5sr dancing has got a nasty hit, but it isn't completely dead in my opinion. Why we can't get out of 5sr that often with clearcasting anymore, but we still have inner focus and SoL and we get longer breaks in fights (supposedly).
The new holy concentration does not only increase i5sr mana regeneration but also oo5sr mana regeneration.
Imagine if you get a couple of seconds break , then cast a inner focus binding heal with a good chance to proc the 50% increased mana regeneration, which should bring our nerfed oo5sr mana regeneration to almost prenerfed levels.
We can only do this every 3 minutes with inner focus, but it is still worth doing, especially if mana gets scarce.
[e3] Here's the Iron Council attempts tonight. Had Distilled Wisdom up tonight, along with crit food, and was tank healing (as primary "nuke" healer) for the first part, then regen until the third part basically. Had a couple of stupid deaths: lag was brutal.
What were your glyph choices? Any thoughts of adding the GS glyph as one of your more permanent glyphs?
Has anyone found someone with Penance glyph on PvP yet btw? If you have I would love to get one made as thats where we have raided Ulduar the last few nights =\. I have been doing researhc on my priest with no luck yet and my herba didn't get xferred.
"Meditation and Improved Power: Word shield have changed locations with each other."
This change makes zero chance unless it is a mistake and they moving meditation to tier 2.
No... The change is just that... Meditation is on the left side of the tree, and Imp PW:S on the right side, because it is linked with Soul Warding (the new 21 pointer)
Regen nerf = all the different (3) ways our regen was nerfed. Just the spirit change alone was a solid chunk of Mp5, but IHC being changed was a dual nerf that ends up at least -100 Mp5, maybe more. Serendipity used to be 50-75 Mp5 for me, and it's gone too. Overall, I'm down at least 250 Mp5, which is roughly Replenishment. Imagine doing a 25-man boss fight without Replenishment on live, and you'll have a good estimate of how bad it is.
They went just a touch overboard. If they want us to have to make decisions, fine. But the fact that we have one cooldown on a ridiculous timer (Fiend), and no real way to regen mana when we do go OOM seems rather short-sighted. The combination of the spirit nerf and the overall fight length increase (lots of up-to-10-minute fights in Ulduar) means we didn't need the massive 3-scale nerfs. I'm hoping they realize this before they push 3.1 to Live.
Anyone who cannot cope with mathematics is not fully human. At best he is a tolerable subhuman who has learned to wear shoes, bathe, and not make messes in the house. - R.A. Heinlein
They went just a touch overboard. If they want us to have to make decisions, fine. But the fact that we have one cooldown on a ridiculous timer (Fiend), and no real way to regen mana when we do go OOM seems rather short-sighted. The combination of the spirit nerf and the overall fight length increase (lots of up-to-10-minute fights in Ulduar) means we didn't need the massive 3-scale nerfs. I'm hoping they realize this before they push 3.1 to Live.
I agree, Nidaba.
Went on the PTR yesterday and couldn't believe the change to our regen. Trying to do everything right, get a SoL proc, use my spirit trinket, use IF, then use binding heal... it was still very ruff.
Really liked the dual spec you put up the other day, were you went half way in both trees, I'm thinking of going that route if all the fights last 10 mins or more. Hopefully blizzard will see this and make some changes.
The formula would be: HC uptime = 1-(1-crit)^(n-1)
...
However, this model is a simplification, as it really doesn't find HC uptime, but the percentage of times after proc'ing a HC that it is renewed before it expires. Finding HC uptime would involve a more complex model that would have to take into account which of the n-1 casts refreshed HC and how much time in average it would take to proc HC once it has expired.
I think your model is a very accurate estimate, that applies to long fights. Here is how I approached the problem, ending up with the same formula.
Consider Holy Concentration (HC) as a Markov chain. The states the system can be are the time remaining on Holy Concentration. Suppose you cast a spell capable of triggering HC every 2 seconds (n = 5, in the above formula). Then we only need to track five states: 8 seconds left, 6 seconds left, 4 seconds left, 2 seconds left, HC inactive (0 seconds left).
Call these 4,3,2,1,0, respectively (seconds remaining divided by the seconds between casts).
Now we need an expression for each state that tells us the expected state in the next time step. For any state x > 0, we have:
,
where w is the probability the fight continues another 2 seconds and c is the probability of a critical strike. The analogous expression for state x = 0 is:
.
