Jayde

After the various discussions in my [Priest] Healing Gear and Optimal Stat Weighting thread some time ago, I decided to sit down to work on a spreadsheet for comparing and contrasting multiple aspects of gear/stat comparison in a flexible way. (Got a bit tired of doing it all by hand, or relying on Lootzor/Pawn which tend to be a bit inaccurate at times.)

Features of this spreadsheet are:
1) Auto-calculation of weightings using my personal method or healing pool
2) Ability to override weightings with your own custom values
3) Gem values based on current weightings
4) Enchant values based on current weightings
5) Sheet for end result of healing pool/HPS/HPM given the input healing stats
6) Adjustments for various talents
7) Ability to toggle the effect of Human/PMC/Buffs/Consumables/Tier 5/Tier 6 bonuses
8) Ability to toggle auto-selection of Epic vs. Blue gems
9) Gear comparison by slot based on selected weightings, including gem selection

To-Do: I will also probably add an "upgrade compare" which will let you put in a set of stats on two sides and show the effect of the upgrade (or downgrade) in weight value/HPM/HPS/pool/etc. Should also implement the T5 4-set bonus on the Renew portion of the sheet

Edit: v1.1 - Updated with some fixes to stat trickledown and base regen
Edit: v1.2 - Added weight method selector, allowing calculation via my method (Downrank Ratio), GH7 Healing Pool, Selected Downrank Healing Pool, and Custom (input in the Calculations sheet)
Edit: v1.3 - Added the 2620 value for base mana in regard to mana pool.. but I may have incorrect data on this. Please post if you have a better way of calculating this.
Edit: v1.4 - Modified healing pool calculations. Added duration component. Fixed some minor errors. Added links to WoWHead in the gear comparison. Cleaned up some name references for people who want to poke around/debug formulas. Changed the way Intellect value is calculated based on fight duration.
Edit: v1.5 - Added support for Mana Potions, Shadowpriest, and Shadowfiend mana regen sources. Implemented a more generic method for calculating stat multipliers (BoK/SoR/etc.) to make sure they are consistant in all places. Fixed some more minor formula errors. Added GH7/Downrank cast number to the 'Healing Results' summery (in addition to noting if you are capped for the duration). Added note for if your Downrank Target is capped (time to possibly look at upranking for the 'Downrank Ratio' method.) Changed the gear comparison so the 3rd place item was compared to the 1st place, not 2nd (so instead of X, -20, -3 you will get X, -20, -23.)
Edit: v1.6 - Large update! Updated for upcoming 2.4 patch Spirit/Int calculations, added some new items, added a bit more info, cleaned up some formulas/errors, removed old patch 2.2 comparisons, converted all formula lookups pointing to secondary worksheets to named ranges for easier debugging
Edit: v1.7 - You can now input your base stats as they appear on your unbuffed character sheet, instead of adding up your gear totals. Added ability to filter gear-based tiers, added all/most of the Sunwell healing gear I could find, updated some inproper stats on the Etherneum Life-Staff, fixed an error where meta sockets were not being considered when comparing head slot items
Edit: v1.8 - Added inital modeling/ranking of Trinkets, added Imperial Tanzanite to the gem list, added modeling for IED and support for selecting which meta gem to use in helm slots, added initial modeling and support for Haste weighting, added a few missing Sunwell patch items, added support for showing healing data for Prayer of Mending (x1 and x5), Prayer of Healing (x1 and x5), and Binding Heal (x1 and x2) in the Calculations sheet for references purposes
Edit: v1.9 - Added much better T5/T6 set bonus support, added some new heroic gems, added the new SSO exalted neck items, fixed the modeling for average adjusted OO5SR% cast times, fixed the modeling on Hyjal exalted ring proc
Edit: v1.10 - Full support for crit calculations (this is toggleable via the 'Enable Crit?' setting), full support for haste, added input for character crit and haste ratings, added weapon oils to the buff list, added a proper Shadowfiend regen model, added Iridescent Fire Opal and Quick Lionseye to the gem list, added proper crit/haste values to the gear sheets due to them being weighted now (e.g. Dark Blessing, Hammer of Attonement), added a couple missing items to the gear sheets

