But shouldnt it be calculated like:
1.5 / (1.06 * 1.05 * 1.03 * 1.25) = 1.043, so haste from gear should only be 4.3%
Yes, haste effects are multiplicative with your haste rating being one of the multiplicative factors -- that's the way I remember it. I don't always have WoA and boomkin aura, and I'm sure others are in the same boat. Still, with just your own 6% haste from enlightenment and borrowed time, you hit the soft cap with 1.5 / (1.06 * 1.25) = 1.132, or 13.2% haste from rating.
But shouldnt it be calculated like:
1.5 / (1.06 * 1.05 * 1.03 * 1.25) = 1.043, so haste from gear should only be 4.3%
From - Casting speed - WoWWiki - Your guide to the World of Warcraft
Note: Haste rating stacks additively with itself but haste stacks multiplicatively. That means that if you have 158 haste rating, you will have +10% haste, no matter how many sources and items that haste rating comes from. If you then use troll Berserking for +30% haste, you would have 110% * 130% = 143% haste.
That is indeed correct. This seems to cause haste to be a lot less valuable for most raiding disc priests as 4.3% is very easely reached, I think this would need us to have two weightings for haste, one when the soft cap has not been reached and another one for when it has.
Well, you have to differentiate between tankhealing and raidhealing (or maybe that should be called raidshielding these days). For tankhealing Borrowed Time will probably not always be aktiv as you would have to cast additional "random" shields into the raid and depending on how hard the tank get's hit it's not always possible or worthwhile to do that. So a little more haste is probably still good in this situation. On the other hand, if you are mainly shielding the raid, both crit and haste are only useful for the penance/prayer of healing between shields. But on most items you have to choose either haste or crit anyway. So in conclusion even a 4,3% haste cap with borrowed time still doesn't really mean that we will should mainly stack crit. Because crit is just as mediocre in situations where you use PW:S alot and subsequently have a high uptime of Borrowed Time.
Well, you have to differentiate between tankhealing and raidhealing (or maybe that should be called raidshielding these days). For tankhealing Borrowed Time will probably not always be aktiv as you would have to cast additional "random" shields into the raid and depending on how hard the tank get's hit it's not always possible or worthwhile to do that. So a little more haste is probably still good in this situation. On the other hand, if you are mainly shielding the raid, both crit and haste are only useful for the penance/prayer of healing between shields. But on most items you have to choose either haste or crit anyway. So in conclusion even a 4,3% haste cap with borrowed time still doesn't really mean that we will should mainly stack crit. Because crit is just as mediocre in situations where you use PW:S alot and subsequently have a high uptime of Borrowed Time.
The problem is that haste above 4.3% is also worthless when you are spamming PW:S. The only solution I can see is stacking spellpower instead.
Are we entirely certain of this? I'm fairly sure that with 12% haste from gear my flash heal cast time is something like 0.97s with Borrowed Time, 3/3 Enlightenment, WoA and the boomkin/ret 3% haste. I can't check this right now as I'm at work and don't have a raid group but I've spent a lot of time worrying about the haste cap so I don't think I'm wrong.
Whatever the maths says, testing in game is always more effective.
Last edited by Silkath : 04/30/09 at 7:25 AM.
Reason: Failing at English
The problem is that haste above 4.3% is also worthless when you are spamming PW:S. The only solution I can see is stacking spellpower instead.
The problem is, that's not really a solution (except for enchanting / socketing). Every (healer) item except some trinkets has spellpower on it anyway. You get to choose between two of crit, haste, spirit and mp5 with the combination of spirit and mp5 being very seldom. So with the itemisation right now it comes down to either stacking crit, stacking haste or balancing the two stats somewhat.
Are we entirely certain of this? I'm fairly sure that with 12% haste from gear my flash heal cast time is something like 0.97s with Borrowed Time, 3/3 Enlightenment, WoA and the boomkin/ret 3% haste. I can't check this right now as I'm at work and don't have a raid group but I've spent a lot of time worrying about the haste cap so I don't think I'm wrong.
