation which all benefit from spell damage - especially if you run command. So how good is spell damage roughly?

Here is my math and assumptions. I need people to give me more accurate numbers and assumptions and I will re-run the numbers to get something more accurate:

1) Assume ideal circumstances - all spells go off every time they are off cooldown. Yes, this obviously isn't true, but for the sake of trying to keep things simple to start I'm fudging.

2) Assume 3.6 speed weapon, 35% melee crit, 7% spell crit, enh shaman and kings. The last two favor str and attack power, the rest is just for a starting point. I also assume the % bonuses are multiplicative, so its 1.1*1.1*1.1 (kings, 10% str talent, and 10% AP) more mileage out of str than base. ALSO: I know my spell coefficients are off. If someone can give me the right ones I'd appreciate it a great deal.

3) Assume following mathematics:

Spell based additions to DPS:

1) 5 Spell damage = 1 dps consecrate (single target assuming)
2) (Spell Damage * .4) / 6 seconds = CS dps from dmg <- adjusted for 2.3
3) (Spell damage * .43)/ 6 seconds) * .2 = HoW DPS from dmg.
4) ((Spell damage * .43)*6.8 PPM)/60 seconds = SoC dps from damage.
5) (Spell damage * .43) / 8 = JoC dps from dmg
6) Exorcism isn't counted for the number I'll present here, but math is (spell damage * .43) / 15

Str based additions to DPS:

1) Attack power / 14 = white dps from str
2) (white dps * weapon speed) / 6 = CS dps from str
3) white dps * .35 = SoB DPS
4) ((white dps * weapon speed) * ((.7 to holy damage * 6.8 PPM)/60 seconds)) = SoC dps from str


The value I get for the value of dmg under the above circumstances is .610001 for SoB, and .916675 for SoC. Spell crit raises the value in favor of dmg, melee crit lowers it, weapon speed lowers it, but not by any massive margins. I will produce precise numbers about that once I get the math corrected, and upload a spreadsheet so people can modify it/see I'm not making things up.

However, these numbers seem to be higher than people were originally predicting the value for dmg conversion. I've seen numbers like .4 etc. I'd like some help correcting this to get an exact value for a spell heavy rotation. It might be higher or lower, but knowing would certainly allow us to evaluate tier X v dps plate etc. Seems like for alliance 'dins dmg might be VERY valuable.

Concluding thoughts

Once again, I stress that I'm extremely limited in my ability to test. My guild was strugging before and it has decided to stop raiding as of tuesday (which was also the day I realized the possibility of theory 2, before I assumed Dmg sucked like everyone else - my gear needs to be redone if I'm right). I've put this out here at last because it seems like I'm not going to get an opportunity to personally test this before coming forward with it, and I may never even get to see the fruits of this if I'm right. So, if you can help me refine the theories or shoot em down, by all means. I put this forward now to try and help further the cause for rets like me everywhere.

I think I might be right, at least partially. If that's true, there is a lot more dps to squeeze out of ret. So, I'm curious what you all think of this. Happy to answer questions, and I appologize if I'm just a newb who made a ton of dumb errors. Have at it!



#50 vokzhen

vokzhen

    Von Kaiser

  • Members
  • 57 posts

Posted 17 October 2007 - 08:19 PM

It seems to be a two-fold problem. Firstly, OOo is not compatible with Excel's macro language at all, so things like the buttons and drop-downs don't function at all. Also, any fields that seem to reference hidden fields don't function. Things such as the totals for stats, the row with values for Blessing of Kings, etc, all give 502 errors (Invalid argument; Function argument is not valid, for example, a negative number for the root function.)

I'd help decipher the issues, but I have never actually worked with macros, and can't figure out what's going on in the hidden cells. Perhaps this will help anyone who's working on this.

If it helps at all, I get macro warning in OOCalc when I load the spriest spreadsheet but get no errors, and I *think* it does incorporate hidden cells, but I might be wrong. As least, I don't see on the sheet where any of the calculations are actually done, unless there's a difference between "cannot be seen by default" and "hidden."

The dropdown menus in your spreadsheet work fine, sorry I didn't clarify that (though they seem to like to break as well, the mage spreadsheet does this). It's just everything else that's messed up :P

#51 Wrl

Wrl

    Von Kaiser

  • Members
  • 33 posts

Posted 17 October 2007 - 08:25 PM

Just a weird calculation mechanic I discovered a while back. Certain conditional multipliers such as Crusade will apply to Seal of Blood and Seal of Command twice.