Version 1.10 can be found here:
priest_healing.xls - FileFront.com

Some tips and notes for this:
-If you want to use your own custom weights and not my methodology/formula, you can input your own weights on the calculation sheet in the 'custom' section and use this instead. This is the -only- thing you need to change. The values of Spirit/Int are simply derived from Healing/Mp5 values as they can both be converted to be relative to those two.

-The main sheet will only show the top 3 items in every slot. However, the Equipment_XXXX worksheets will list 30 items. If your tier of items aren't in the top 3, just reference the Equipment_XXXX sheets and look at the Rank/Total columns.

---

Anyhow, this sheet was thrown together rather quickly over the course of a couple evenings, so I have no doubt there are some errors on my part lurking around. If anyone finds any or has any feature requests, feel free to post here.

Hope it is useful.

 

expired321

I honestly don't see the benefits to gain from using this spreadsheet, the methods you use for calculating your aep system are totally irrelevant to the actual weight of the stat, also your calculations of base character sheet stats are inaccurate/incomplete.

The rounding on your intellect and spirit calculations is off causing slight differences in the original values from what they actually should be, but these are amplified throughout the calculations and cause a lot of inaccuracies. Also, you failed to take into account the base OO5SR regen you have before spirit calculations (31.25 mp5) which causes huge differences in base stats and will really throw off the results of your calculations. Also, you forgot to take into account the base mana of your character as well as any mana gains coming from sources other than int (intellect only makes up for about 80% of your mana). Lastly, in end game healing stamina often plays a very big role so I would advise putting an option in to include the weight of stamina in gear calculations as well.

You also were not considering a large ammount of buffs that the average priest would have in a raiding situation, which plays a big role seeing as how drastically the weight of a priests stats depends on the other stats the priest has at the time. You need to include buffs such as AI, fort, and consumables/buff foods/oils.

Once these things are sorted out it will give a much more accurate representation of your stats which are required if attempting to set up an average equivalency point system. However, the way in which you calculate the value of mp5 (and in turn the rest of the values) makes little to no sense to me. It would appear as if you are calculating mp5 as a ratio of the healing you lose from downranking verses the mana you "don't lose" that you would have if you hadn't of downranked. so in effect you are simply stating a modified version of the HPM changes of downranking which is in no way at all related to the actual value of a point of mp5. This method of calculation will tell you little to no relevant information about the realistic weight of your stats.

As far as I know the only relevant way to derive an AEP system from holy priest gear is to set up an equation that will define the total throughput of your stats in terms of raw healing done (assuming you use all of the mana you get over the course of a fight) and then take the derivatives of the single stat values to come up with the benefits you would recieve from adding 1 of a certain stat to your current gear/spec/buffs etc. The equation I for this is "Healing Gain" = (Number of casts of Greater Heal from base mana + Number of casts of Greater Heal gained from regen over duration of fight) x (Healing done by Greater Heal). By expanding on this basic formula you are able to determine an AEP system that would give you the stat weights, and in term gear/buff/gem/enchant/etc weights that would give you the most to work with for the variables of a given fight.

I don't mean to bash you or anything but this spreedsheet simply does not provide an accurate look into stat weights for a priest, and as a result provides a gear ranking system that is irrelevant to almost every factor of the game.

If anyone is interestead in a spreedsheet based on the logic I used in this post I am currently working on one that I've put a good 200+ hours worth of time and theorycrafting in to. It currently is fully operational, however I haven't had the time to make it user friendly yet, and I'm guessing it would confuse the hell out of everyone but myself. If there is a demand for an accurate spreadsheet though I will definately put the work into completing it and throw it up here for people to use.

