The Retribution Concordance - 4.3, Voice of the DPS - Page 34

Neocaine · 01/21/11, 10:31 PM

Originally Posted by Elimine

I'm kind of surprised that there hasn't been any discussion about the Fury of Angerforge trinket. I'm new, so I apologize if I'm missing something extremely obvious, however an on-use trinket that has a cooldown which aligns with our DPS cooldowns (AV/Zealotry) and provides the same amount of strength as a ICD proc trinket (Heart of Rage, etc) makes it seem like a competitive trinket. I do not have the trinket and cannot test it, but has anyone on PTR been able to get their hands on it ever since the ICD on the stacks were lowered?

I have this trinket and have tested it both on live and on the PTR.

On live it seems to be very restrictive and underwhelming regarding the ICD, the stacks regularly drop off since you have a 5 second window to refresh it (buff lasts 15 seconds and the ICD is 10 seconds). This becomes even more problematic on fights with any sort of break in damage dealing during those precious five seconds.

On the PTR, the ICD was lowered from 10 seconds to 7, which although better, is still not enough. You don't get the benefit of an on-use trinket that you can bundle with AW (unless the stars align), or a purely proc based buff that would be up for longer.

Assuming a stationary fight in which you always trigger the proc every 7 seconds, you can get this up 35 seconds into the fight but the on-use still shares a cool down with other on-use ones. The assumption that you can proc it every 7 seconds is pretty big too, since it only procs on melee attacks.

EDIT: I just logged on now and did a few rounds on the dummies and it seems the ICD is now 5 seconds.

moonsond · 01/21/11, 11:23 PM

IMO, the biggest problem about our damage output is the balance between our serious burst when all 3 cooldowns are used and the overal low level of damage after they fall off. Retribution paladin is actually able to pull some serious and competetive damage from cooldowns, especially from GoAK, on short high-damage phases, think of Magmaw's head or Ascendant Council 3rd phase; but just half a minute later our DPS can drop by 30-40%, especially if some movement or target change is involved.
The ptr' changes to SoT stacikng will address the issues with dynamic fights, I hope; however, i really don't see any positive changes to our burst/sustained DPS balance mechanics, and GoAK would be the best subject to such. There is no trouble with its design in general, but the difference between our damage with or without it is way too large.
Not to mention our aggro, which is somewhat high. Even though i don't put out as much DPS as other melee classes, aggro balance on burst phases can be very stressful even now, when tanks didn't (or can't yet) build enough agro on the target.

An obvious solution would be to nerf STR stacking from GoAK, but increase the duration (by some deep-placed Retribution talent) of the "pet".
Also, some passive reduction of our threat level would be fine, we don't have any since the 4.0.1 and the removal of Fanatism.

Incadelico · 01/22/11, 12:16 AM

Originally Posted by Redcape

3. The best rotation I could find was:

Inq
TV
CS
HoW
Exo
Judge
HW
Cons

However! The difference between moving HoW/Exo up and down through TV/CS was virtually nonexistent. Into 20k dps the difference between those rotations was less than 30 dps, which is so far below the margin of error that is can be safely ignored. Basically the different rotations are closer together than ever before so just make sure you use all the good attacks (HoW, Exo, Inq, CS, TV) before the bad ones (HW, Judge, Cons) and you are playing optimally. More precision than that isn't possible due to variances in gear/situation and modelling limitations.

With the priority of CS moved up and the randomness in HP gen removed, should we still be using CS->2x filler->CS?
Otherwise we could have a sort of basic rotation with certain "soft" haste caps.
Unless my math is off, using a 3.5sec CS we could get away with using 2 3HoPw TV's during a 30sec inq (w/o 4pc T11) and then refreshing it right as it expires. This is, of course, assuming that divine purpose only procs when we'd be using a filler. (which isn't entirely impossible during several cycles)

Kinmaul · 01/22/11, 2:27 AM

Originally Posted by Avitus

Still above the other secondary stats, there's no significant change in the other stat rankings.

Edit: Beaten.