Whatever the maths says, testing in game is always more effective.
I am not sure either. I just linked a similar thread from a boomkin forum about haste. Can anyone confirm the above math about haste is the same ingame or not? Maybe ingame tooltip of casttime is not correct, who knows?
And I agree with you Glasswizard. T8 gear is a good mix between crit and haste also, so maybe they want us to mix it.
1 second is the global cooldown limit. You can still cast spell quicker than 1 second (for istance if you have enough haste to come down 1 second, you can cast them during Malygos vortex like an istant, at least from what I've read). The only limit is chain casting. Basically, suppose that you have enough haste to cast flash heal in 0.7 seconds, you must still wait 0.3 second before casting the next spell.
I scanned this thread since today will be my first day to really need my Disc-2nd-Specc in Ulduar. I stumbled over some things while reading.
About thread-issues of PS on tanks:
I'm a little bit crazy and tank in one raid with my warrior and heal in another with my priest. As a tank I cannot think about a single boss I've seen in Ulduar, where PS would be a problem on a tank. Given that it would not be applied at or shortly after the pull. Since there will be other CDs available (at least the tank will have at least one) it will not be needed on the pull so that should never be the case anyway.
A short note on today bosses: Unlike BC most bosses are tauntable. So even if a DD would pull aggro because of PS (or would come near to the 110/130%-mark the tank can simply taunt the boss so it should never even turn around. PS to safe a tank is possible at least if there are not 5 priests chain-PSing the tank.
About the theorie of stacking SP instead of crit:
Glasswizard mentioned something there. The same thing is discussed in the "WotLK Healing Compendium v3.1: Here There Be Theorycraft (and Dargons?)"-thread for holys.
You cannot do it with the given gear. SP is something like a fixed value for a given item-lvl on your gear (with the exception of trinkets and the decision between staff or mainhand-offhand-combo). The same goes for int. The only points where you can make a decision between SP and crit are gems, food and elexier/flask. And sometimes even there you may want to take a socket-bonus or maintain a meta. The difference for those options is somewhere between 100 and 300 points. Most of that decisions you would make anyway - so they are no gearing-decisions anyway.
So the gearing decision comes down to which 2 stats you would like to get: crit, haste, spirit or MP5 (or hit). So the question should not be if you would like more SP over more crit, or if you should get lost of something of your 30% crit to pick up more SP. The question should be if you would like to trade some crit for haste or spirit or MP5. Since spirit and MP5 only help you to get more mana they will not be the answer if you are looking for more throughput.
Sometimes you find both haste and crit on one item, in this cases you cannot trade crit for anything considering throughput. In all other cases the decision is between crit and haste - not crit and SP. Most of you may have reached the haste-softcap anyway. So more haste would help less than more crit, even if you already got 30% (or whatever) and use shield alot. (Especially then, because of BT.)
About role of Disc:
The comparison to druids is valid up to some point. When they do tank-healing most times they also do some raid-healing. (Or at least did it in BC - I'm not an expert on druids.) Because holy-paladins struggle to raid heal in many cases you do not need to define a tank healer as someone who can only heal one (or two) targets and ignores everybody else. Disc priests are good tank healers. The argument that sometimes (or given your raid-setup most times) they will be more useful if they concentrate more on the raid that don't make them less tank-healer.
Discs can be very valuable on the raid in some cases. But if they are the only ones on the raid they would have a much harder time than a holy priest or a shaman (or maybe a druid - again I'm not an expert on them) in most encounters.
So while their mitigation on the raid may make a huge difference they are not really raid healers.
Since I'm mostly holy I may have mixed something up. Please correct me there.
Glyph of Penance: Now increases critical strike chance by 5% instead of its old effect.
How big of a nerf/buff/nothing is that going to be? I'm not good with the maths, so I can't figure out if the increase in DA procs will be enough to counter the pure throughput.