 

Jayde

As a point, there are no rounding on stats in the various cells other than display. I am not using the ROUND() function for any purpose, thus formulas referencing cells will maintain consistant regardless of if I set the display of a cell for reader-friendliness. (I highly doubt end-users want to see am 8 precision decimal value.)

As for the base values of the character, I presumed people will input their Intellect/Spirit from their character sheet, which would include the base values. You are correct that I forgot to include the base mana, and will add that.

I will add support for more buffs as I continue working on this, which are mostly trivial to add given how the sheet is set up on my working copy. Although experimenting with these didn't yield a huge difference other than looking at healing pool.

As for your argument on AEP types of calculations and relevance, I would say it is entirely up for debate there. You could make a system basing the weight solely based on healing pool, however I don't agree that it is the most valuable metric. That said, you are free to look at it that way if you wish. (You are also free to input your own weight values and have it compute the gear comparison based on those.)

However, since I do compute the effect on healing pool throughout, it would be trivial for me to add an option to rank based on pure effect of healing pool over X duration of a fight or healing pool until OOM. Will consider adding that to the next version.

The fact of the matter is, all healers tend to weight stats differently. Some Priests will weight based on throughput, others will weight based on longevity, and others will weight based on healing pool. I don't feel there is a "right" or "wrong" answer here, and most Priests have their own ideas as to how weights should be. I believe my default method of calculating it works pretty well for me, however I presume many people will punch in their own Mp5 weight and let it work from there.

As for the gain? There is a lot of mathwork when comparing gear, sockets, socket bonuses, etc. that people have to do. This sheet does most of it for them. That's why I created it, and I presume many others may find similar use from it--even if they have their own weights to punch it. I would say it is far from irrelevant, even if you have a different view as to how to rank things.

 

expired321

Actually I was referring to the fact that wow doesn't use a straightforward rounding system but in fact one that is quite illogical in my opinion, but hey there's blizzard for you. The rounding methods actually differ from stat to stat as well, best way to come up with these is to experiment with different gear setups and see how your stats change from what you would expect them to be. If you'd like though i could give a more detailed description of my findings when I've got more time.

What i was referring to is the fact that just as int doesn't account for 100% of your mana pool, spirit and mp5 does not account for 100% of your regen. There is in fact a base value of 31.25 mp5 while NOT casting that needs to be added to the regen value.

I understand and completely agree with the fact that different people will weight their stats differently, however unless i am completely misinterpreting the formulas used in your spreadsheet your method for calculating the weight of mp5 does not even take into account the characters stats. You can easily verify this by changing the base character values of mp5, intellect, and spirit (when imp DS and Spiritual Guidance are checked off) and you will see that NONE of these stats (except for the healing gain from spirit) effect your stat weights in ANY way. I don't know whether this is simply flawed logic, or a mistake in the construction of your spreedsheet, but as it stands at the moment your stat weights are based of of nothing more than your +healing.

I assume that it is readily apparent why an AEP system based off of a single stat is completely useless but on the off chance that is not I'll do my best to explain why. As your gear changes (and this goes for almost any class/spec not just a holy priest) the weighting of your stats will change, and almost every stat in the game suffers from diminishing returns at some level. What I mean by this is the more you stack a single stat, the less value you gain from a single point of that stat when compared to the gain you would achieve from another stat. A hypothetical situation to prove this as well as to show why I believe your stat weight calculations to be extremely inaccurate would be to compare a priest with 0 mp5 and 2000 healing to a priest with 2000 mp5 and 2000 healing (I realize these numbers are completely unrealistic but I am using them only to show how drastically off your spreadsheet is in its current state). Now, according to your spreadsheet both of these priests would have EXACTLY the same stat values as one another (3.26 mp5=1 healing). In reality, however, the priest with 0 mp5 would have a DRASTICALLY higher gain from a single point of mp5 than the priest with 2000 mp5, for the priest with 0 mp5 would be able to cast a few heals and then be useless for the rest of the fight so would obviously benefit from a much higher regen, where as the priest with 2000 mp5 would have WAY more than enough regen for doing anything but chain casting Prayer of Fortitude, and would gain 0 benefit from acquiring any more.