So now we have a system of 5 equations and we wish to solve for the expected count of HC uptime, given we start at state x = 0 (HC inactive). One can do this with highschool algebra, by solving V(4) = 1 + w(1-c)V(3) + wcV(4) for V(4), substituting the solution into V(3), solving V(3) for V(3) in the same way, and continuing down to V(0). What we end up with is an expression for the expected uptime, starting at x = 0. This is (pasting from my Mathematica notebook):
This is the expected number of 2 second intervals that HC is up. If we want instead expected *proportion* of uptime, we divide by the expected length of the fight, which is 1/(1-w)---this is a simple infinite geometric series, so some of you will recognize this closed form expectation. In any event, we now have an expression for the expected proportion of uptime, starting at x = 0 (again pasting from my Mathematica notebook):
If you let w -> 1, then this becomes (after much simplifying):
.
Huzzah. If w < 1, then the expected uptime is lower, because we start with HC down. But note that for w = 0.99, we expect a fight of 200 seconds, and then 1-(1-c)^4 is a nearly perfect approximation.
I wrote a small stochastic simulation to verify all this, which you can paste into R (which is free for download) and check it all out yourself.
simhc <- function( pc=0.25 , n=10000 , waitcast=2 ) {
dcast <- 0
d <- 0
hc <- {}
for ( i in 1:n ) {
dcast <- ifelse( dcast-1 > 0 , dcast-1 , 0 )
if ( dcast == 0 ) {
d <- ifelse( runif(1) < pc , 9 , d )
dcast <- waitcast
}
d <- ifelse( d-1 > 0 , d-1 , 0 )
hc[i] <- ifelse( d > 0 , 1 , 0 )
}
mean(hc)
}
hc <- {}
pclist <- 0:10/20
i <- 1
for ( p in pclist ) {
hc[i] <- simhc(pc=p,n=9999)
i <- i + 1
}
plot(pclist,hc,type="l" , xlab="crit" , ylab="proportion HC uptime")
curve( 1-(1-x)^4 , from=0 , to=0.5 , add=TRUE , col="red" )
The black curve will be the simulation results. The red curve is the analytical estimate, assuming w=1 and n=4.
Now, this isn't a completely general model. I solved with a fixed n = 4 (2 second cast intervals), and this model assumes even cast intervals. We can't be sure that the quality of the estimate doesn't depend strongly upon n, until we solve for either a more general model or more other n values. Thankfully, that's not a hard task. But a model of uneven cast intervals would be a little trickier, perhaps.
But absent the additional modeling effort, I am at least comfortable to conduct rough theorycraft using the 1-(1-c)^n approximation. It seems pretty good, to get a sense of the uptime.
This is a purely qualitative comparison, but watching the Hodir kill video that has been linked on mmochampion/WoR, the Disc priest in that video never comes close to running OOM. Now, obviously the factors are different, 10 man vs. 25 man being the most obvious, but I'm wondering, given the extent of the mana regen nerf for holy, is how is Disc holding up comparatively?
My question is.... where are these random people who don't show a glyph of PWS heal getting rapture from?
Well it looks like you guys had a second discipline priest in the raid that has no idea how to play the class, Jochen according to the WWS log. It looks like he cast 1 flash heal and 40 PWS (40 Borrowed Time effects on himself) during that Grobbulus fight, with only 6 providing a Rapture return. So the two of you got a total of 15 mana returns due to Rapture and the rest of the raid got 15 energize effects due to Rapture so that makes sense. I am guessing he did not have the PWS glyph.
Serendipity probably changed my playstyle last night more than any other change, including the regen changes. Pre-nerf of Fusion Punch, we had 2 holy priests assigned to healing our MT and while the incoming damage was relatively high, we kept one priest on the MT full time (weaving FH/GH), while the other was free to FH/CoH the raid and then drop .9 second GH bombs on the tank. Reactive healing with GH was a lot more realistic which meant that once my 3 stack was built, i could take some mana naps as necessary. Later, we were able to get one of our holy paladins in so I switched over to raid healing and the combination of the buffed CoH and serendipity-hasted PoH is every bit as powerful as it looks on paper.
Some other quick thoughts after two nights of testing:
- I'm now planning on picking up ToF over BR to start. I realize in some builds these aren't mutually exclusive, but when I'm allocating points I've usually worked it out so I'd pick up one over the other and if the other encounters are anything like what we've seen so far, it's seems likely that tanks and the raid will spike below 50% on a regular enough basis that I think I'll get more total healing from ToF. Some numerical proof would obviously be good but with the servers being as unstable as they were it wasn't exactly easy to switch mid-fight so I'll have to be content to try it out on another fight.