What this translates to is that gearing up has been extremely trivialized, all people need to do now is hunt for ilvl, regardless of stat combo, assuming equal STR values. The nightmare scenario of gearscore actually becoming an accurate representation of gear value (and not just an overhyped value laymen froth at the mouth over and veterans wag their fingers at) is upon us.

Past hit/exp caps, according to napkin math there's no point in reforging either, as having your stats in equilibrium as opposed to stacking 1 stat makes little difference. Note: In practice it's probably not a bad idea however.

I would say for the sake of sanity (considering we have the added difficulty of juggling a low non-dual wield hit cap and a very low expertise cap) this is a step forward. For the sake of "fun", it's definitely a step back.

For the people that are actually reading this forum gearing has always been trivialized. The Ret community here is robust and min/maxing is taken very seriously. Therefore at any given point the community knows exactly what stats/gear are the best and what is sub-optimal. Is it really such a bad thing that our secondary stats are relatively balanced? Hypothetically if crit suddenly became the best stat then everyone would gear/gem/enchant for crit to optimize their dps. However in truth the real determining factor for EJ ret paladin forum members wouldn't be their knowledge of the class, but rather the rng loot that drops week after week while farming bosses. If you didn't get your hypothetical crit gear drops then your knowledge of the class does nothing for your actual dps when compared to someone who got lucky with the loot table.

Making every stat relatively equal may lead to bland gearing choices, but given that every change to the spec is put to the mathematical grindstone it's not like there has ever been much of a decision (unless you count being at the mercy of a the rng loot table a decision). With the addition of reforging the individual stats of a piece of gear also lose a lot of importance when compared to previous expansions. The homogenizing of our dps rotation, like it was in wrath, is a much bigger threat to the difference of an average ret paladin and a great one then gear selection at this point in the game.

Avitus · 01/22/11, 5:05 AM

Originally Posted by Kinmaul

For the people that are actually reading this forum gearing has always been trivialized. The Ret community here is robust and min/maxing is taken very seriously.

I'm unsure why you latched on to my post to create a semantic argument (or lecture me about a community I've been engaged in since SSC = day 1), it's one thing to know/understand the stat weights and cherry pick your items based on preferred stats (and reforge into superior stats if you get unlucky with drops) - read: "educated choice that makes an impact", it's another to be able to take ANY dps plate of a certain ilvl and being guaranteed that it's just as good as all the alternatives - read: "trivial gearing, choice has no impact".

"Is it really such a bad thing that our secondary stats are relatively balanced?" that's not what I said, maybe you should actually read what you quote.

Originally Posted by Kinmaul

The homogenizing of our dps rotation, like it was in wrath, is a much bigger threat to the difference of an average ret paladin and a great one then gear selection at this point in the game.

As someone who used to juggle 4 different ranks of Consecration back in TBC, I agree that this is the bigger issue. I don't think being the bigger issue makes all other discussions taboo however.

Vistana · 01/22/11, 6:15 AM

I haven't seen anything in the thread about this, so I might as well ask.
Does Divine Storm double dip from the effects of AW?; if I have Avenging Wrath active, I'm getting the 20% more damage, so I'm getting more healing because it scales from damage. But AW also affects healing too, so would it be like:
100% normal damage + 20% extra damage from AW = 120% normal damage with it up (naturally)
At this stage, Divine Storm would be healing for 25% of this damage, which is 30% (?). But then we factor in that we're getting 20% more healing from AW, so we're now healing for 37.5% of our damage?
Excuse if the math is wrong, I'm quite tired >< but if this is true, then possibly in the future with Mastery affecting DS and Inq affecting Mastery, Divine Storm's heal could be semi-significant.

Tiara · 01/22/11, 7:12 AM

About the threat issue, we got Hand of Salvation exactly for that reason. I wouldn't like it if Blizzard dumbs it down and cancels all risk of bursting by lowering threat. This might change in higher tiers though, depending on how well tank threat scales. At the moment during my highest bursts i produce ~40-45k tps for 10 seconds or so which can be scary in the first 30 seconds of the fight. I run with a paladin tank so if he pops his wings he should stay above me 9/10 times. The other time is where i use salvation

Originally Posted by Vistana

Excuse if the math is wrong, I'm quite tired >< but if this is true, then possibly in the future with Mastery affecting DS and Inq affecting Mastery, Divine Storm's heal could be semi-significant.