1 second is the global cooldown limit. You can still cast spell quicker than 1 second (for istance if you have enough haste to come down 1 second, you can cast them during Malygos vortex like an istant, at least from what I've read). The only limit is chain casting. Basically, suppose that you have enough haste to cast flash heal in 0.7 seconds, you must still wait 0.3 second before casting the next spell.
While this is true, it is also less of a factor for healers than DPS people. Someone in a DPS role is going to be hindered by the GCD limit much more than someone in a healing role. As a healer, you are not always chain casting, especially as a priest. Furthermore, while there is no benefit to say, pumping out Wrath as a druid in 0.7 seconds, there IS a benefit in pumping out a Flash Heal in 0.7 seconds as opposed to 1 second. Faster heals are always more likely to save someone's life, and they allow you to heal more reactively (thus saving mana).
Glyph of Penance: Now increases critical strike chance by 5% instead of its old effect.
* Divine Hymn: Healing and healing scaling reduced by 30%. Buff on affected players changed from 15% to 10%.
* Renewed Hope: Effect can no longer be dispelled.
* Soul Warding: Mana cost reduction is now 15% down from 30%.
Glyph of Penance: Now increases critical strike chance by 5% instead of its old effect.
How big of a nerf/buff/nothing is that going to be? I'm not good with the maths, so I can't figure out if the increase in DA procs will be enough to counter the pure throughput.
This is a back-of-envelope calculation, and I'm sure there are people here who can correct me if I'm wrong. I am leaving some things out, like HPS from spells cast between Penance casts and the extra healing from an actual Penance tick critting. But this should give a rough idea of what this change is going to do.
There's really no way that this is a buff. 5% crit increases the chance that each Penance tick will crit by 5%. This doesn't affect any other spells. Say your Penance is ticking for 4000 (a nice round number, might not be accurate, but I'll use the same number for both.) Each tick has an extra 5% chance to crit. Say you start with a base crit of 25%. This will give you a 73.4% chance of gaining at least 1 DA proc per Penance cast. Each DA proc is 30% of the damage healed. So for each penance cast you have a 73.4% chance of getting a 1200 health DA proc.
Adding an extra 5%, you get an 87% chance of getting a DA proc on a Penance cast. This is an improvement of 13%. And as this only works on Penance, with the regular cooldown of 8 seconds, you have an extra 14.3% chance of getting 1200 healing every 8 seconds. Let's assume that you will always get an extra DA proc 14.3% of the time. If we do this, you end up with effectively 21.45 HPS added by Glyph of Penance. I'll be charitable and call it 25 HPS.
Glyph of Penance, in its current form, will let you cast Penance 2 seconds earlier, shortening the cooldown to 6 seconds. With an 8 second cooldown, Penance is about 1500 HPS (with 4000 point ticks.) With a cooldown of 6 seconds, Penance is about 2000 HPS. (Both not considering DA procs.)
Decreasing the cooldown by 2 seconds thus yields you about 500 HPS on average for the contrived numbers I've picked.
I'd say 500 HPS >> 25 HPS. Glyph of Penance increases throughput roughly 20x more in its current form than it does on the PTR. This is a huge nerf. And probably not at all worth a major glyph slot if the change goes through. I'd rather glyph PoH or Holy Nova, to be honest.
EDIT: Even if we include the extra HPS from the extra 14.5% chance of Penance crits, this comes out to about 36 HPS, so total this ends up being about 62 HPS by glyphing the new Penance glyph. Still over 8x less than the one on Live, and it's probability based whereas simply shortening the cooldown is a guaranteed increase. I'm becoming more sad the more I think about this.
EDIT2: Corrected a math error which propagated through to other calculations.
This is a back-of-envelope calculation, and I'm sure there are people here who can correct me if I'm wrong. I am leaving some things out, like HPS from spells cast between Penance casts and the extra healing from an actual Penance tick critting. But this should give a rough idea of what this change is going to do.