With sanctity aura off, I was hitting for more than 35% of my white damage when using a low dps, low damage range weapon. I created a pretty big sample size and found that it was hitting noticeably harder (I believe it was 3%, but I don't have the numbers anymore).

I believe that crusade first applies to your base melee damage, then when blood determines its damage it applies the modifier again.

I don't believe this is the case for Sanctity Aura, because that shows up on your character sheet specifically.

If anyone wanted to run this test again to back this up, that would rock.

#52 flyingtoastr

flyingtoastr

    Bald Bull

  • Members
  • 2,257 posts

Posted 17 October 2007 - 08:28 PM

Obviously these numbers will be off for 2.3. If someone would be kind enough to supply me the new mana costs for 2.3 (and if someone would let me know if the mana cost on CS has changed), I can adjust all these numbers. Assume HoW, Exorcism go down and CS goes up for now. This means that the standard rotation most rets use goes up in mana cost and the new one will not be as expensive as before.


2.3 Mana costs are:
Exorcism (Rank VII): 340 mana
Hammer of Wrath (Rank IV): 440 mana

Consecrate mana costs remained constant.

Crusader Strike remained at 236 mana on both Live and the PTR (for myself) which somehow works out to be ~6% of my base mana (as opposed to the 8% that is advertised). I must have something wrong here, but the fact remains that the cost is constant.

Good post BTW.

#53 Arkhubar

Arkhubar

    Glass Joe

  • Members
  • 7 posts

Posted 17 October 2007 - 08:33 PM

2.3 Mana costs are:
Exorcism (Rank VII): 340 mana
Hammer of Wrath (Rank IV): 440 mana

Consecrate mana costs remained constant.

Crusader Strike remained at 236 mana on both Live and the PTR (for myself) which somehow works out to be ~6% of my base mana (as opposed to the 8% that is advertised). I must have something wrong here, but the fact remains that the cost is constant.

Good post BTW.


Thanks, I'll go in and adjust all my numbers presently.

I just wish I had useful personal data to back it up, I've had the first theory for months, just variety of things have conspired against me to prevent me from having a WWS log really worth a damn, and now I'm guildless. :P

edit: updated based on new mana costs. Unless I'm crazy Exorcism is the same, HoW is now a good amount less and CS is a good amount more in MPS costs. Makes mana generation even more critical than before.

#54 Arkhubar

Arkhubar

    Glass Joe

  • Members
  • 7 posts

Posted 17 October 2007 - 08:48 PM

Ok, attempt to model WF in the spreadsheet is as follows, double check my thoughts:

Add 20% of whatever the original benefit of +dmg/str was (if WF is true) for the following abilities:

White DPS, SoB DPS, SoC DPS.

So for +dmg only SoC benefits to +dmg get a boost, for str +dmg gets a boost. The coefficients I get with WF on are .490719 SoB and .763424 SoC in terms of how much str 1 dmg is equal to. Again, assuming Kings and Enh shaman, without enh the numbers go up to .539791 and .839766. edit: fun fact, without WF, without enh shaman, +dmg exceeds the value of str for retadins using SoC assuming you can use Cons CS HoW SoC JoC every time without fail.

Once I make sure most of the basic assumptions are right and I didn't miss anything silly I'll upload the spreadsheet (anyone got a good method to do that?)

#55 Artaxz

Artaxz

    Piston Honda

  • Members
  • 144 posts

Posted 17 October 2007 - 09:05 PM

Do we have any idea what value to place for each state for purposes of itemization? For instance, the enhancement shaman thread gives fairly well tested values (2.0 str, 1.0 Ap, 1.61 crit, 1.17 hit I believe). I fully understand that STR would be the stat given the most weight (perhaps 2.2 based on the previous listed figures), while agi and crit would be fairly low. As a BE, haste is worth about the same as crit I believe, but I have no real idea on how to prioritize gear.

My guild is asking me to experiment with ret in 2.3, and I have access to any gear. What I am really asking for is item values to use with Lootzor I suppose. My ret gear has some BT trash drops, and pretty much the best from TK/SSC (1800 ap, 20 crit, some haste, and hit capped). I would think with the 8% crit in the ret tree that I would have 100% uptime with vengeance, and could just stack STR and haste.

I looked at T6 ret and it is underwhelming. Any geared ret paladins that can offer gear choices?

