3.2 PTR - Retribution Discussion - Page 11

path411 · 06/28/09, 12:08 PM

Originally Posted by Farast

But remember that Prot Pallies are almost certainly going to pick it up so unless your raid only has DK tanks you shouldn't need it.

It isn't a guarantee that prot will be grabbing it, as it requires them to drop an additional 2 points out of conviction for it. (Though it is definitely worth grabbing if it's not being covered by someone else) Though it will be something to work out, per guild basis, depending on how many ret/prot paladins you typically bring.

Also, it's worth mentioning warriors/druids typically don't fully talent imp demo shout/roar. And I'm pretty sure 2/2 vindication is equivalent to the fully talented debuff. (I think untalented to talented is in the range of an extra ~5% damage reduction from the boss).

Redcape · 06/28/09, 12:44 PM

Originally Posted by flyingtoastr

As usual, it doesn't matter what the raw damage is (otherwise Exorcism would handily take first priority). DPS based on effective cooldowns is still the rule of thumb for determining priority. At a 4 second cooldown CS loses some 20% of its total DPS by extending the effective cooldown by just one second. Judgement, at 8 seconds, loses only ~11% of its DPS for each second the effective cooldown is extended. Thus Judgement's DPS has to be that much larger than CS's to be worth more in the priority.

This isn't necessarily true with our new rotation. I will explain:

First, define GCD Jammed (GJ) as a situation where you have all GCDs spoken for. You do not have random gaps to fill things in. Define GCD Open (GO) as a situation where you have gaps in your rotation to fit things in on a regular basis.

When you are GO the mantra of 'highest dps first' is definitely true. You should delay by the smallest amount of dps possible because you know that that high damage ability with a long cooldown will fit in shortly. This currently describes our situation as of 3.1 which is why we use highest dps first.

When you are GJ though the math is different. You may not be able to fit in a high damage ability for a long time because you simply have no space, so you want to hit the highest damage ability first and not worry about cooldowns. Example:

Bonk: 1k damage, 3 sec cooldown
Bash: 1k damage, 3 sec cooldown
Crush: 4k damage, 20 sec cooldown

If these three abilities are all we have access to a 'highest dps' priority system our rotation would be

Bonk
Bash
Repeat

Clearly that is awful as we never use our Crush ability! The question is: Does changing CS to 4 seconds actually push us into a GJ situation? By pure Patchwerk 0 latency standards, it does not. However, real fights are not like that and we have to shift, watch the screen, listen to vent, etc. and nobody has a 0 latency. Doing those things reduces the number of abilities we can use and pushes us close to GJ. Given that we should be dropping out SS every minute, DP every minute, stunning every minute (situational, obv) Salving every 2 minutes, cleansing (situational), BOPing, DiSacing, LOH, etc. etc. we absolutely should be GJ if our utility is coming into play.

If we ARE GJ then we need to revert to a 'highest DAMAGE' priority system instead. I think in 3.1 we are pretty much never GJ, even on fights like Hodir where pretty much the full gamut of our utility is used. In 3.2 I think we will absolutely be GJ on that fight and on some others.

Modeling rotations assuming zero latency, perfect coordination, etc. only goes so far. If you want to find out if we are REALLY GJ or GO then you should go fight Ignis and look at the WWS. What are the ACTUAL effective cooldowns of your abilities? They probably are a lot worse than the spreadsheet cooldowns and that would suggest a differing priority should be considered.

Until we have these sorts of results from a relatively trivial tanknspank to work with I don't think we will know for sure whether we have crossed to GJ territory. I look forward to those results.

flyingtoastr · 06/28/09, 1:21 PM

You missed the major point of my post Redcape.

Doing a Bonk > Bash rotation without using Crush does a total of 13000 damage over 21 seconds.

Now, if you include Crush as often as possible (assuming you don't waste free time the eCD is 21 seconds) you a Bonk every 21 seconds for a Crush. However you gain the damage that Crush does. You total damage for those 21 seconds is now 16000. As you can see, trading one Bonk or Bash for the higher damage Crush has resulted in more damage over the same period of time! Why is this?

We'll use the acronyms DPS/ic for "Damage per Second per Ideal Cooldown" and DPS/ec for "Damage per Second per Effective Cooldown". Ideal is the base cooldown of the ability, effective is what you get in practice, naturally.

Bonk and Bash have a DPS/ic of 333 each, while Crush has a DPS/ic of 200.