- The initial heal of empowered renew is...meh. On fights were I was primarily tank healing, it accounted for 2-3% of my total healing done. While raid healing, the highest I saw was 7%. I currently use renew a lot more often than many people that post here, and renew ticks accounted for 12% of my healing for the night so once they fix empowered renew crits to proc HC, it'll probably be worth it just for that if you use renew a lot.
- Both fights I've tested and watching videos of the other fights really made me value the mobility that SoL affords a lot more. When the talents first came out, my initial reaction was that I didn't want to lose potential HC uptime to SoL non-crits but after actually running some of the fights, I've changed my mind.
- Compared to some others here, I actually didn't feel the hit from the regen nerfs as badly as I expected. For buffs I had Kings, Mana Spring, Mana Tide and Replenishment and I switched from the haste-heavy set I use on live to a more balanced crit set, so I had ~ 350 crit rating/350 haste. HC uptime was pretty good and it really felt like the fight was designed as people expected: pretty intense burst with some periods of lighter healing afterwards. As people theorized before we could test, these fights would have been so preferential towards spirit-based regen as to really hurt class balance.
Overall, while I don't love the regen nerfs, I understand where they were coming from a lot better. It may be a bit heavy-handed, and if so I'd imagine they'll tune it, but I think the nerfs were in the right vein. It really did help keep people a bit more focused on what they were doing rather than just casting heals on anything the second it takes damage, which I think was part of their goal. The other changes were amazing and I still wonder if serendipity will make it through completely intact.
Ya, I wouldn't be surprised if they drop Serendipity to 10/20/30% instead of 20/40/60. 0.9 second GHeals just seem ... insane.
Anyone who cannot cope with mathematics is not fully human. At best he is a tolerable subhuman who has learned to wear shoes, bathe, and not make messes in the house. - R.A. Heinlein
Bleh, I entirely missed seeing that. We had asked if there were any other disc priests and they all said they were holy, guess its possible he switched at some point. He had 40 borrowed time gains with only 1 heal being tossed the entire fight. Can't believe I missed that (was doing my math at 4:30 am though to be fair =P .)
From our 7sh attempts on Iron Council I really wasn't having mana issues as Disc even without a replenishment though I did have a tide in group. With a replenishment I really think I wouldn't even have to worry about mana on that fight as I had a fiend up on the final phase. More testing has to be done sure, but at least for this encounter I didn't feel a squeeze.
Current Disc Build I have ran for Hodir and Council: Talent Calculator - World of Warcraft
- I'm now planning on picking up ToF over BR to start. I realize in some builds these aren't mutually exclusive, but when I'm allocating points I've usually worked it out so I'd pick up one over the other and if the other encounters are anything like what we've seen so far, it's seems likely that tanks and the raid will spike below 50% on a regular enough basis that I think I'll get more total healing from ToF. Some numerical proof would obviously be good but with the servers being as unstable as they were it wasn't exactly easy to switch mid-fight so I'll have to be content to try it out on another fight.
- The initial heal of empowered renew is...meh. On fights were I was primarily tank healing, it accounted for 2-3% of my total healing done. While raid healing, the highest I saw was 7%. I currently use renew a lot more often than many people that post here, and renew ticks accounted for 12% of my healing for the night so once they fix empowered renew crits to proc HC, it'll probably be worth it just for that if you use renew a lot.
Mavv two things quick, could you post your proposed build where you take ToF over BR. And the initial heal of empowered renew is currently broken, so the 2-3% of the healing it was doing was on yourself only. The bug is that it is only healing yourself with the instant 15% and not your target, I honestly think that when it gets fixed it will be *very* nice.
Interesting. I had noticed the Emp Renew bug earlier in the day, but by the time we were raiding I had SCT setup which seemed to indicate my initial heal was going toward my intended target and I never bothered to check it against recount. Clearly if that's the case my comments are invalid for now.
As far as build goes, http://talent.mmo-champion.com/?prie...0&version=9626 will be what I test next time we're in. Again, I'm making a lot of assumptions of the limited experience we have so far but it really feels like they're moving more toward a model of controlled burst and in that context I don't think that there will be as much PoH spam as I had originally assumed when I saw the patch notes. Also, if after more testing with the fixed Emp Renew I'm not more impressed, that would be a place to salvage some points for either BR or Healing Prayers.
