Malodorous

I've thought about it a lot actually, i just don't know why you see haste so valuable. I've talked about it with everyone in my guild that reads EJ and knows about comparing gear. Not one person can come to a conclusion of why haste is worth anything. I guess it really comes down to if you are still having mana problems.

 

• constantius

If you don't understand the value of a little haste, I'm sorry, but we can't help you. There's dozens of pages of posts here, many of which mention haste, and why it is valuable in moderation. No-one is advocating stacking it to the moon, but having 21% raid-buffed is nice. It drops the cast-time on your primary nuke heal (GHeal) by 0.5 seconds, reduces your GCD overall, and generally makes you a more responsive healer.

If you were to drop all that haste off your gear (even to the extent of using sub-optimal pieces for your belt / ring / chest / bracers), you would gain a marginal amount of crit (~ 5%). Since I'm already running 27% crit raid-buffed, along with 20% haste, I'm not exactly sure where you think you're gaining.

Finally, fix your profile.

[e] I just looked at your profile. You have haste items. Why are you wearing them if you think haste is worthless? In fact, you have a *lot* of haste items.

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

 

MavSteele

Mathematically, haste and spellpower are our most effective pure throughput stats with crit being a distant third. It's "always on", and its effects on cast time and GCD mean that it improves literally every spell we cast.

Despite our attempts here, however, healing is often more an art than a science due it's reactive nature, and it's in this sense that I find haste to be most valuable. Missing a heal with <.2 of a second to go is something that has happened to many, if not all healers. At higher gear/skill levels it is rarely a stupid mistake that gets people killed, and almost never a throughput issue. More often, the problem is not having a heal hit the target before they take more damage that kills them. While it's difficult to put a value on haste's "just-in-time" effects, it's a tangible difference.

 

Malodorous

Ive spend hours trying to find posts with mathematically figures about hast. It's pretty hard looking through 50 posts that talk about anything and everything. I understand about GCD and reducing my Gheal .5. But the argument people bring to me when i try to debate my geming for hast is that spell power doesn't have a dimension return.

Whats wrong with my profile?

And I said not one person can tell me why haste is good, i never said i saw it as worthless im just looking for some math on the subject. I figured the only way to see if i liked haste was to re-gem all my gear, so i did just that. Ive looked over a lot of priests here and lots have gemed hast.


Mavsteele thank you for the response i agree its more of a nature and i have been playing a priest from MC days so i know exactly what you are talking about.

 

siegfried

Your logic is confusing. You say you can see the purpose of reducing GCD and reducing Gheal cast, but you say nobody can give you a good reason? Those are the reasons. Gemming is ofc your prerogative, but when you say to stack spell power(?) it doesn't make sense because it is on all the pieces. The 4 choices you have on pieces is: 1) spirit/haste, 2) spirit/crit, 3) spirit/mp5, or 4) crit/haste. That is why constantius lays down simple guidelines for crit and haste. It really is quite easy.

TL;DR: Stop complaining that you can't find a good reason for haste because there is diminishing returns on crit, and it is essentially your only other option. (Plus they have already shown you 2 reasons which you quoted in your post.)

 

Malodorous

I don't have any need to reduce my GCD now that they are putting cool down on CoH is my point. I don't really need to reduce my Gheal because I find my self almost never using it now with the way raiding is going. Maybe after they nerf CoH i will use it more, but i really dont see my self using it.

If i stack spell power in my gems i can get more effective use of my renews and even my flash heals. In order to make flash heal more efficient you want it to over heal so you get your mana return from my understanding.

I never complained one time, i just stated the reason for my posts.

I'm starting to think you guys are experiencing different environments then i am. I find my self spamming flash heal to compete with our pallys. We have no resto shamans so i am 100% of the time raid healing.

 

Isin

Your post is making my brain hurt.

You do realize that GH is just an example right? Flash heal is also improved by haste.

You keep asking for math, but haste is one of those stats where the math is basically provided for you in-game.

 

Malodorous

I'm glad your brain hurts must mean you cant read well, i didn't say flash heal because who needs a faster flash heal. If you need a faster flash heal roll a pally. We have instant heals aka bubble. I'm talking about the math of effectiveness vs other stats, but i guess your brain hurts from the thinking involved.
Ive decided that hast is worthless and if you do the math spell power is 10x more effective in end game raiding. Spell power will effect every spell i have, hast will not. Flame on i wont be reading anymore, all of gotten out of this is people who decided that talking down to people over the internets is cool, grats?

 

siegfried

You are now being combative. Haste will affect every spell you have. Spell power will affect every spell you have. All gear you take will have spell power on it. If you don't like crit, take the pieces with haste on it. If you don't like either, take the pieces with mp5. Gem with all spell power. Post a WWS with all mp5 gear and full spell power gems. We will compare. Very, very, very simple.