There's really no way that this is a buff. 5% crit increases the chance that each Penance tick will crit by 5%. This doesn't affect any other spells. Say your Penance is ticking for 4000 (a nice round number, might not be accurate, but I'll use the same number for both.) Each tick has an extra 5% chance to crit. Say you start with a base crit of 25%. This will give you a 73.4% chance of gaining at least 1 DA proc per Penance cast. Each DA proc is 30% of the damage healed. So for each penance cast you have a 73.4% chance of getting a 1200 health DA proc.
Adding an extra 5%, you get an 87% chance of getting a DA proc on a Penance cast. This is an improvement of 12%. And as this only works on Penance, with the regular cooldown of 8 seconds, you have an extra 12% chance of getting 1200 healing every 8 seconds. Let's assume that you will always get an extra DA proc 12% of the time. If we do this, you end up with effectively 18 HPS added by Glyph of Penance. I'll be charitable and call it 20 HPS.
Glyph of Penance, in its current form, will let you cast Penance 2 seconds earlier, shortening the cooldown to 6 seconds. With an 8 second cooldown, Penance is about 1500 HPS (with 4000 point ticks.) With a cooldown of 6 seconds, Penance is about 2000 HPS. (Both not considering DA procs.)
Decreasing the cooldown by 2 seconds thus yields you about 500 HPS on average for the contrived numbers I've picked.
I'd say 500 HPS >> 20 HPS. Glyph of Penance increases throughput roughly 20x more in its current form than it does on the PTR. This is a huge nerf.
I agree with everything you said, including the numbers you posted. However, as a hypothetical situation, if you start getting overgeared for Ulduar, (and so does the tank you are healing), the need for casting Penance reduces inversely to the tank's gear level. If we assume this, will there at one point be a cap at which your target's consistency of damage/amount damage will let it be better to have 5% extra crit chance than reducing your cooldown of Penance by 2 seconds? For sake of discussion, we can say that at n amount of parry/dodge/block and armor, the need to use Penance right when it's available is at about 20%. Can anyone figure out if this will then mean that it is now better to use the new glyph?
One thing I noticed is that it doesn't seem to take slot bonuses into account - where there is a yellow slot with a +7 SP bonus, it still recommends a 19 SP gem instead of a 9 SP/8 Crit gem (or similar), so take the recommended gems with a grain of salt.
I know you said it was back-of-envelope but that maths looks pretty worthless to me. The current Glyph is not a 500HPS increase (even assuming your Penance value is correct), firstly because it doesn't take 2sec off but rather 1.6 given that everyone has it talented, and secondly because that 500HPS increase assumes you're standing around doing nothing for the rest of the fight. You need to look at how much of an increase in HPS it is over the real alternative which is casting FH or GH (or whatever).
Additionally the measurement of "chance of getting a DA" isn't a relevant way to calculate HPS. If you have 5% more crit chance that's 5% more chance to get a DA of a certain size per tick, and that gives you the HPS. It's not an x% chance to get one-or-more procs which you then count as always being one proc.
Not that I'm that arguing the new glyph is better (or worse), just that this maths seems far too loose.
I agree with everything you said, including the numbers you posted. However, as a hypothetical situation, if you start getting overgeared for Ulduar, (and so does the tank you are healing), the need for casting Penance reduces inversely to the tank's gear level. If we assume this, will there at one point be a cap at which your target's consistency of damage/amount damage will let it be better to have 5% extra crit chance than reducing your cooldown of Penance by 2 seconds? For sake of discussion, we can say that at n amount of parry/dodge/block and armor, the need to use Penance right when it's available is at about 20%. Can anyone figure out if this will then mean that it is now better to use the new glyph?
I don't see why a spell should be more useful when you overgear the raid instance and I doubt anyone would care if that was the case. Penance is the best heal for Discipline priests and nerfing it is not the way to go considering how low our HPS is already. I'm assuming this was done with PvP reasoning so at least we can hope for a way to get back that throughput before the PTR patch goes live. We should wait for an answer from Ghostcrawler for now.
As a sidenote has anyone noticed high Glyph of PW:S crits on protection paladins? I think there's a conflict/misbehaviour with Touched by the Light.