One important thing to remember is how much mana over the course of a fight surge of light will save us. If you look at Nid's learning attempts on Hordir last night he gained 41 SoL procs in 19 mins and 23 seconds, an average of just over 2 a minute. Each time you SoL proc it saves you ~625 mana, to get the same benefite from Healing Prayers you would need to cast four PoH's a minute (I don't know about you but that is a lot of them and a consumption of ~6k mana a minute).
Don't forget about the savings to PoM from Healing Prayers.
If you cast PoM once every 10 seconds (6 casts per min) you save 696 mana per minute
If you cast PoM once every 7 seconds (around 8.5 casts) you save 986 per minute
That's significant savings by itself, with PoH an added bonus.
Ya, I wouldn't be surprised if they drop Serendipity to 10/20/30% instead of 20/40/60. 0.9 second GHeals just seem ... insane.
Not really, we need to cast 3 subpar "tankheals" in order to get that 0,9 sec GH. Overall single target-HPS is still low compared to paladins and druids (for holypriest). Don't get me wrong, 0,9 sec GH is good, but you will never choose a holypriest over a paladin or druid for single target-healing anyway, so don't see the issue.
Serendipity is simply amazing for raidhealing though. Raidhealing on Freya tonight with all those hasted PoHs was probably the most fun healing I've done since KJ.
Not really, we need to cast 3 subpar "tankheals" in order to get that 0,9 sec GH. Overall single target-HPS is still low compared to paladins and druids (for holypriest). Don't get me wrong, 0,9 sec GH is good, but you will never choose a holypriest over a paladin or druid for single target-healing anyway, so don't see the issue.
Serendipity is simply amazing for raidhealing though. Raidhealing on Freya tonight with all those hasted PoHs was probably the most fun healing I've done since KJ.
I agree in principle, but what it does is make a holy priest a very strong complementary healer to a holy paladin or resto druid. Throwing flash heals around on the raid and being able to have such a strong reactive heal (especially when combined with ToF, but that's a separate issue) really made me feel a lot more flexible than I have in the past. I'm OK with the GH part, but the 1.1 second PoH's just seem like a lot of HPS to be tossing around.
I need to sit down and do the math, but I have a suspicion that FH->GH->FH->GH ends up being some pretty insane HpS. Remember that FH was at least 80% of the HpS of GHeal, and when it procs a -20% effect on the cast-time of GHeal, you start to see some pretty impressive throughput numbers.
If anyone feels bored and gets to it before me, the model is:
1) Consider FH->GH->FH->GH, with standard crit/spell. What's expected throughput?
2) Then consider FH->FH->GH->FH->FH->GH, and compare.
3) Compare this to a paladin with 20% baseline haste (+WoA+Moonkin/Ret Aura), and average HLs of 11k.
[e] One further note to Rukli: yes, yes you will choose a priest over a druid or a paladin, especially for the second slot. Until druids and paladins get an AC-increasing proc, priests are more mitigation for similar (slightly less) HpS. And with the nerf to rolling Lifeblooms, druids aren't necessarily going to be as powerful at multi-tank healing in future, which opens up more slots for us.
Last edited by constantius : 02/28/09 at 8:53 PM.
Anyone who cannot cope with mathematics is not fully human. At best he is a tolerable subhuman who has learned to wear shoes, bathe, and not make messes in the house. - R.A. Heinlein
Not really, we need to cast 3 subpar "tankheals" in order to get that 0,9 sec GH. Overall single target-HPS is still low compared to paladins and druids (for holypriest). Don't get me wrong, 0,9 sec GH is good, but you will never choose a holypriest over a paladin or druid for single target-healing anyway, so don't see the issue.
Serendipity is simply amazing for raidhealing though. Raidhealing on Freya tonight with all those hasted PoHs was probably the most fun healing I've done since KJ.
Rukli did you find you were hard pressed throwing aroudn 1 second PoH's?
I've been thinking about models for uptimes of HC and trying to figure out a 'soft' cap for crit if you will. But my brain is just not working for math at the moment, stupid homework.
Empowered renew is not subtracted from the heal. It is a flat out 15% instant heal, followed by 25% strength ticks (assuming you're glyphed which if you go the whole renew route you should take ALL the talents and the glyph). It works like this:
Is the renew glyph fixed? If so this is awesome. Empowered renew not only gives 15% more healing it also gives 15% more scaling. Adding the renew glyph really makes renew quite powerfull
If this is the case at 2500 spellpower fully talented renew would heal for
(1400+(1.5/3.5*1.88+1.15)*spellpower)*1.05*1.15*1.1*1.15 = 9890 non crit, or 10535 with a crit. The initial heal is 1290 and glyphed ticks would be 2150.
