Glyph of Divine Storm - Page 6

gmedina · 10/09/09, 1:28 PM

Originally Posted by Zurm

Allow us, for a second, to analyze the fact that it is "always" easing the burden of healers by taking a look at this past week's kill of heroic Twin Valkyr 25, one of the higher AoE damage fights in the game, and there is consistent AoE. By choosing this encounter, I'm maximizing the effective healing of DS, so this is the exception rather than the rule.

Fight in question: Heroic Twin Valkyr 25

Duration - 4:08
Total effective healing done - 11959754
My DS healing - 35808 * (effective: 1 - 31.6 %) = 24492.7
My DS healing if I had the Glyph - 39188.3

Difference - 14695.6 healing
Additional contribution it would have made to raid healing is approximately 0.12%. And that's assuming that it didn't increase your percentage of overheal, which it almost certainly would.

Now let's look at my use of exoricsm, and see what it would have given. I was using Judge/Consec/SoV glyphs, so I can add the damage in.

Total damage done - 39081945
My exoricsm damage - 141109
With Glyph - 169330.8

Difference - 28221.8 damage
Additional DPS contribution - 0.07%.

Now obviously everyone will have differening opinions here. The order of magnitude involved is the same. For me, I'll take the 2.6x smaller percentage of the DPS contribtion since my raid role is what? That's right, DPS. And hey, look at the actual damage/healing amounts, damage is higher almost by a factor of 2! If the new 2pc makes us use DS twice as often, I will consider it broken, and would be greatly surprised if it isn't nerfed. That has many PvP liabilities.

Again, I picked a fight in your favor. I'm sure most other fights would end up even more in mine. I want to stress that while DS is a smart heal, there is a gap between target picking and actual healing, meaning it often times will actually overheal. I am NOT arguing that the DS glyph adds no value; rather, I am arguing that the value it adds is not worth the tradeoff, even of the exorcism glyph.

And this whole idea of the DS glyph is a flawed one in the first place because by wearing that much expertise is not ideal. You SHOULD be using most of the SoV glyph (unless you're a dwarf). Don't be an ilvl whore and just get upgrades beacuse you see a bigger STR number. The SoV glyph, when fully or mostly used, proves a very significant DPS increase; one that no other glyph other than Consec or Judgement can even hold a candle to.

Please, stop throwing these "but it could" or "it still helps" arguements at me. I'm showing you numbers as to why it just doesn't make sense. And regarding Fmorisson's post, you are arguing symantics. I was hoping for a lock ;P

So what your saying is that if a raid group has a choice between adding 2% more to healing at cost of 1% dps no raid should take that? Conversley if your priest could spec into a talent that would increase raid dps by 2% at a cost of 1% to healing, they should not take it?

So if we have a situation below

We can have one person increase healing by 2% at a cost of 1% dps.
We can have one person increase dps by 2% at the cost of 1% healing.

You would argue your are better off having each grab 1% more healing and 1% more dps?

Instead of having both grab the 2% benefit? Standard Economics would tell you your are being inefficient and all current understading of risk benefit anlysis would tell you that you are wrong.

• Zurm · 10/09/09, 1:54 PM

When the total amount of healing versus damage we are looking at is in an almost 1:4 ratio, yes, I would pick 1% more damage than 2% more healing. Again, while the % increases for healing and damage showed the DS glyph to be better, the actual amounts were greatly in favor of exorcism.

Originally Posted by gmedina

So if we have a situation below...

I'm confused as to what exactly you are arguing here. How does this situation relate to what we are arguing? When does us getting DS glyph influence a healer increasing the raid DPS other than smiting out of boredom (which a .12% increase will not cause)? Furthermore, healing is not the same as DPS. Healing involves a lot more waste, because you are hitting for more against targets with many times lower HP pools, so that .12% increase is an ideal extrapolation, not a realistic contribution like the DPS is more likely to be. In a typical raid, you have 6 healers and 17 DPS. Only one of those 17 DPS can overkill, and then only once (assuming a single target fight). All 6 healers can overheal (plus additional random hybrid heals from JoL, DS, and imp LOTP), and each of them, many, many times throughout the fight.

EDIT: Also, it's not really necessary to quote my whole post. That's a lot of wasted space and you're not addressing all my points!

Capstone · 10/09/09, 2:48 PM

Just as a quick rejoinder, I'd like to echo comments made in the main Retribution thread: don't assume everyone has the same access to upgrades as you. It's entirely possible that the only drops available to some folks (RNG or whatever) push them high enough on expertise to warrant dropping the glyph. And choosing 2% extra healing over 1% extra damage - while it wouldn't be my personal choice - doesn't make you a terrible player.