/end discussion and lets get back to something useful

 

Beliandra

We're talking sustained usage. Hymn of Hope = 8% of your maximum mana each cast, 5 minute cooldown, so that's 8% of your mana every 300 seconds, or 0.1333% every 5 seconds.

But yes, you can optimize it by channeling the Hymn when the 300 int proc is up, for an extra 360+ mana.

 

Lazare

The Disc weights are bizarrely off, as I believe has been discussed before. You can more or less break stats down by their impact on throughput or regen. How to value throughput versus regen is open to debate - most weight scales would rank regen significantly higher than wowhead does, but it's not obviously wrong. However, even if you accept that, the disc scale wowhead supplies isn't internally consistent. Spirit and mp5 are both of use purely for regen, but mp5 is a vastly more powerful regen stat. No matter how important you think regen is, mp5 should always be given a much higher weight than spirit; the wowhead scale weights them the same, and that is absolutely wrong. Similar criticisms can be leveled at the throughput stats. It's ranking spellpower and crit as roughly even, and both being well above haste. That's probably wrong as well; I think most would put spellpower and haste as roughly even, with crit a distant third.

Holy looks a lot more plausible, although even there it's a bit dodgy. Much as with disc, crit seems weighted awfully high, and all regen stats seem awfully low. I have no idea where the scales come from, but they seem to have some systematic issues.

 

Veldefice

I just had a couple of gear questions since I've been wrestling around with the idea of changing some pieces.

Currently I have:

[Valorous Shoulderpads of Faith] and [Heroes' Gloves of Faith] (no luck on conq gloves off Sarth for me yet =/)

Last night I managed to pick up [Gloves of Token Respect] and I've had [Mantle of the Locusts] sitting in my bank for a while.

I have 25% crit in raid with about... 15% haste? (not exactly sure, but my GCD is 1.3 sec).

So, I'm thinking of swapping to both of these offset pieces for more haste. I would lose the crit from both pieces, but I just wanted to get some opinions to see if this change is really worth it especially with the CoH nerf soon.

Another thing I wanted to ask was... changing gems from [Luminous Monarch Topaz]to either [Reckless Monarch Topaz]... or the crit one was worth it at this point. For all red slots, I put in [Runed Scarlet Ruby] instead of an orange, but I can always change that to up some haste or crit.

I dunno, but I just wanted some input before I start spending gold on something I might not need to do. I'm happy with how I am right now, but I obviously want to be better. Oh.. and yes, I'm a spirit stacker.

 

• Richelieu

Any set of weighting scales can only be a rough approximation even when appied to identical talent builds, since the relative values of stats vary depending on the nature of the fight (e.g. mana demands vs. throughput demands), the mixture of buffs you bring, and both the relative and absolute amounts of Spirit and Int your gear has.

By way of comparison, here are some Holy weightings I derived from Rawr recently with Rawr set to optimize "Overall" value (both thruput and regen) in "Raid-Heal" mode in my usual raiding gear set (mostly 213 epics) and with a mix of typical raid buffs and consumables, for a 5-minute fight with an 85% ISR assumption:

Spell Power: 100
MP5: 118
Haste: 73
Int: 90
Spirit: 74
Crit: 53

I'm sure if I changed the Rawr assumptions I could also come up with numbers similar to WoWhead's. Weight scales are only useful to the extent the assumptions behind them match the situation you are gearing for. A good approach IMO is to become familiar with Rawr and play with it to improve understanding of what gear/stat combinations are theoretically optimal for different roles and situations (tank heal vs. raid heal, burst heal vs. endurance heal etc). Then test the theoretical recommendations in raids and see if they hold up. Theory is sometimes wrong in practice--a common example is valuing "on-use" trinket procs using the assumption that the player will always use them maximally in the stress of actual raiding.

On the subject of Haste, I think that is one stat where actual practice matters more than weights produced by theorycrafting. Theorycrafting can only value Haste with confidence in terms of its effect on theoretical throughput in sustained chain-casting situations (and to a very small degree, its potential effect on OO5SR %). In reality, there are often fights and healing roles where haste's theoretical contribution has only minor practical benefit. (Example: there are many tank-healing situations where one could comfortably keep the tank out of danger by casting unhasted heals). But as others have said above, a major practical benefit of haste is being able to land an emergency heal faster. Having a good amount of haste *does* save deaths from happening sometimes. Haste translates into some unknown X% increased chance of saving a fellow raid member from death in a difficult fight. Saving one raid member from death can be equivalent to contributing hundreds of thousands of damage points to the raid in some situations. There is no way to accurately factor that value of Haste into theorycrafted weight charts.