I really hope going into detail more will help you understand why your spreadsheet is so flawed and if you are still confused say so and i will do my best to explain it in a way you may be able to better understand.

I didn't mean to imply that a spreadsheet provides little to no gain, a good one provides a vast wealth of knowledge that breaks down gear decisions for you and presents them in a much more logical, and controlled manner. Your spreadsheet however, based on the above logic, does not provide a system that is even remotely close to accurate so as a result the conclusions produced from it contain absolutely no value whatsoever and thus it makes this particular spreadsheet offer absolutely no useful or accurate information on a method of gearing yourself.

 

Jayde

The argument of stat balancing is really not tied to an AEP system however, although I understand your logic. Even with a DPS spreadsheet, AEP systems are generally guidelines with assumptions that the player will meet certain balancing requirements on their own.

For instance, +hit is worth quite a lot for a Mage (and is the most efficient value per budget on an item) however is should be quite obvious to the Mage that any +hit once you are hit capped will provide no further value. But, retroactively changing the AEP for that would downprioritize it inappropriately when looking as a "general" overview.

As for my methodology on the default method, it's explained in my other thread.. but, as I have already said, I expect many people to plug in their own ratio based on personal experience. I will add a weighting method based on healing pool however, for people who prefer that without having to toy with the numbers.

As with any AEP system, it is a way of looking at items in a general sense assuming the end-user will eventually prioritize certain things over another. Doing this automagically somewhat defeats the purpose, as the weights would fluctuate heavily.

Some brainwork always needs to be done when considering weights, however being able to have a quick overview based on personal preference/weighting is always useful (especially when evaluating gems/socket bonuses and similar tier items) which is why the sheet was created.

 

expired321

I think you have entirely missed my point. First of all what you have set up in this spread sheet is an AEP system, or in other words a system of values that are applied to different stats in order to give them a weight. What your spreadsheet is designed to do is to generate these values (for i assume the numbers under the Values section are in fact the values you give the stats) based on the players current gear and then plug those values into the currently available pieces of gear in order to compare and rank them. Now where your system is flawed (and no i don't meant simply that our opinions differ, i mean that your spreadsheet flat out does not work) is that the values are generated without taking the characters stats into consideration. What your spreadsheet is currently saying (yet again whether this be an error in constructing it or a flaw in your logic i don't know) is that because Greater Heal Rank # has a healing per mana efficiency of ##, therefor 1 point of +healing equals x points of mp5. You are comparing 2 things that are completely irrelavent of one another in order to obtain a result that shows absolutely nothing.

Now, if what you are trying to do is provide a static AEP system that is not based on gear or stats and you are not in fact attempting to provide a then the construction of a spreadsheet is simply unneeded, for the purpose of a spreadsheet is to provide calculations based on changing variables (IE to provide a system of stat weights based off of your current stats). If you were in fact intending to simply provide a static AEP system then you need to do nothing more than simply list what your opinion is on what is the best gear for raiding. And if this is the case, then opening a new thread for this is not at all needed, there are plenty of more relevant places to provide this info.

 

Jayde

Uploaded a new version with fixes to the base regen stat and some other calculations. Also added a buff option which will allow toggleable buffs in the Buff sheet.

As the results with my base stats from gear input ends up matching my character sheet excatly, it seems to be correct now. Thanks for the base regen note.

"And if this is the case, then opening a new thread for this is not at all needed, there are plenty of more relevant places to provide this info."

Is there, really? I've been browsing around and posting in these forums for quite some time, even when I was primarily playing my Mage, and when I swapped to playing my Priest primarily in raids I was a bit shocked at the lack of resources for Holy Priests compared to every other class. There are very few things floating around even remotely statistical compared to pretty much any other class/spec/whatever.