The issue again comes to your DPS/ec. Even though Crush's DPS/ic is lower than that of both Bonk and Bash, extending the eCD of Bonk to 3.5 seconds was still worth the loss in that ability's DPS. With a 3.5 second eCD Bonk's DPS is reduced by roughly 14.3% to ~285, but you gain the 190 DPS from Crush, easily balancing out the loss. That is why it works out better to extend the eCD on Bonk, because the gain from using Crush outweighs the loss from Bonk - not any convoluted GCD lock mechanics.

I'll go ahead and quote myself:

Originally Posted by flyingtoastr

Thus Judgement's DPS has to be that much larger than CS's to be worth more in the priority.

Judgement can be higher prioritized than CS easily. It all comes down to whether the DPS/ec of Judgement is more than the loss of the DPS/ec from pushing back CS. It is entirely possible this could be the case in 3.2 given Judgement's damage. Further testing will be required.

cannadrys · 06/28/09, 9:39 PM

Originally Posted by HamSlammer

I'm personally not worried about mana in a raid situation (10man, possibly), as we're very mana positive. The CS change accounts for a -131 MP5 swing assuming an impossible rotation (all abilities done on CD). In a realistic situation with CS having an effective CD of 4.5 seconds, it's only a -90 MP5 swing.

To me, the big issue about mana is PvP. We're losing Glyph of Seal of Blood, which is nigh mandatory in some setups/against some comps, and a weaker initial Judgement, making absorbs even more detrimental to our staying power in PvP. And the only solution Blizzard has provided me is the Glyph of Seal of Command, which helps the Seal no sane Paladin would use.

I would agree with the above, confirmed by my own testing which I though I would share here.

I did some testing by simulating raid buffed mana stats. I have an offspec holy gear set. I subbed out some pieces of gear - making sure not to pick up any MP5, while still maintaining hit+expertise caps (as to not miss any JoW procs). Since the holy gear has Int. on it - I got my mana to 7883, which is within about 50 mana of what I have fully raid buffed. I used untalented Blessing of Wisdom on myself, and judged wisdom, which would be the mana regen buffs you would have in a raid.

In my testing on a dummy I was using a CS>J>DS>CONS>EXO rotation, working in divine plea on cooldown. After 8 minutes, I had ~56% of my mana left. Glyphs used were Consecration, Judgement, Exorcosm.

I'm not at all worried about mana regen in a 25 man raid. In a 10 man raid where you may be missing buffs, I woudl suspect mana to be a probable concern.

• frmorrison · 06/28/09, 11:11 PM

Originally Posted by path411

Also, it's worth mentioning warriors/druids typically don't fully talent imp demo shout/roar. And I'm pretty sure 2/2 vindication is equivalent to the fully talented debuff. (I think untalented to talented is in the range of an extra ~5% damage reduction from the boss).

Correct, most Ferals/Warriors do not spec imp Demo Shout and since Vindication is equal to fully talented version, it adds a lot of utility.

Originally Posted by flyingtoastr

Judgement can be higher prioritized than CS easily. It all comes down to whether the DPS/ec of Judgement is more than the loss of the DPS/ec from pushing back CS. It is entirely possible this could be the case in 3.2 given Judgement's damage. Further testing will be required.

Worrying too much about 3.2 Priorities is a little premature. 3.2 is likely will come out in September, so there may be more Paladins changes coming in the coming months on the PTR.

Glutton · 06/29/09, 1:14 AM

Originally Posted by path411

It isn't a guarantee that prot will be grabbing it, as it requires them to drop an additional 2 points out of conviction for it. (Though it is definitely worth grabbing if it's not being covered by someone else) Though it will be something to work out, per guild basis, depending on how many ret/prot paladins you typically bring.

Also, it's worth mentioning warriors/druids typically don't fully talent imp demo shout/roar. And I'm pretty sure 2/2 vindication is equivalent to the fully talented debuff. (I think untalented to talented is in the range of an extra ~5% damage reduction from the boss).

Protection Paladins pick so many DPS/threat talents these days because there's a deficit of mitigation oriented talents. The excess threat we produce isn't doing a whole lot of good beyond giving us some extra padding to use our band aid cooldown (Salvation). I for one will have no qualms about dropping 2% crit for the ability to lower the boss's DPS by ~15%.