(EDIT TO ADD AN OBSERVATION: I should say there is no way for *us* to to accurately quantify the death-prevention value of Haste. But Blizzard conceivably could with their data--e.g., they could perform a multiple regression analysis of data from thousands of successful 25-man fights using the healers' stats (including haste) as the independent variables and total raid deaths as the dependent variable.)

 

• constantius

Per GC's request for suggestions on the priest (as a healer) class, I've started a discussion on the other thread:

http://elitistjerks.com/1054218-post168.html

Please keep it to that thread so all the material / suggestions / discussion is in one place.

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

 

Morakk

I would switch to the nonset gloves, if only for the extra int and spirit. Yay higher ilvl.
I'm less sure that I would want to switch the shoulders.

If you don't have gold burning a hole in your pocket I wouldn't regem. Pretty much any combination of spell power, intellect, spirit, crit, and haste is doing it right, or close enough.

Once 3.0.8 hits, I would reevaluate things. If you never ever have to work hard to have mana, then it would be time to gem away from int/spirit and maybe pick up some of the haste+crit items. [Cosmic Lights], [Signet of Manifested Pain], [Valorous Handwraps of Faith], a cloak, etc.

 

Distomos

It is really nice. As I said before, though: it is so difficult to keep your spirit higher without using a +spirit trinket in your second trinket slot or constant elixirs. As it stands, my spirit is just 11 lower than my intellect without any buffs whatsoever. In a perfect world, I would like to use a +spellpower trinket and still have the trinket proc even in a five man setting. This would not happen without a spirit scroll. You will find yourself sacrificing +intellect in terms of enchants (i.e. bracer) just to ensure spirit will indeed proc. You will also find yourself analyzing potential gear choices to a nauseating degree. So, to summarize: be prepared for a headache. Very often, I feel like pulling my hair out.

While this might be a bit of a shameless plug for my friend (Shadowed), if you're at all concerned with OO5SR as you should be (and particularly if you are considering the aforementioned Darkmoon choice), do try giving Pentagon a try. I asked him to post it this afternoon; we have been using it for a while. With RegenFu no longer being maintained and requiring FuBar to function, it's a helpful option albeit not providing detailed reports (i.e. best mana regeneration, etc) as RegenFu did. Change the color of the OO5SR background and add something like PowerAuras to the exact duration of your Greatness buff and it's difficult not to take advantage of it. :)

 

• constantius

If your spirit is slightly lower than your intellect unbuffed, it will be higher buffed. Think about the way buffs work.

Assume you get AI and DS. DS > AI. Assume you don't get DS. Scroll = AI. But then we get SoR scaling spirit, but not intellect. So long as you get +value equal for both, Spirit will gain more per point of raid buffs.

Say you have 1000/975, for example, and you have DS in the raid.

Gotw: +54 to both
DS: +80 to spirit
AI: +60 to int
Food: assume N/A
Flask: assume N/A
BoK: 10%

Intellect: (1000+54+60)*1.1 = 1225
Spirit: (975+54+80)*1.1 + (54+80)*0.05 = 1227

Went from -25 to +2. It only goes up if you actually keep them even pre-buffs, or if you (worst-case) use the 40 stam / 40 spirit food.

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

 

Promethia

I've recently finished extracting data on absorbed damage for discipline priests from ~500,000 lines of a Naxx25 combat log.

For boss encounters, the average absorbed damage was 32.7% of effective healing done. That is, for every 100 healing done, there was an additional 32.7 points of damage absorbed on average. This value is not significantly different from the 29.7% I measured a couple months ago using a Mount Hyjal log.

However, I also noticed that the amount of absorbed damage varied considerably. I computed a normal confidence interval (alpha = 0.05), and that gave me a range of 17% - 48% of healing done, which seems like an awfully wide interval to me. So keep in mind that "it depends" on the encounter, and your mileage may vary. Also, on trash clearing it was lower: ~22% of healing done.

After digging a little deeper, I could see that most the variance was in the damage absorbed by PW: Shield. The amount of damage absorbed by Divine Aegis was much more stable: 13% +/- 4% (which incidentally agrees with theoretical estimates assuming 20% crit and 40% overhealing). So not only can I say "your mileage may vary" but more precisely I can say that it will vary mostly depending on how much you use PW:S.

I still like the fairly simple hueristic of absorbed damage being 1/4 to 1/3 of healing done, but keep in mind that for discipline priests extensively using PW:S, it could be more than that.