Therefore, I figured I would post it should other Priests find it as useful/interesting as myself and a number of healer friends I have shared it with do. You are certainly free to stick with your own sheet if you prefer.

I, for one, don't enjoy whipping out a calculator/notepad/spreadsheet every time I want to compare two items considering sockets, socket bonuses, set bonuses, etc. I'd imagine most others feel the same way. Hense the sheet and hense the post.

You are free to disagree with the premise of my sheet, however it does provide a lot of things in one place that are pretty handy--even if you use your own weightings. (The interesting thing is, despite possible objections to my method of calculation, in the other thread and in general talking with other Priests, the end result of weightings aren't that deviant from various other methods. But feel free to post in the other tree about it.)

 

Lazare

Assertion 1: Once we make assumptions regarding talents, set bonuses, fight length, and the time spent inside the FSR, spirit, int, and mp5, are equivalent; all can be expressed in terms of mana.

Assertion 2: The fundamental priest stats are therefore +healing and mana. The key function of a model is to determine how to weight these two key stats relative to each other.

I think everyone in this thread would agree with those two assertions. The hard part is working out how to calculate the tradeoff.

Jayde is, I believe, arguing that there is a "natural" tradeoff presented by the game. Using a lower rank heal is like using a higher rank heal while wearing less +healing and more mp5 gear. Dropping from GH7 to GH6 is very similar to losing 330 +healing and gaining 128 mp5.

There are a few problems here, I think.

1) It ignores HPS. This is understandable, since that's extremely hard to evaluate, and alternative "total healing pool" systems do the same, but it's still a flaw.

2) This isn't a "natural" tradeoff; it's an arbitrary one. Bliz might of had a plan when they made the spell, but they might have just fished them out of a hat for all I know. Even if they did have a plan, why do we think that particular tradeoff reflects the effectiveness of two stats in some particular encounter?

3) It assumes that these two heals are identical, and that we're indifferent to which one we cast. The logic goes, I think: "I don't care if I cast GH7 or GH6. Going from GH7 to GH6 is like going from 330 +healing to 128 mp5. Therefore I don't care if I get an extra 33 healing or 13 mp5 from my new gloves." It seems to this that we haven't proven step 1. Are we indifferent to ranks? I doubt it. I usually cast the heal which seems best; I rarely pick a heal at random from one or more that seem equivalent. But if we're not indifferent to ranks, then that means we're not indifferent to the tradeoff. If GH7 is better than GH6, then 330 +healing is better than 128 mp5. And if that's true, then we haven't established an equivalence, but simply a bound.

In short, I think Jayde has done a great job of making a spreadsheet to do the heavy lifting of an AEP. I also think Jayde has done a good job of calculating the healing/mana tradeoffs inherent in downranking. What I'm not quite seeing is why that tradeoff is meaningful.

Also: If I understand the implications correctly, the spreadsheet strongly implies that as gear levels increase, healing becomes increasingly unimportant relevant to other stats. In other words, mp5 doesn't have diminishing returns (or if it does, they're overwhelmed by the diminishing returns to +healing. That seems odd to me; my experience has been that, if anything, it's mp5 which has the greater diminishing returns. Does that seem right to you?

 

expired321

This is where I am saying his formula is incorrect, if you look carefully at the way his formula is set up, or use the method I stated above and change the base character values of mp5, intellect, and spirit (when imp DS and Spiritual Guidance are checked off) in the spreadsheet you will see that NONE of these stats (except for the healing gain from spirit) effect your stat weights in ANY way. When setting up his formula he failed to incorporate mp5 and intellect in any part of the calculation which is why I have said that this formula and as a result this spreadsheet/formula is wrong. The lack of a diminishing return on mp5 being present in his spreadsheet is a result of this. The AEP values in his spreadsheet are modified by nothing other than his +healing (feel free to play around with the values and observe the results to prove this beyond a shadow of a doubt), and the fact remains that you cannot render an AEP system off of only one single, it simply will not show you any relevant information.

