Maybe this is a stupid question or had been answered already but i don't find the answer among all those post.
What's the crit/haste cap for holy and discipline unbuffed ? I play with both spec and atm use the same gear for boths pec wich is not ideal..And I really wonder if my haste is not too low. I use extremely a lot PoH glyphed in both spec, so 234 haste might be too less for progression in ulduar.
I just skimmed through the last four or so pages of this thread and there is plenty of discussion about crit vs haste vs other stats, potential caps, whether the caps matter, how to weight those stats, and various other stuff. Since I can't quote the entire thread I would suggest you actually go and read it.
For simplicity's sake you can assume that Spell Power is your best throughput stat and getting as much of it as possible is good. Largely we don't get to choose how much SP we get on our gear though and have to pick between Spirit, MP5, Crit and Haste. What you need is enough regen that you don't go OOM and then a balance of Crit and Haste. The more you spam shields, the less useful these two stats are (since they can't crit and BT limits Haste) but it's largely impossible to replace them with anything else except regen which is pointless if you don't need it.
(4) Set: Casting Power Word: Shield also grants you 250 spell power for 5 seconds.
Talent: Borrowed Time
... and increases the amount absorbed by your Power Word: Shield equal to 40% of your spell power.
I think the set bonus could increase the shield by 100 depending on which affect came first. Anyone know?
As far as I know it goes like this:
Lets say u have 2500 spellpower (raidbuffed or whatever).
PW:S rank14: absorbs 2230 damage.
Extra absorb gained from spellpower: 2500 * 0.8068 = 2017
Extra absorb gained from Borrowed Time: 2500 * 0.40 = 1000
Semi Total absorb: 2230 + 2017 + 1000 = 5247
Then extra absorb from Imp.PW:S : 5247 * 1.15 = 6034.05
So first shield thrown gives 6034.05 absorbtion.
Second shield, now with 250 extra spellpower:
PW:S rank14: absorbs 2230 damage.
Extra absorb gained from spellpower: 2750 * 0.8068 = 2218.7
Extra absorb gained from Borrowed Time: 2750 * 0.40 = 1100
Semi Total absorb: 2230 + 2218.7 + 1100 = 5548.7
Then extra absorb from Imp.PW:S : 5548.7 * 1.15 = 6381.005
So the extra 250 spellpower is converted into 6381.005 - 6034.05 = 346.995 extra absorbtion.
(Same as calculating: (250*0.40 + 250*0.8068) * 1.15, just thought example was more clear )
This is ofc the gain from 4set bonus, when only looking at PW:S.
Correct me if I am wrong here :-)
As far as I know it goes like this:
Lets say u have 2500 spellpower (raidbuffed or whatever).
PW:S rank14: absorbs 2230 damage.
Extra absorb gained from spellpower: 2500 * 0.8068 = 2017
Extra absorb gained from Borrowed Time: 2500 * 0.40 = 1000
Semi Total absorb: 2230 + 2017 + 1000 = 5247
Then extra absorb from Imp.PW:S : 5247 * 1.15 = 6034.05
So first shield thrown gives 6034.05 absorbtion.
Second shield, now with 250 extra spellpower:
PW:S rank14: absorbs 2230 damage.
Extra absorb gained from spellpower: 2750 * 0.8068 = 2218.7
Extra absorb gained from Borrowed Time: 2750 * 0.40 = 1100
Semi Total absorb: 2230 + 2218.7 + 1100 = 5548.7
Then extra absorb from Imp.PW:S : 5548.7 * 1.15 = 6381.005
So the extra 250 spellpower is converted into 6381.005 - 6034.05 = 346.995 extra absorbtion.
(Same as calculating: (250*0.40 + 250*0.8068) * 1.15, just thought example was more clear )
This is ofc the gain from 4set bonus, when only looking at PW:S.
Correct me if I am wrong here :-)
Your first idea was to measure how long a given healer can sustain a certain rate of healing, a value one could measure in minutes per 1k hps. However, the fundamental problem is that time and hps are not anti-proportional to each other, so one has to fix an arbitrary hps value and just assume it is a good estimate. Again this is completely arbitrary and depending on class, spec and encounter may be wildly off.
...
Since [infintely sustainable hps] is class and spec independent(again, perhaps split in single target and raid-wide), one can simply add the values of all healers to get the raid sustained healing ability. Compare it with the similar sustained incoming damage of a specific encounter and one has an estimate of the number of healers necessary.