I know you said it was back-of-envelope but that maths looks pretty worthless to me. The current Glyph is not a 500HPS increase (even assuming your Penance value is correct), firstly because it doesn't take 2sec off but rather 1.6 given that everyone has it talented, and secondly because that 500HPS increase assumes you're standing around doing nothing for the rest of the fight. You need to look at how much of an increase in HPS it is over the real alternative which is casting FH or GH (or whatever).
Additionally the measurement of "chance of getting a DA" isn't a relevant way to calculate HPS. If you have 5% more crit chance that's 5% more chance to get a DA of a certain size per tick, and that gives you the HPS. It's not an x% chance to get one-or-more procs which you then count as always being one proc.
Not that I'm that arguing the new glyph is better (or worse), just that this maths seems far too loose.
I'd love to see an improved calculation, I don't really have time for one at the moment. I was really simply trying to show that it's indeed a nerf, as it clearly is.
Maybe if I get more time later I'll fiddle around with it, but maybe somebody's already done a similar calculation and would be willing to provide a more accurate picture than I'm currently able to.
Assuming 30% raid buffed crit, 5% more crit is 3.7% more healing. 35% and it drops to 3.5%.
This might be the worst glyph for a healing spell in the entire game. I certainly can't think of a glyph that does less offhand.
I'm leaning towards Prayer of healing myself.
P.S. Envelope math says Penance is roughly 7000 HPS, Flash Heal roughly 5000 HPS. You can easily do envelope on 1.4*2000 versus 3.5% of 7000.
In actual fact healing is not about throughput. It's about ability to respond to spikes. This means the best heal will be down more often than it would otherwise be down. In Ulduar? You really need it.
The glyph will give us around a 3-4% increase on Penance's average throughput per cast.
(5%*0.95/(1.30))=3.6% for a 30% crit rating.
This is pretty bad, and is far weaker than the current glyph. A direct numerical comparison is not particularly useful due to the different natures of the two versions.
Well, the calculation might not be worth making at this point, but I calculated the probability of one of the penance ticks critting incorrectly. They're all independent, so using the correct formula gives, for 25% crit, a 57% chance of at least one tick critting whereas for 30% crit I get a 65.7% chance of DA proc'ing. This is a difference of 9%. For what it's worth, this is worse than what I calculated above.
Maybe a numerical comparison isn't really useful, but I thought I'd attempt it anyway. How else can we quantify how "bad" the nerf is?
If this glyph change go through it will hurt the ability to handle spikes on tanks in a pretty big way. And with HPS being significantly lower than that of holy paladins the disc priests needed that ability to lessen spikes. It might push disc into the role of mostly PW:S spammer than tank healer.
I do think the changes are a knee jerk reaction to the complaints of rogue+priest success in the early days of Arena season 6 and hope with some perspective they will give it some more thought.
If this change goes through I would switch to Glyph of PoH. Even as disc priest I get plenty of opportunity to use PoH in Ulduar and it would give me pretty good value.
I calculated out just how much of a nerf this is based on my gear (~2166 Spellpower)
First, here is a declaration of sorts.
Currently, Penance has a 6.4 sec CD with the glyph
10 CD
-2sec from glyph
8sec * .8 (20% reduction from aspiration)
6.4sec CD
The change of the glyph from -2sec CD to +5% crit will make it have an 8 sec CD
10 CD
* .8
8sec CD.
We will use this to determine a comparable frame of time. In particular, the Least Common Denomonator of the time is 32 SEC.
In this 32 seconds, we can ....
Get off 5 Penance casts with the -2sec glyph
Get off 4 Penance casts with the +5% crit glyph.
Each Penance cast has 3 BOLTS. (15 bolts w/ -2sec glyph, 12 bolts with +5% crit glyph)
Using my spellpower and recount over a period of time, I found my heal amount: average HEAL 3390
average CRIT HEAL. 5060
The values for crit% and non crit in these tests are. Crit--------non-crit
30%-------70%
35%------65%
40%------60%
45%------55%
50%------50%
55%------45%
The formula used to get the HEAL amounts was
(Penance BOLTS * normal heal amount * non-crit%)+(penance bolts * crit heal amount * crit%)
The formula used to get the Divine Aegis amounts was
(penance bolts * crit% * crit heal amount * .3 )
.3 being how much Divine Aegis shields for out of a crit heal.