For fights such as Iron Council, it's quite useful to have multiple people (especially the tank who's holding a mob) who can put up a Demoralizing Shout equivalent. Remember, 3.2 has already hinted at not one, but TWO Jormungars (!).

jere · 06/29/09, 8:51 AM

Just remember that not every protection paladin runs a 0/53/18 or similar. Some do put all their extra points in holy (for LoH and AM) or fill up the prot tree more. Not everyone specs the optimal threat specs. For those that do, vindication will be an easy pick up with minimal threat loss. It won't be tough for me as I have 0/53/18, but many I know do not go with that spec.

Malleus · 06/29/09, 8:55 AM

Originally Posted by Chicken

Alternatively, if you start off with Judgement because you don't start off right next to whatever you're fighting (Which is pretty likely), we end up with this:

If we don't start off right next to whatever we're fighting, why would we not start with Exorcism? It has a longer range, so can be cast sooner, and as we'll get the mana back from the Judgement that follows it's essentially a clearcast.

• Chicken · 06/29/09, 8:57 AM

Originally Posted by Malleus

If we don't start off right next to whatever we're fighting, why would we not start with Exorcism? It has a longer range, so can be cast sooner, and as we'll get the mana back from the Judgement that follows it's essentially a clearcast.

I assume you want to move to whatever you're attacking, with the new cast time on Exorcism, you're essentially delaying your time before you're next to the target by 1.5 seconds. Exception would be if the mob is being pulled towards your position I suppose.

Naughton · 06/29/09, 9:49 AM

Did a Naxx25 run last night on the PTR. Was in full Conq gear, T1 Furious Wep and libram, and bout half BIS offset items. Noticed that Judge of Veng was hitting very hard. Crits were regularly in the 10k-15k range. Was going to post a WWS parse of it but somehow the combat log got corrupt and is unusable now. I used the FCFS rotation, with CS at the top then Judge and DS. Mana really wasn't an issue although I will say it did feel like i wasn't using consecrate as often as I do on live. This might be to explain my lack of mana issues.

Exemplar · 06/29/09, 10:19 AM

Originally Posted by Arikah

I'm curious about what is making the most impact on our dps, so bear with my odd requests...

Exemplar/Endo/Redcape, can one of you try modeling CS as it is on live (6s version) and see how much dps the seal changes are doing by themself?

Sorry - don't really read EJ on weekends (too busy raiding, mostly read at work).

All numbers are Patchwerk-style-maximums in my current gear with raid buffs per spreadsheet modeling:
- (live) CS(6s) + (old)SoB = 5777
- (ptr) CS(4s) + (new)SoV = 6150
- (theory) CS(6s) + (new)SoV = 5779
- (theory) CS(6s) + (new)SoV + Exo(6s) = 6647

On the last I actually roughed Exo in as 7 seconds, not 6, to account for some cooldown clashes, etc. CS is roughed at 4.5 sec. I have yet to run python scripts to get true no-lag effective cooldowns. These numbers are probably plus or minus 100 DPS. If/when Endo replies expect more accurate information (since Rawr internally calculates effective cooldowns).

So faster CS is indeed a DPS buff. However, I have assumed we have not reduced Consecration up-time and still spam literally everything on CD. If we have to reduce Cons due to a mana issue, it could be quite detrimental.

A CD reset on Exo would be a huge DPS boost. I really do not see this happening. It'd also cause GJ (Global Cooldown Jam) per Redcape's explanations and as you can see from Chicken's rough-up - free GCD are few and far between. Using Divine Plea is a DPS "loss" - although not, since we'll go OOM without it which would be far worse.

Personally - haven't tested on PTR, but I'm unhappy. It's more spammage, not less. The only thought is up-front. We'll theorycraft once, get best priorities, sit until a future patch change. AoW resetting Exo would only change our initial priority setup, raiding would still be whack-a-mole. Looking forward eagerly to whatever Blizzard changes next to see if it will be less spam-buttons-on-cooldown and more requiring player consideration.

I have a question for those on PTR. Can you drum up at least two paladins and both Judge Light then both Judge Wisdom off and on for a while. Do they:
A) always stack?
B) never stack?
C) occasionally stack?

Want to compare/contrast. Multi-JoW or JoL procs could still be useful. Likely best to test on something like a Blasted Lands mob - the Dummies appear to give unusual results based on what everyone says on live.