There are some good thoughts there, including one potential use for knowing the HPS you can sustain indefinitely. You could potentially add this up across healers, and if it is enough to cover the average incoming DPS indefinitely, you should feel pretty good about your sustainability.
I am slightly surprised there has been so little discussion about how to size up healing longevity. It is a fundamental issue, but a tricky one because we're trying to reduce a complicated nonlinear relationship down to a number. For illustration purposes, consider the following graph as a conceptual illustration of the problem:
So how do we best summarize that information? We cannot really say our longevity is any particular length of time because we may last anywhere from ~1 minute to indefinitely long, depending on our healing output. On the other hand, our maximum "sustainable hps" similarly covers a wide range depending on how long we need to sustain it. More to the point: you cannot meaningfully talk about longevity without also talking about HPS. The two are intertwined.
One thing that does stand out to me is that if you specify a relatively long time period and figure your sustainable hps, the result is really not that much higher than the hps you could sustain indefinitely. Both values are on the right sided tail of curve [the above example curve asymptotically approaches 3000 hps, just fyi], and the two values should correlate well with each other. If one is raised or lowered by some percentage, the other will move to a similar extent (percentagewise).
However, I don't think the left side of that curve should be ignored. We all agree peak hps is important, but I want to have some idea how long that peak hps could be maintained.
Another approach that is sometimes used in science to summarize similar nonlinear relationships is to compute the time one could maintain exactly one half his/her maximal hps (or something like that) -- a 1/2 max hps constant. For instance, if your max hps is 10K, then you look at how long you can sustain 5K hps. That may be appealing because it gives you a point in the middle of the curve, which might be closer to where you're operating, rather than on the tails of the curve. On the other hand, healers are often on either the right or left tail of the curve, rather than in the middle.
Originally Posted by Elimbras
I prefer to set up the fight length, but that's a personal opinion. The first reason for this is that I have a better estimation of fight length, which I also find more stable between fights that for needed hps.
The interesting thing here is that I agree that fight length is much easier to predict than needed hps, yet that is partly why I want a longevity metric in time, not hps. It is all about which problem you want to be easier, but my primary argument against a longevity metric measured in heal per second is that I don't know what to do with it. Trying to figure out how much sustained hps I'll need on each individual fight seems pretty difficult. It varies both within and between fights and according to raid composition, etc. How can we expect anyone to figure out "on the spot" whether X sustained hps is enough?
I understand your point, though, since for the same reasons it is more difficult to estimate a time from a hps than the other way around. That is, an estimate of sustained hps should be more reliable because there is less uncertainty about your assumed time period.
Originally Posted by Jamora
An additional advantage is that it is directly comparable to peak HPS.
Originally Posted by Elimbras
there is a natural way to merge sustained hps and peak hps metrics. You need to be more clear about hidden weights when merging longevity and peak hps : you're adding apples and oranges.
No, don't go there. While it is tempting to think that sustained hps and peak hps can be merged more easily because they have the same units (heal per second), semantically they are not the same thing. If you combine peak hps and sustained hps in any way, you still must weight each appropriately. Otherwise you do end up with hidden weights: you'll assume both have equal value because they have equal units.
Because throughput and longevity are somewhat orthogonal concepts (and indeed are the two different dimensions on the graph above), it might make more sense to combine them multiplicatively anyway. I mean, we could evaluate the "size" of a table by adding together the length and width, but most of us would tend to favor multiplying the length and width to get an area, even though the units are the same. But regardless I really really think you have to weight both factors no matter what. Units don't really change that; they just alter the weights.
Maybe this is a stupid question or had been answered already but i don't find the answer among all those post.
What's the crit/haste cap for holy and discipline unbuffed ? I play with both spec and atm use the same gear for boths pec wich is not ideal..And I really wonder if my haste is not too low. I use extremely a lot PoH glyphed in both spec, so 234 haste might be too less for progression in ulduar.
There is no simple answer for this question.
Different Haste effects stack multiplicative and there is a cap of 1 second for gcd.
32,79 Rating = 1% Haste.
The Formula is : New Casting Time = (Base Casting Time)/(1 + (% Spell Haste / 100))
A raiding Disciplin Priest should have baseline:
Enlightment: 6%
Wrath of Air Totem: 5%
Improved Moonkin Form/Swift Retribution: 3%
And sometimes
Power Infusion: 20% or Heroism/Bloodlust: 30%
Borrowed Time: 25%
These buffs will give you 14.64 % haste (1.06*1.05*1.04=1.14639) or with BT 43.29%
(1.06*1.05*1.04*1.25=1.4329)
These caps are only true for 1.5 second casts and instants
Disc will need
self buffed: 41,51% = 1361.11 Haste Rating
Raid: 30.85% = 1011.42 Haste Rating
Raid + PI: 9.03% = 296.35 Haste Rating
Raid + BT: 4.68% = 153.34 Haste Rating
Raid + H/B: 0.65% = 21.33 Haste Rating
Conclusion:
Disciplin:
You should have at least 154 Haste Rating to reach the Haste cap with PW:S spam and given that Crit provides no benefit to PW:S the Crit Rating cap is 0.