 

Tattersail

Promethia thanks a lot for the information! Do you have data on the amount of absorbed damage that came from PW:Shield relative to DA, and the number of PW:Shields that were cast? And also interesting to know would be what % you got for what encounter, and how much PW:Shield was used on each encounter

 

Abygail

Found the thread, sorry.

 

Crow

As a side note sqrt(x) on x values ranging from 1000 - 2000 is actually is roughly linear. We can approximate it by something like sqrt(x) = 0.013x+19 and our error is roughly 1%. If you assume that the mana regen formula is 5*0.005575*spirit*(0.013*int+19), then your I5SR regen would be incorrect by around 3 mp5 at most. Saying that regen doesn't scale linearly with int while being correct, isn't really true. Regen is scalling linearly with both spi and int, only that the weight on spirit is greater.

 

Havoc12

I have already done this calculation. The answer is that at 2000 intellect 1 spi ~ 1 int with 10% ooFSR, including pretty much everything except hymn of hope, but that includes the extra starting mana. Without the extra starting mana, spirit is better or equal to int for regen, for most high end priests.


Sorry but you should spent a little more time thinking about it before you posted. If you think before you post you will avoid wasting both your time and mine. If you want to argue for the sake of arguing this is not the place to do it.

1) I did saph10 with 2 healers and could maintain sub80% FSR. If you can already cast all the spells you need then neither spirit nor int has any value for you. If mana is an issue then ooFSR maximisation is the most effective way to go.

2) How could you have possibly failed to spot that its 3:2 in favour of intellect. If you don't have a good grasp of the model, then go read up on it. I made a post with all the info you need further back.

3) You should really read posts carefully and think your posts through. The original poster said that if you have nothing to do and you can take a regen break you should use shadowfiend. If you have a chance for an ooFSR break and you break it to cast shadowfiend then you don't know how to play your priest. The situation you describe is ridiculous. So much so that it takes several lines to point out its ridiculousness
a) The first and most obvious, is that your mana is desperately low and you remember shadowfiend AFTER you took your regen break instead of chaining it at the end of a cast BEFORE you take your regen break? Real smart.
b) The + 300 proc only happens after a cast and it lasts 15 seconds. If you chain shadowfiend after a cast to take advantage of the proc, it does not break your ooFSR time. If you wait to regen and then cast shadowfiend not only you break your ooFSR regen, you also lose the benefit of the proc because it will fall off before shadowfiend is done returning mana.
c) This is a proc that comes up fairly frequently and shadow fiend is a 5 minute CD.

4) @1000spi 1500 int, 300 int adds 103 ooFSR. 300 spirit adds 320 ooFSR That is more than 3 times not half as much. Replenishment with 300 more int is 56mp5. shadowfiend and HoH you have no guarantee that you will be able to use within the +300 int buff. Realistically you need to determine uptime and divide the 33mp5 and 5.85 mp5 that the extra int gives to them. That pretty much takes them out of the equation. At 80% FSR spirit gives you about 40% more regen than intellect if you discount the increase in mana pool. I dont know if the proc is affected by BoK and SoR, but if it does the differences are larger. The larger the stats get the better spirit becomes.

Int procs without the extra mana are simply a waste, since you dont get nearly 50% of the benefit from intellect, which is the increased mana pool.

 

• constantius

Sapphiron isn't necessarily a big issue when it comes to 2 healers, especially as a priest. If you really want to challenge 5SR time, try Sartharion. Since you haven't done Sarth+1 (or, well, any Sartharion at all), you may not know just how hairy it gets. Sapphiron is a trivial fight comparatively. Cast PoM, collect epics.

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

 

Morakk

I think I could have written more clearly. It seemed to me that people were implicitly assuming something along the lines of what you've shown. I decided to actually check whether 330 was a large enough interval for it to matter, and found out that it isn't and that intellect still nets you close to as much regen at 1530 as 1200, which was not something I trusted my intuition to.

 

Havoc12

The poster above me talked about Sapphiron so I responded about Sapphiron. If you want to say that there is an encounter where you can't get much ooFSR time, that does not tell me anything, unless you are trying to say that 80% FSR is an unreasonably low assumption.

As far as hairy fights, any fight can be hairy. Our tank on faerlina (achievement) could be burst down to zero from 100% in under 2 seconds. Keeping up the tank and solo healing the raid (no poison totem) was not too hard but definately not trivial. The horsemen where quite tricky with our set up (2 pally tanks and the ppl tanking the horsemen at the back needing constant healing). I also did not have enough HP to eat up a solo meteor strike, which meant I had to time running back and forth quite carefully (I actually got killed by it in the end). The fight was difficult because I had very little time to actually heal. There you go a hairy fight with lots of ooFSR.