These 2 numbers are added together to get a net total
(Heal amount)+(Divine Aegis amount)
Here is the data for healing over 32 sec with -2sec glyph(15 bolts). Crit----non-crit----Heal amount---Divine Aegis amt---Total
30%--------70%--------58365--------6831--------65196
35%-------65%--------59618--------7970--------67587
40%-------60%--------60870--------9108--------69978
45%-------55%--------62123--------10247--------72369
50%-------50%--------63375--------11385--------74760
55%-------45%--------64628--------12524--------77151
Here is the data for healing over 32sec with +5% crit glyph (12bolts) Crit----non-crit----Heal amount---Divine Aegis amt---Total
30%-------70%-------46692-------5464-------52156
35%-------65%-------47694-------6375-------54069
40%------- 60%-------48696-------7286-------55982
45% -------55%-------49698-------8197-------57895
50% -------50%-------50700-------9108-------59808
55% -------45%-------51702-------10018-------61720
And here is the difference between the 2 as a percent of change of -2sec glyph.
The Heal, Divine Aegis and Total values of 30% CRIT, with -2sec glyph, are being compared to the values of the 35% CRIT with the +5% glyph. 35% to 40% etc.
CHANGE IN VALUES OVER 32 SEC. Crit compared------Heal amtt--------Heal amt %--------DA amtt-------DA %--------Totalt-------Total %
30% to 35%.........-10671.............-18%.............-1594...............-23%.........-13517.........-21%
35% to 40%........-10922.............-18%.. ..........-1822.............-23%..........-13996..........-21%
40% to 45%........-11172.............-18%..............-2049.............-23%.........-14474.........-21%
45% to 50%........-11423.............-18%..............-2277.............-23%.........-14952.........-21%
50% to 55%........-11673.............-18%..............-2505.............-23%.........-15430.........-21%
Based on my 2166 Spellpower.
The Change in the glyph will net the following over 32 sec.
a -18% margin of HEALING
a -23% margin of DIVINE AEGIS
a -21% margin in total.
For the hell of it, I calculated out just how much crit the glyph would have to add in order for it to be equal in terms of output of the -2sec glyph (over 32 sec).
Healing over 32 sec without -2sec glyph: Crit---non-crit---Heal amount---DA amt--Total
70%.......30%.......54708.......12751.2.......67459.2
So, in conclusion:
1) The change of the Penance glyph from -2sec to +5% is a definite NERF. About 20% to be exact.
2) In order for it to be as good as it currently is, it would have to add 40% crit to penance.
I wouldn't mind seeing a glyph that increases the crit % of Penance by 40%. This would add more to the whole Disc priest thing of shielding. However, leaving it as it is...+5% crit...The glyph inst even worth getting anymore.
Iwachiten your calculation assumes that you are not casting any spells for those 2 seconds (or less due to haste) instead of casting penance. Now it is still going to be a pretty big nerf, but it is not a full 21%.
Iwachiten your calculation assumes that you are not casting any spells for those 2 seconds (or less due to haste) instead of casting penance. Now it is still going to be a pretty big nerf, but it is not a full 21%.
Correct.
This assumes using Penance "at every CD", timing of spells aside.
This is strictly looking at Penance. Calculating haste, borrowed time, etc would be a huge undertaking because the order of spells as well as haste levels vary widely from player to player.
However, the missing glyph makes the timing of spells effectively SLIDE OVER 1.6 sec. Which is a little above the time needed to cast a Flash Heal or PW:S.
So instead of casting Penance for a truck load of heals and shields for 618 mana, we will be casting the much weaker Flash Heal (521man), or now expensive PW:S (666 mana....assuming no weakened soul is on target already).
04/29/09, 8:29 PM
Similar Threads