When you are MT or PoH healing everything between 154 and 1012 Haste Rating is a HPS gain (aside from various PI,BT or H/B uptimes)
In pure HPS calculations Crit Rating is worse than Haste Rating, even with DA, the only benefits are DA shields when your target is over 100% HP.
In my opinion, you should gem (SP,SP/INT,SP/SPI) and enchant Spellpower and aim at least towards 297 Haste Rating through Items. There will be enough Crit Rating on the other Items to support Inspiration and a decent DA Procc Rate.
Holy:
Holy needs only enough Crit Rating to ensure almost 100% Holy Concentration uptime (np with coh/poh) and as much Haste Rating as possible
This is my best english, so don`t complain.
Edit: There is no Haste Cap, aside from the 1 second global cooldown limit.
Last edited by Aiel : 09/21/09 at 6:44 AM.
Reason: miscalculation
Aiel, the thing is though that if my memory serves me well, Enlightenment, Wrath of Air totem and Moonkin Form give haste that is additive with haste rating, rather than multiplicative. Borrowed Time, on the other hand, is multiplicative.
Therefore, total haste required to reach GCD cap is 20% (1 - 1.5/1.25). Enlightenment, WoA and moonkin Form give a total of 14% haste, hence 6% haste will put you at GCD cap.
I must admitt that I have never checked whether Enlightment, WoA and Moonkin Aura stack multiplicative or additive but untill now i was under the impression that only Haste Rating from Items is additive and all other effects are multiplicative.
And all of them stack, aside from Heroism and Icy Veins which stack not with Power Infusion.
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
No, don't go there. While it is tempting to think that sustained hps and peak hps can be merged more easily because they have the same units (heal per second), semantically they are not the same thing. If you combine peak hps and sustained hps in any way, you still must weight each appropriately.
I completely agree with your point, however you misunderstood me. The advantage of both values being in hps is not that they are somehow easier to merge into one single metric(they aren't, since they serve different puposes). However, the difference now has a meaning: It shows by how much a healer can increase his healing output if the situation demands it(of course there are limitations, like the weakened soul debuff). It sounds obvious, but a different metric for longlevity does not allow for that. If we had some value like [time to oom] for an arbitrary hps value, it would not be comparable to [peak hps] this way. For example, for a holy paladin, the difference will be small, while for a holy priest the difference is very large. Or perhaps in a better way: The quotient [peak hps]/[sustained hps over infinite time] will be close to 1 for holy paladins(compared to other healers). It is common knowledge, paladins can spam holy lights almost without pause. However, these metrics did not know that, but we can recover that fact by looking at the quotient. At the same time it implies that said holy paladin cannot cover for other healers if something happens, the healer setup is build so that he always operates near his limit.
You also mentioned that the area close to peak hps warrants more attention. While I agree, in my opinion it is a minefield if we try to put a number to it. There are so many factors: cooldowns(like penance, IF, hymn), buffs and debuffs(weakened soul, mortal strike effects), positioning(in case of raid healing),..., so I doubt that we can do better than some value like peak hps. One just has to remember that the peak hps do not reflect reality, but peak hps should be strongly correlated to the maximal healing in the real world(nice, considering the virtual game as real world ).
I read in the middle of that post a lot of information about stats weights
can smb pls give me combination lets say 85% power to 15% longevity
thanks in advance
Your only real choice will be between Haste and Crit Items. Almost all Ulduar cloth healing gear has Stam, Int, Spirit and SP and due to Blessing of Kings and FSR, Spirit Items are better than MP5.
Your only real choice will be between Haste and Crit Items. Almost all Ulduar cloth healing gear has Stam, Int, Spirit and SP and due to Blessing of Kings and FSR, Spirit Items are better than MP5.
Could please provide more math on this? At least you contradict clearly the OP in a single sentence and advise the opposite.
IMO haste is better than crit not just for output but even when it comes to DA stacks. Boss fights have a time limit and don't go on forever, so when you can cast more spells within a certain time-frame, you are also getting more crits in, even though your crit rating hasn't changed.