#56 Prepared

Prepared

    Von Kaiser

  • Members
  • 85 posts

Posted 17 October 2007 - 09:24 PM

If it helps at all, I get macro warning in OOCalc when I load the spriest spreadsheet but get no errors, and I *think* it does incorporate hidden cells, but I might be wrong. As least, I don't see on the sheet where any of the calculations are actually done, unless there's a difference between "cannot be seen by default" and "hidden."

The dropdown menus in your spreadsheet work fine, sorry I didn't clarify that (though they seem to like to break as well, the mage spreadsheet does this). It's just everything else that's messed up :P


I managed to load a development version of OOo which runs some VB script stuff just fine. I did receive an error on enable macros, but the buttons actually worked. However, the calculations are still off, and I get a lot of 502 errors all over the place. This makes me think the errors there are independent of the actual macros.

#57 Cromfel

Cromfel

    Don Flamenco

  • Members
  • 462 posts

Posted 17 October 2007 - 09:59 PM

I have been trying to find values for JoW and JoL how they effect certain classes. For reference I took Blood legion WWS reports, calculated class attack "ticks" per second and tried to find HPS and MPS values for each class.

So far it looks like this...

Judgement of Light
Rogue 142 HPS
Warrior (A) 55 HPS
Shaman (En) 101 HPS

Judgement of Wisdom
Mage 25 MPS (125 mana per 5sec)

Anyone have better way to measure classes theoretical interaction with bosses that could trigger JoL and JoW? Is there better ways to present how judgements effect your raid members? Im open for all sugestions, this is just what I thought could be good way to show just few raw numbers how each judgement add.
.:. Retribution Paladin Hideout .:. http://cromfel.battlefield.fi/

#58 Ragnor

Ragnor

    King Hippo

  • Members
  • 606 posts

Posted 17 October 2007 - 10:01 PM

I have a question about what kind of group a ret paladin needs to be in to be successful. Currently we have a melee group that looks like this:

DPS Warrior (fury)
DPS Warrior (fury)
Rogue
Rogue
Enhancement Shaman


The melee group makeup really depends on the total raid makeup, obviously you are looking at total raid dps. If one of the rogues is daggers they would be the first to go from that group probably, of course this assuming you can slot them in somewhere else with lotp,bs, tsa etc otherwise it's probably not worth it.
The universe is run by the complex interweaving of three elements. Energy, matter, and enlightened self-interest.

www.retpaladin.com

#59 bellator

bellator

    King Hippo

  • Members
  • 580 posts

Posted 17 October 2007 - 11:50 PM

Have been spending some time this evening on the dps model.

Unfortunately most of that was spent on creating an "I Win" button in which it would analyse every combination of item to come out with the perfect gear. Unfortunately i overestimated the processing power of my computer and when the timer i had built in to keep track of it's progress told me i had 50 million millenia left until the calculation was done i decided it wasn't worth it. Even filtering out items < item level 140 still left me with an 80million combination / 37 day calculation.

On the plus side learnt some new coding, so implementing a nice new section which will compare individual items against all items of the same type in the database. Not a universal theory unfortunately but a nice extra. Will try to get it out tomorrow.

#60 flyingtoastr

flyingtoastr

    Bald Bull

  • Members
  • 2,257 posts

Posted 18 October 2007 - 12:16 AM

I have been trying to find values for JoW and JoL how they effect certain classes. For reference I took Blood legion WWS reports, calculated class attack "ticks" per second and tried to find HPS and MPS values for each class.

So far it looks like this...

Judgement of Light
Rogue 142 HPS
Warrior (A) 55 HPS
Shaman (En) 101 HPS

Judgement of Wisdom
Mage 25 MPS (125 mana per 5sec)

Anyone have better way to measure classes theoretical interaction with bosses that could trigger JoL and JoW? Is there better ways to present how judgements effect your raid members? Im open for all sugestions, this is just what I thought could be good way to show just few raw numbers how each judgement add.


The problem with saying "JoW gives x mana per 5 seconds" is that JoW depends greatly on your rate of attack and the kinds of damage you're doing. In other words its entirely dependant on your raid makeup and the specs of everyone hitting the mob. I had a WWS a while back (I think its deleted by now) of a Mag raid where you can see just how much of a difference there is between say a mage (spamming spells) versus a spriest (dots and such). In fact, there are huge differences in spec as well, our destro lock had almost triple the procs as an affliction lock.

I think as a blanket statement you could say JoW restores a ton of mana. But coming up with exact numbers for every class and spec in the game, not counting outside variables, would be a very difficult if not impossible task.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users