Even with the changes i doubt DS would ever become a factor. As ret you're brought to do dps, if you need to heal use lay on hands, holy radiance or (preferably not) word of glory. Using DS over CS just because you want/need the heal is probably the wrong view on how to play ret if you want to play to your character's strengths. Having some decent healing on one of your attacks is nice but not if it hampers your dps as significantly as using DS over CS will do. This might change though if they decide to let DS generate 1 HoPo. At that point i can see paladins switching to DS during heavy aoe phases but right now, nah.

Vistana · 01/22/11, 7:36 AM

Originally Posted by Tiara

At that point i can see paladins switching to DS during heavy aoe phases but right now, nah.

I wasn't suggesting DS > CS for the heal or whatsoever, it was more just a thought when I was doing Heroic Maloriak and there was quite a lot of damage from the Vile Swills going out, and that if this situation arose later , perhaps in the next tier, where you needed to AOE hard with quite a bit of incoming damage going around, that DS may be more beneficial for the raid at the loss of a little personal DPS. I wouldn't say "we need the heal, let's DS for good measure", because right now it's not a lot, but I'm saying later with the patch and some more geare, it (DS and it's heal) could be more than just damage with a side of heal. It could be a whole bowl of each! - or something.

ironspants · 01/22/11, 8:19 AM

I don't think that the healing aspect of DS will ever be a great factor in raiding. sure, you will do more damage and healing with DS when better gear is available in later tiers of raiding, but the incoming damage will also be higher, as it was in every tier of raiding before.

In my opinion it is better to concentrate on single target damage, even more with the incoming changes to censure stacking, and throw in a tower of radiance when neccesary.

Tiara · 01/22/11, 8:55 AM

Keep in mind that losing out on single target dps also means the boss is going down slower, thus the healers need to heal more. I also think later tiers don't change this, since you might be healing more with DS then, but your single target dps should also rise keeping the dps loss in place.

Originally Posted by Vistana

perhaps in the next tier, where you needed to AOE hard with quite a bit of incoming damage going around, that DS may be more beneficial for the raid at the loss of a little personal DPS.

If it's an AOE situation, that would warrant the use of DS already regardless of incoming damage.

solacespecs · 01/22/11, 12:13 PM

I think the point to focus on is what, based on the current PTR build, is the break point for the # of mobs to use DS.

Talking about what the next content patch might be like and how it will affect DS is theoretical and useless for us currently.

I personally don't use the PTR, but with the latest build it would be a good idea for someone to go do some number crunching and provide some data, particularly with DS.

Mewee · 01/22/11, 12:24 PM

What may in fact become viable next patch for multitargets is censure spreading. Having it stack in about 2 GCDs, and running for a full 15 seconds, it becomes quite easy to keep it running on 3 or 4 targets simultaneously. I believe this was in fact the preferred way for Ferals to AoE before their bleeds got nerfed, by tab-spreading Rake on more than one target for massive aoe damage. Now our Censure is not as strong as their bleeds, but if thing stay as they are on the PTR, it is almost effortless to apply - the rampup for each mob you want to inflict, is 2-3 gcd, but the maintenance upkeep is only 1 GCD or a melee swing. I could easily see this being a decent DPS gain if done properly.

chippydip · 01/22/11, 3:43 PM

Originally Posted by Danath

Yes, but at the very least would these (slight, in the end) changes mean that 50 Mastery enchant is now better than 32 AP on Feet slot? Seems possible to me, taking for granted those stat weights.

Yeah, 4.0.6 should be the end of wrath enchants for us (50 str for bracers and 50 mastery/hit/haste for boots). I'm actually starting to think that the "naked" stat weights are undervaluing secondary stats relative to str a bit and that 50 hit/haste may be slightly better than 32 AP already.