Aiel, the thing is though that if my memory serves me well, Enlightenment, Wrath of Air totem and Moonkin Form give haste that is additive with haste rating, rather than multiplicative. Borrowed Time, on the other hand, is multiplicative.
Therefore, total haste required to reach GCD cap is 20% (1 - 1.5/1.25). Enlightenment, WoA and moonkin Form give a total of 14% haste, hence 6% haste will put you at GCD cap.
Enlightment, Wrath of Air totem and Moonkin Aura are multiplicative.
With 350 Haste Rating my Resurrection cast time is 7.88 Seconds
IMO haste is better than crit not just for output but even when it comes to DA stacks. Boss fights have a time limit and don't go on forever, so when you can cast more spells within a certain time-frame, you are also getting more crits in, even though your crit rating hasn't changed.
I agree that haste is better throughput stat, even once DA is included.
However, I'd like to know how you justify that haste is better for DA stacks. Sure, you cast more heals and get more crits. However, the crit spells increasses is usually lower that the increases you get from crit stats.
Just some napkin math computation:
Assume you already have x% spell crit, y% spell haste, and that during a time frame, you can spam N heals.
Presently, your expected number of DA stacks is N*x.
If you increases your haste by one 1%, you'll be dividing your casting time by (1+y+0.01) instead of (1+y). Therefore, the number of heals you will be able to cast will be N*(1+0.01/(1+y)), and you'll get roughly N*x*(1+0.01/y) crits. You gained about N*x*0.01/(1+y) DA stacks.
Now, you increases your crit number by 1% also. Your expected number of crits will be N*(x+0.01), leading to a gain of N*0.01. The Da profit of haste is higher if and only if x/(1+y) is higher than 1, which is not possible.
If you consider that you need about 1.5 more crit rating than haste rating to increase your stat by 1%, crit is more beneficial for DA only if x*1,5/(1+y) >1, which won't happen unless you already have more than 66% of crit.
I agree that haste is better throughput stat, even once DA is included.
However, I'd like to know how you justify that haste is better for DA stacks. Sure, you cast more heals and get more crits. However, the crit spells increasses is usually lower that the increases you get from crit stats.
Just some napkin math computation:
Assume you already have x% spell crit, y% spell haste, and that during a time frame, you can spam N heals.
Presently, your expected number of DA stacks is N*x.
If you increases your haste by one 1%, you'll be dividing your casting time by (1+y+0.01) instead of (1+y). Therefore, the number of heals you will be able to cast will be N*(1+0.01/(1+y)), and you'll get roughly N*x*(1+0.01/y) crits. You gained about N*x*0.01/(1+y) DA stacks.
Now, you increases your crit number by 1% also. Your expected number of crits will be N*(x+0.01), leading to a gain of N*0.01. The Da profit of haste is higher if and only if x/(1+y) is higher than 1, which is not possible.
If you consider that you need about 1.5 more crit rating than haste rating to increase your stat by 1%, crit is more beneficial for DA only if x*1,5/(1+y) >1, which won't happen unless you already have more than 66% of crit.
I think you misunderstood what I said. I never meant to say that 1 haste is better than 1 crit rating to get DA procs. I meant that the faster casts from haste not only increase output, but do give more crits when you take into account that boss fights have a time limit- be it an enrage timer or just a healer's mana.
Let's say you have 30% crit, you get 30 crits out of every 100 spells you cast. If you get enough haste to cast 110 spells within the same time frame, you get 3 more crits, even though your crit is still at 30%, which does mean extra DA procs during the fight.
As a throughput stat the amount to which you value crit or haste has a lot to do with indivdual playstyle and raid role, and everyone will have their own personal weight. Someone with incredibly fast reaction times or someone who's raid role is to pretty much spam heal a tank will most likely value haste over crit (assuming no extreme lack of either stat); on the other hand someone with bad Latency or frame rates might well value crit much higher than haste.
Specifically to Mocha's point on DA though:
I don't think that you can really use DA as a reason to value haste more than crit. If you are looking for more DA procs, crit is hands down the winner.
Take the example used above about haste increasing cast rate, thus increasing procs. You are correct in that your number of casts would increase thus the number of DA procs would also increase, as you demonstrated. Unfortunatly in your example, what you missed was the "Opportunity Cost" of increasing your haste.
So, you cast 100 times on a fight @ 30% crit chance, getting on average 30 crits. Increasing your haste so that you can cast 110 times in that fight should at that same rate give you 33ish. The haste needed to give you this number is 10%.