===============================
20% reduction in cast time is 1/0.8 = 1.25 = 25% haste. When you cast FH you increase GH HPS by 25%. 40% reduction is 1/0.6 = 1.6667 = 66.67% haste. 60% reduction in cast time is 1/0.4 = 2.5 = 150% haste.
At 2500 sp GH is ~4k HPS, while fh is ~3333 HPS. FH-GH spam is ~4150 HPS. fhx3-GH spam is ~ ~5000 HPS
Each 20% reduction in cast time gives a disproportionately larger increase in GH HPS. So 1x serendipity stack increases gheal throughput by 25% while 3 stacks increase throughput by 150% which is 6 times larger than 25%.
However the main advantage is dealing with spikes and OH, this change makes holy priest much better tank healers than they were and I consider them much better than druids for solo healing a tank.
Tank healing is about dealing with spikes.
Holy priests are now extremely good tank healers, as they can uses flash while damage is low and whenever a spike comes up respond with an extremely fast gheal. Priests now beat the socks off druid as a solo tank healer and they are very close to paladins. Serendipity has increased holy tank healing ability by 10x.
Empowered renew is not subtracted from the heal. It is a flat out 15% instant heal, followed by 25% strength ticks (assuming you're glyphed which if you go the whole renew route you should take ALL the talents and the glyph). It works like this:
Is the renew glyph fixed? If so this is awesome. Empowered renew not only gives 15% more healing it also gives 15% more scaling. Adding the renew glyph really makes renew quite powerfull
If this is the case at 2500 spellpower fully talented renew would heal for
(1400+(1.5/3.5*1.88+1.15)*spellpower)*1.05*1.15*1.1*1.15 = 9890 non crit, or 10535 with a crit. The initial heal is 1290 and glyphed ticks would be 2150.
Rukli did you find you were hard pressed throwing aroudn 1 second PoH's?
No, not at all - the fight is easy. Mana was not an issue at all.
PoH is just extremly powerful on the encounter as you have very regular AoE-dmg (you know exactly when it comes). Having a hasted PoH ready procing ToF is spectacular HPS and HPM. I even casted quite a few non-hasted PoHs just to keep mana efficency high, as said the fight was easy and you were not forced into spamheal mode. We ran six healers though and could probably have done it with four without much trouble.
On topic of holypriest tankhealing at the moment .. I guess it comes down to personal preference and what healers you have in your guild. Of course holypriests can do great tankhealing, but saying that 0,9 sec GH is sooo much better than Swiftmend doesn't make any sense to me when you have to cast loads of Flash Heals to get it. I also think that a discpriest or a restoshaman is a lot better if you want armor buff.
Sorry, but to me Serendipity is not what makes holypriests great at tankhealing, it's what makes us good at raidhealing / assisting on tanks when needed. We're extremly flexible now when even Renew is worth throwing around.
I guess MavSteele said it better: (...) but what it (Serendipity) does is make a holy priest a very strong complementary healer to a holy paladin or resto druid. Throwing flash heals around on the raid and being able to have such a strong reactive heal (especially when combined with ToF, but that's a separate issue) really made me feel a lot more flexible than I have in the past. I'm OK with the GH part, but the 1.1 second PoH's just seem like a lot of HPS to be tossing around.
I need to sit down and do the math, but I have a suspicion that FH->GH->FH->GH ends up being some pretty insane HpS. Remember that FH was at least 80% of the HpS of GHeal, and when it procs a -20% effect on the cast-time of GHeal, you start to see some pretty impressive throughput numbers.
If anyone feels bored and gets to it before me, the model is:
1) Consider FH->GH->FH->GH, with standard crit/spell. What's expected throughput?
2) Then consider FH->FH->GH->FH->FH->GH, and compare.
3) Compare this to a paladin with 20% baseline haste (+WoA+Moonkin/Ret Aura), and average HLs of 11k.
[e] One further note to Rukli: yes, yes you will choose a priest over a druid or a paladin, especially for the second slot. Until druids and paladins get an AC-increasing proc, priests are more mitigation for similar (slightly less) HpS. And with the nerf to rolling Lifeblooms, druids aren't necessarily going to be as powerful at multi-tank healing in future, which opens up more slots for us.
I mathed Serendipity out here, although I did forget to account for GCD capping on the FHx3 rotation. Short story is that FHx2->GH is definitely a large improvement over straight GH spam or FH->GH, and FHx3->GH is likely a small improvement over that. (Feel free to point out any other glaring errors.)
02/28/09, 10:41 AM