I added damage calculations to my rotation simulator and run some tests to see what I got for stat weights. Nailing down exact numbers with a simulator is quite difficult due to the RNG involved in our rotation. Over the course of a 6 minute fight, I'm getting a StdDev in dps of about 650 and a Mean around 18.4k. To try and minimize the error, I repeated 10 hour tests a couple hundred times, but the StdDev was still nearly 50dps. The numbers seemed fairly stable, though, so here's the results for a +100 stat increase:

Stat	4.0.3a	4.0.6
str	~200	~220
crit	~80	~90
haste	~80	~85
mastery	~40	~90

Same ballpark as Redcape's number, which is good. Overall, its looking to me like 4.0.6 is bringing mastery to where it should be and providing an overall ~10% dps gain (I'm seeing overall dps go from low 18k to low 20k in a 359 gear setup). Haste seems not to gain quite as much, probably due to HoL no longer proccing from auto-attacks.

I also ran some Inq refresh tests. For the most part it doesn't seem to matter a whole lot as many of the options I tested fell within the ~50dps StdDev of a 10 hour test. Still, I ran 30 tests each testing refresh times between 0 and 10 seconds left on Inq and refreshing at 1-3 HP min and then looked at the best 10 results to get a feel for what works best ("3s" means refresh at 3 stacks if less than 3 seconds remain, "2hp" means refresh at 2 or more HP if Inq has fallen off):

4.0.3a no 4pT11
0s 3hp 18088.3

1s 3hp 18102.5
1s 2hp 18078.2

2s 3hp 18098.3
2s 2hp 18094.6

3s 3hp 18100.1
3s 1hp 18089.8
3s 2hp 18087.5

4s 3hp 18086.2
4s 1hp 18081.1

Without 4pT11, refreshing at 3hp looks like its probably the best option. Refreshing with a small timer left looks best (which will result in less Inq uptime).

4.0.3a w/ 4pT11
0s 1hp 18374.4

1s 2hp 18400.8
1s 1hp 18398.9

2s 1hp 18398.8
2s 2hp 18379.2

3s 1hp 18396.4
3s 2hp 18392.2
3s 3hp 18376.5

4s 1hp 18384.5
4s 2hp 18373.5

With 4pT11, refreshing with whatever hp you have seems best. Again, refreshing with minimal overlap looks good with less downtime due to refreshing at 1+HP.

4.0.6 no 4pT11
0s 2hp 20201.2
0s 3hp 20201.0

1s 2hp 20211.7
1s 3hp 20196.1

2s 2hp 20253.5
2s 3hp 20211.0

3s 3hp 20227.0
3s 2hp 20210.0
3s 1hp 20197.0

4s 2hp 20202.7

Here we see 2+hp refresh edging out 3hp at the shorter times (more holy damage from the new mastery makes it more important to get Inq up again once it falls). This still isn't enough that longer overlap points start showing up in the top 10, however.

4.0.6 w/ 4pT11
0s 1hp 20656.5

1s 1hp 20635.5
1s 2hp 20625.2

2s 1hp 20645.9
2s 2hp 20617.3

3s 1hp 20650.5
3s 2hp 20627.0

4s 1hp 20638.4
4s 2hp 20607.3

6s 2hp 20599.8

Here refreshing ASAP is universally the best choice. 4pT11 means those lower hp Inqs are a more efficient use of the HP and 4.0.6 means letting Inq drop is a bigger dps loss. It still isn't worth refreshing with a 3hp Inq very soon before it drops off.

This made me curious, so I tested refreshing only at 1HP while Inq was still up and then at any hp if it falls (4.0.6 w/ 4pT11):
5s 20668.9
6s 20666.5
1s 20662.5
0s 20654.4
2s 20647.1
8s 20646.3
7s 20645.1
4s 20642.5
3s 20641.1
9s 20606.5

These are slightly better results with refresh times from 0s to 8s falling within a 20640-20670 dps range. (No real preference for a specific time within that range, but falling off noticeably for 9s or more). I tried a few other permutations of the refresh rules, but couldn't come up with anything better than this.