Redcape · 10/09/09, 3:23 PM

I think a couple of points are in order here:

1. Picking a fight against an single, undead target where your effective cooldown on DS was 15 seconds probably isn't a fair choice. Exorcism is substantially better against undead and DS is at its worst against a single target. Clearly you are never going to hit a 10 second cooldown, but I would generally expect to see more Divine Storms in that timespan. Certainly we expect to face a lot of undead in Icecrown (all?) but I don't think basing our benchmarks on a single undead target is a good way to get unbiased data.

(Note, this is me looking at the WOL and checking the whole fight length by the graph, it is entirely possible that the mob actually died a bit before that and the effective cooldown should be more like 14 s.)

2. Saying that the burden of proof is on the other side and defining an exceptionally narrow range of proof isn't reasonable. By your standards of proof you need to produce a fight where it is absolutely proven that without the extra damage on Exorcism the fight could not have been won. Both of these proofs would not be particularly relevant because finding them would be very random and not statistically convincing. Just because I can find 1 bear in the world that can dance gracefully and 1 human that cannot does not prove that bears are better dancers than humans.

This sort of proof would be far more dependent on the zeal of the searcher and the definition of the condition than the actual benefit of the effect being tested.

All that said, given current mechanics I am not suggesting that everyone who uses GoEx is wrong or bad. I merely suggest that GoDS is useful and worthy of consideration, and that I personally think it is more powerful from the point of view of maximizing the chance of the raid to defeat the boss. That said, it is exceedingly difficult to accurately determine the value of the glyph because of the enormity and complexity of a WOW raid combat and as such I cannot find fault with people who choose to go with GoEx; I think the waters here are sufficiently muddy to make both choices rational. I do think that refusing to discuss the topic or wishing away the thread is a mistake given that the whole point of these forums is to have informed debate.

If the current t10 bonuses go live the current situation with its (imo muddy waters) will be dramatically altered. The benefit of the GoDS will increase dramatically (doubling would not be surprising) and GoEx will drop equally dramatically. (halving would be similarly unsurprising) At that point I definitely expect GoDS to be flat out the better choice, monumentally so in a situation with consistent multiple targets for DS to affect.

gmedina · 10/09/09, 4:21 PM

Originally Posted by Zurm

When the total amount of healing versus damage we are looking at is in an almost 1:4 ratio, yes, I would pick 1% more damage than 2% more healing. Again, while the % increases for healing and damage showed the DS glyph to be better, the actual amounts were greatly in favor of exorcism.

I'm confused as to what exactly you are arguing here. How does this situation relate to what we are arguing? When does us getting DS glyph influence a healer increasing the raid DPS other than smiting out of boredom (which a .12% increase will not cause)?

I was attempting to show the relative value of the the glyph by showing it's counter point. So hypothetically speaking if you had a priest talent that could provide a debuff on the boss that would increase dps 2% at a cost of 1% healing, using your rules they would never take this. At the same time the Ret pally would not take the 2% healing increase which comes at a cost of 1% dps. So applying your rules each would say i need to maximize my single contribution regardless of other costs. Economics has already proven that maximising individual contribution is never the most efficient in any system.

The fact that priests don't have this ability doesn't invalidate the value of the Ret pallies it just more clearly shows you the domain issue.

• frmorrison · 10/09/09, 6:47 PM

Here are some of my numbers, from Jaxx-H :

Total damage: 2699963
Exo: 96732
With Glyph - 116078
0.7% damage increase

Divine Storm healing: 47862 with 17% overheal (I am pretty sure WoL shows the actual heal, the overheal is for reference).
w/glyph: 59828
Total Healing: 185021 (Holy was doing Light, I had wisdom but I did do a 71k raid wall and JoL did 143862).
6.5% healing increase
Assuming I was doing Light and no raid wall, it would get a 4.7% healing increase.

So your results depend on your healers somewhat.

Wrathblood · 10/09/09, 7:44 PM

Ash, so that means you gave up ~20k damage from the Exo Glyph to gain ~12k healing with the DS glyph, and it sounds like that's a trade you're happy with, but I don't think a lot of rets would take that trade.. While the % healing gained is larger, its because you're obviously dpsing rather than healing and thus have a much larger base of damage done to compare against.

IMO, we need to see what the proc data looks like from the PTR and what that does to our ability usage before we can figure out how strong a case can be made for DS vs Exo. Obviously relative value will vary dramatically from raid to raid, but I think its entirely possible Zurm will still be right, I just think he's jumping the gun.

Edit - appalling typos. posting from BB ftl.

flyingtoastr · 10/09/09, 7:51 PM

Originally Posted by Wrathblood

Ash, so that means you gave up ~20k damage from the Exo Glyph to gain ~12k healing with the DS glyph, and it sounds like that's a trade you're happy with, but I don't think a lot of rets would take that trade.