Since roughly 32.79 haste rating = 1% haste, you are adding 328 haste.
Now the important part, is that since the iLvl value of haste and crit is even, for every point of haste you gain, you gave up the opportunity to gain a point of crit.
Since crit sits at 45.91 rating = 1% crit, in this same 328 haste that you added in the above example, you could have instead added 7% crit
Go back now to that original 100 casts. Now in the same timeframe, you once again cast 100 casts. This time at 37% crit rate. This means you would average 37 DA procs during that time.
So, ΔHaste : ΔCrit is 1 : 2.3 for DA procs.
Now, I would even take it a step further and weigh DA even heavier on the side of crit. Unless you are called upon to solo heal a tank frequently, most Discipline priests find themselves reactive healing while spamming shields and popping PoMs. When reactive healing, a higher crit rate is going to give you a better chance of landing a DA proc on any one specific cast, thus any time lost not on the tank increases the value of crit for this. Additionally, haste has no impact on PoM nor on PW:S (assuming you are not below 6% haste...in which case increasing your haste should be a no brainer anyways), while crit DOES have an impact on PoM.
Again, to sum up, I am by no means arguing here that crit is better than haste across the board...but DA is strongly weighted in the favor of crit over haste.
Playstyle and other factors need to weigh the throughput vaules though.
I completely agree with your point, however you misunderstood me. The advantage of both values being in hps is not that they are somehow easier to merge into one single metric(they aren't, since they serve different puposes). However, the difference now has a meaning: It shows by how much a healer can increase his healing output if the situation demands it...
You also mentioned that the area close to peak hps warrants more attention. While I agree, in my opinion it is a minefield if we try to put a number to it.
Apologies for misunderstanding, and yeah, the different you mention will give you a kind of HPS range within which you operate.
I am not sure putting a number on peak hps longevity is all that much more difficult than figuring a peak hps, however. Definitely you have to make a lot of assumptions but all the assumptions about casting pattern are required to figure your peak HPS anyway. The main additional thing you need to know is how much mana you burn while doing whatever it is you do to generate your peak HPS. If you know that, your mana pool, and your mana regen, you can calculate your longevity (in seconds) from:
or from some similar formula. That could be done.
I guess I view it as an ordered pair: given some casting rotation or scheme, you can figure a both a hps and an associated longevity based on it.
Last edited by Promethia : 06/07/09 at 8:46 AM.
Reason: Corrected error in formula
This isn't entirely about Disc but I'm looking for a Priest's view on it rather than a Shaman's.
I recently ran Ulduar-10 with me as Disc and a Resto Shaman, we are both around half 226-geared and the rest 213, we are regularly healing in Ulduar-25 (up to Mimiron currently) and don't have any issues but obviously there are various other healers. The DPS in this 10-man was quite low, I wasn't running OOM but the Shaman was struggling a bit for mana.
We had trouble with Ignis keeping the tanks up when the Shaman got potted, and then on Council we really couldn't find a tanking/healing strat that worked, the tanks just kept dying when all three were still up. In the end we got the boomkin to heal and we managed it but it still wasn't that smooth. Should one of us be able to solo heal the tank with Big+Medium on him?
Then on Hodir we kept wiping on Frozen Blows, we'd survive one or two but then the raid damage was just too much and a few people would die, or the tank would die. Is Disc + Shaman just a bad combination for 10-man? My feeling is that it is, but that the Shaman should be able to handle the raid damage better than he was. As far as I'm aware we shouldn't need three healers for that fight (we tried it and it didn't help particularly but the Druid hasn't really healed since the expansion).
But concerning your question: you should elaborate your actual healing strategy a bit more. From the shaman going oom and "Shaman should be able to handle the raid damage better than he was" I'm guessing it was mainly disc healing the tank and shaman healing the group. For a start on most encounters I'd do it the other way round. With pre shielding, glyphed Holy Nova and hasted Prayer of Healing a disc has a much higher potential for healing/absorbing raid damage than a shaman. Disc + Shaman is not a bad combination, you should be able to heal most of ulduar but it of course also depends a lot on how the other 8 players are geared/played.
The shadow helm provides both crit and haste, similar spellpower, and when gemmed it provides even more Int. I am confused how the Holy helm out weighs that. The 4 piece from Holy is the clear winner but the helm is not required for the 4 piece so that can't be it.
I also looked at the other pieces in comparison possibly thinking that there was another peice that was far better. This was not the case though. All the Holy pieces are clear winners except for the helm. Unless I am just missing something major.
06/04/09, 1:13 AM
).
Similar Threads