(If anyone is curious, the code for the simulator is available here: https://github.com/chippydip/RetRotationSim but you'll need to have a C# compiler and the patients to look through the code and figure out what's going on to make any use of it at this point ;-))

Aviendha · 01/22/11, 5:21 PM

I just went and tested on the PTR and can confirm that Glyph of Exorcism now does damage according to the actual damage done to the target. However, the DoT damage is calculated regardless of critical hits (if your exo crit'ed, the glyph damage would be the same as if you hadn't crit'ed). I can also report that the glyph still "refreshes" its damage on each cast and does not "stack", thus ignoring any leftover dots left on each recast.

This most definitely increases the value of Glyph of Exo, which was already the best third glyph choice alongside TV and Seal of Truth(though there was some debate).

EDIT : Mmmmh, it seems actually more complicated than that on second thought. The glyph's DoT damage reflects somehow the average damage of my exorcism, but small variations in exo damage don't seem to affect the DoT :

Alvi's Exorcism hits Raider's Training Dummy for 8103 Holy.
Raider's Training Dummy suffers 541 Holy damage from Alvi's Exorcism.
Raider's Training Dummy suffers 540 Holy damage from Alvi's Exorcism.
Alvi's Exorcism dissipates from Raider's Training Dummy.
Raider's Training Dummy suffers 540 Holy damage from Alvi's Exorcism.

Alvi's Exorcism hits Raider's Training Dummy for 7965 Holy.
Raider's Training Dummy is afflicted by Alvi's Exorcism.
Raider's Training Dummy suffers 540 Holy damage from Alvi's Exorcism.
Raider's Training Dummy suffers 810 Holy damage from Alvi's Exorcism.(Critical)
Raider's Training Dummy suffers 540 Holy damage from Alvi's Exorcism.
Alvi's Exorcism dissipates from Raider's Training Dummy.

Then with AoW proc :

Alvi's Exorcism hits Raider's Training Dummy for 15197 Holy.
Raider's Training Dummy is afflicted by Alvi's Exorcism.
Raider's Training Dummy suffers 1548 Holy damage from Alvi's Exorcism.(Critical)
Raider's Training Dummy suffers 1032 Holy damage from Alvi's Exorcism.
Raider's Training Dummy suffers 1032 Holy damage from Alvi's Exorcism.
Alvi's Exorcism dissipates from Raider's Training Dummy.

chippydip · 01/22/11, 7:55 PM

I spent the afternoon playing with more rotational variations in my simulator (all with the 4.0.6 mechanics).

Moving CS ahead of TV in the priority list (when holy power is less than 3) resulted in a dps gain. This seems reasonable since it should basically switch TV, CS into CS, TV when possible (putting CS on cooldown sooner and resulting in more holy power generation in the long run).

Moving HoW ahead of TV also resulted in a small dps increase. Most likely this is due to a similar effect (reordering TV, HoW into HoW, TV for fillers between CS when possible). This has the downside of potentially wasting a DP proc on HoW, but presumably the upside of potentially squeezing one more HoW into an AW is enough to counteract this effect.

I also noticed that it didn't really matter where I put Inq in the priority list as long as it was ahead of J/HW/Cons. All of these rotations resulted in almost exactly the same dps (~21.2k):
Inq > CS (below 3hp) > HoW > TV > Exo > others
CS (below 3hp) > Inq > HoW > TV > Exo > others
CS (below 3hp) > HoW > Inq > TV > Exo > others
CS (below 3hp) > HoW > TV > Inq > Exo > others
CS (below 3hp) > HoW > TV > Exo > Inq > others
Most likely this is just a result of the damage you lose by using a couple abilities without Inq up equals the extra damage you get by getting things on cooldown sooner and using up DP and Exo procs.

(I also noticed that using AW and Zeal on cooldown was better then waiting for 3HP to keep them stacked, but I suspect this is just a result of being able to fit a few more AWs into the long sim time and that in real world use its best to stack them since presumably a 120 vs 125 second effectively cooldown on AW won't get you an extra AW in most realistic fights.)