• Zurm · 06/29/09, 10:42 AM

Originally Posted by Exemplar

I have a question for those on PTR. Can you drum up at least two paladins and both Judge Light then both Judge Wisdom off and on for a while. Do they:
A) always stack?
B) never stack?
C) occasionally stack?

I can check these with Endoscient tonight if no one does before then.

Along these lines, I also wanted to bring up the topic of judging in 3.2. It's accepted practice on live to have the ret paladin judge light, and the holy/prot to deal with wisdom. With the nerf/change to JoL in 3.2, I've been thinking it would be more beneficial for us to go back to JoW, and let the healer types decide if JoL is appropriate for the given fight. This ensures 100% uptime for DPSers who have potential mana issues (hunters and us, primarily) and properly seperates healing and utility between the specs. The only real downside is not being able to top two meters at once!

SiliconSeraph · 06/29/09, 1:01 PM

Originally Posted by Zurm

I can check these with Endoscient tonight if no one does before then.

Along these lines, I also wanted to bring up the topic of judging in 3.2. It's accepted practice on live to have the ret paladin judge light, and the holy/prot to deal with wisdom. With the nerf/change to JoL in 3.2, I've been thinking it would be more beneficial for us to go back to JoW, and let the healer types decide if JoL is appropriate for the given fight. This ensures 100% uptime for DPSers who have potential mana issues (hunters and us, primarily) and properly seperates healing and utility between the specs. The only real downside is not being able to top two meters at once!

I'm not sure that I'd agree. If you're running with a prot pally then sure six in one, half dozen in the other when all Light judgments are equal, but in the case of ret and holy wouldn't it still be generally better to favor light for the 100% up time?

I'm sure this will be influenced by a variety of factors (Replenishment changes, caster DPS vs melee DPS, AOE content, etc) and these can be delt with on a situational basis but even after all judgments are equal, the HPS from JoL stands a solid shot at still being a more valuable choice. Guess we'll need to look at the actual content to make a real call. It would be interesting to see some encounters that make the most of a situational decision for which to favor for higher up time.

Raanis · 06/29/09, 1:54 PM

Originally Posted by SiliconSeraph

I'm not sure that I'd agree. If you're running with a prot pally then sure six in one, half dozen in the other when all Light judgments are equal, but in the case of ret and holy wouldn't it still be generally better to favor light for the 100% up time?

I'm sure this will be influenced by a variety of factors (Replenishment changes, caster DPS vs melee DPS, AOE content, etc) and these can be delt with on a situational basis but even after all judgments are equal, the HPS from JoL stands a solid shot at still being a more valuable choice. Guess we'll need to look at the actual content to make a real call. It would be interesting to see some encounters that make the most of a situational decision for which to favor for higher up time.

It would probably be most dependent on your raid setup and the prevalence of AoE damage in CC (HELLO Ulduar), however I would agree with Zurm on keeping up JoW. I normally raid with anywhere from 1 to 4 other paladins (one prot, 2 holy, one ret), but only 2 of them are truly consistent enough to count upon for daily raiding (prot and holy). On nights where the prot pally isn't there, a boss will only have JoL up 100% due to how much healing it triggers, despite any potential mana issues on my own part- just pop a potion and it's more than enough to last the rest of the fight.

However, due to the change to JoL on the PTR, it will be healing for less than half what it currently heals for when cast by a geared ret, assuming a 28k health raid buffed attacker (which only one or two of my melee have hit due to lack thus far of 25 hard mode gear). Based on my own personal observations in the raids I lead and the changes to replenishment on the PTR, it would probably be more beneficial to maintain JoW on the boss rather than JoL. Obviously on AoE heavy fights like Mimiron P2/P4 and on oddball fights like Vezax, it would be better to keep light up to help with the healing load a bit, but JoL will most likely be taking a secondary priority for judgement buffs.

Again, all of this comes from my own raid leading experience, personal results may vary, and as always the PTR is subject to change. Above and beyond though at this time on the PTR the only distinction between who should be casting JoL when 100% uptime of two judgements is possible would be whoever has the most ranks in Divinity.

• Chicken · 06/29/09, 2:30 PM

Originally Posted by Raanis

Above and beyond though at this time on the PTR the only distinction between who should be casting JoL when 100% uptime of two judgements is possible would be whoever has the most ranks in Divinity.

Only the healing received part of the Divinity talent is applying to Judgement of Light on the PTR. Not the healing done part, so how much Divinity you have specced only affects the healing on yourself, not the healing on others.