The real debate he is when the 2-piece T10 bonus hits. Even assuming a conservative proc chance (only on whites) you're still looking at pushing Exorcism back a lot and gaining a lot more Divine Storm damage, which has the effect of both devaluing the exorcism glyph further while also make the total healing from Divine Storm higher.

Jamora · 10/10/09, 6:30 AM

This discussion is kind of interesting. It is about healing vs. damage importance, role purity and many personal opinions. Since I am in the situation that I am at the expertise cap without the SoV Glyph I have to choose which glyph to use now until I can shed some expertise through upgrades, I wanted to compare the value of both.

The events that the heal from divine storm actually saves somebody from death is extremely rare, just as the event that only a well timed Exo prevented a wipe, so I will compare the average properties, in the form of a singe target buff.

Note: Vales taken from RAWR, from my priest and paladin.

Glyph of Exorcism: ~35 dps
This corresponds for me to a 50 AP buff.

I assume now that the heal from Divine Storm would replace a Flash Heal from my priest, which has a mana efficiency of 10.5 hpm.

Divine Storm: ~450 dps -> 112.5 hps -> 53.6 mp5
Glyph: additional 67.5 hps -> 32.1 mp5

Now, this conversion probably is not 100% efficient (I ignore overheal, the FH it replaces would on average overheal as well), so lets say it is only 70% of the original value.

So in terms of buffs, I have to decide whether a 50 AP buff to myself is more valuable than a 22 mp5 buff to a healer, which is kind of a wash. Neither is very impressive, and until we start hard modes next week I do not know how our healers will cope with the increased raid damage, and neither do I know how the DDs will fare with the increased damage requirements.

But at least I can compare 50 AP to 22 mp5(or more in case of multiple targets), while 0.12% increased raid healing is quite incomprehensible.

Of course, if the T10 set bonus greatly reduces the effective cooldown of DS and there is even more expertise on gear I acquire in IC, then the Glyph of DS looks more valuable.

Neraya · 10/12/09, 6:03 AM

Originally Posted by Zurm

So, here I propose a challenge. If someone can provide ONE example via combat logs where Glyph of Divine Storm (or the Divine Storm heal, for that matter) did or could have saved someone from dying, I will rest my case. If more than one is found, I'll even support the cause. I tried to find one, and couldn't. This does NOT mean find a case where someone died with <1k overkill or whatever your DS heals for. This means find an example of a timely DS heal that either saved someone's life or had it been marginally stronger would have. The logs would ideally be hosted on WoL or another major site... and don't try to doctor them, I'll know!

Finding evidence that the glyph did actually save someone somewhere at some time during a boss fight is i.m.o. futile. For one, the DS heal doesn't need to have happened a split second before the final kill. It could have been several seconds earlier (yes, with healers slacking and not healing that person up for a while). It could even be a case of someone receiving a DS heal, followed by a real heal or some HOT ticks not healing them to full followed by a big blow that just barely kills them so the GoDS would have saved them. Good luck plowing through the logs trying to find it

And even if you do find such a case, it still doesn't proove anything. Is it potentially possible for it to happen: Yes, definately. But for every case you can find where DS saved someone's life, I can probbaly find 2 or more '1% wipes' where the extra bit of damage from exo would have killed the boss.

It's wrong to try to find a case of "with the glyph, <boss ability> would not have killed someone". You can say the same about speccing 5/5 divinity for some bigger heals from JoL which a lot of ret paladins are doing. Show me a case where the extra healing from divinity would have not killed someone, chances are, you won't find one either, but that doesn't invalidate the merit of the talent.

Other than the occasional gimmick fight like vezzax, nobody really counts on stuff like heals happening at random from stuff like JoL, DS, and Vampiric Embrace. But they do help in any fight. Not necessarily by actually being the thing that saves someone from dying, but it takes some pressure off the healers, it gives them that little bit extra breathing room. It may be the just the thing that gives them time to heal person A that really needs it because thanks to DS/JOL/VE/... person B doesn't get that next heal but it goes to A first. It's this bit of choice offered to healers that makes theorizing avout the actual merit of a 'random heal' like DS impossible. DS can save someone's life not because how much it actually healed, but because it made healers prioritise someone else that would have died otherwise. No amount of log plowing will teach you if this happened.

GoDS definately has merit. And with T10 that merit gets better. Whether it'll be good enough to take it over Exo... I can't tell yet, I guess it'll depend on how many Exorcisms we'll still end up doing. But if it does end up in a situation where the amount of exorcisms get reduced to near nothing, GoDS is definately a decent replacement glyph for exo.
For the time being... in 3.2 I'll stick with the extra bit of DPS. There's too many DPS checks to skip over it. With T10... It'll depend on how many Exo's we end up doing over the course of a long fight and the fight mechanics.