I'm simply trying to warn people that if they use this spreadsheet in its current state that they will end up getting results that are based on a nonsense formula that have no merit in the actual mechanics of the game.

 

Jayde

I have uploaded a new version 1.2 which allows you to toggle different weight calculation modes. Default is still 'Downrank Ratio' (my personal method of preference), however you now have GH7 and GHX healing pool available in the drop-down as well. You can also specify a Custom weight in the Calculations sheet and toggle between them. Hopefully this will help people who prefer to weight based on healing pool.

This way, one can argue about my calculation method in the other thread, and leave this one for the spreadsheet specifically.

(Personally, I view the inherit healing/efficiency tradeoff provided by up and downranking to be a natural way to base the weight of efficiency. Lower ranks have higher efficiency but a lower coefficient, thus as healing increases upranking vs. downranking tradeoffs represent pulling the weight of healing and efficiency in different directions. This is arguable, of course, but I would prefer you discuss that in the other thread.)

The spreadsheet was intended as a resource for those not wanting to do math manually, not to say that I think my stat method (which was the default) is perfect. I admit completely that not everyone will view stat weightings the same as me.

On that note, if anyone has any other suggestions for other weighting options to build in to the sheet, I would be happy to add them.

 

galzohar

This is the primary flaw of any healing spreadsheet I've seen. And TBH it's hard to impossible to evaluate. Personally I prefer spending the time to try figure out a solution to this.

If you watch the healadin thread I offered a possible solution for paladins (which is rather contraversy, as it makes +healing by far the best stat), but paladins are different since they can use a lower HP/s higher HP/mana heal with a significant difference in both values, thus HP/s could be translated to HP/mana due to the extra HP/s allowing more use of FoL. Obviously this isn't accurate, but until someone shows me otherwise, at least in "first-order" it's pretty good.

For priests it's different because last time I checked (correct me with numbers if it's wrong), with +healing in the 2k+- zone, different ranks of GH have more or less the same efficiency, which means downranking simply lowers your overhealing while reducing your OOFSR time, which may or may not be a good thing depending on the fight (as in, how much overheal reduction VS how much OOFSR time). Ignoring changes in overheal reduction and OOFSR time, different ranks of GH are generally the same from the prespective of a spreadsheet that tries to maximize the "amount of healing you do with the mana you have over X minutes" (assuming you can't just spam max rank non-stop and not go oom, which is generally quite true).

So how would priests value extra HP/s? To me it seems that as long as you're not needing a bigger max rank GH, additional HP/s is useless. Don't get me wrong, +healing isn't useless as it gives more efficiency and increases the "healing done with your mana over X minutes", it's just that it seems to me that the HP/s portion of it would net 0 benefit to a priest (again, as long as you're not needing bigger max rank heals).

Thinking about it a little more, priests don't just cast GH/FH/BH... Renew and PoM are highly efficient spells that have limited use (cooldown, can't cast more than 1 renew on the same target is what I mean by "limited use"), so basically higher hp/s means those spells heal for more which means you need to heal less with your other spells. The more of your overall healing PoM/renew are, the more efficient you get, which is where HP/s might actually increase your effectiveness.

Of course if your max rank heals aren't big enough and people die because of it, forget everything I said and max your hp/s... But afaik this generally doesn't happen, and even if it does you're still not using any standard spreadsheet for it...

 

Irise

Awesome work Jayde. I haven't had a chance to reply to your PM but I will definitely do so and take a closer look at the spreadsheet it once my work schedule loosens up.

Would it also be possible for you to include a tab which has HPS, HPM, etc. for all Priest spells much like Nightshroud's spreadsheet. It would also be nice for your original post to have the new 2.3 coefficients for Binding Heal, PW:S and CoH.

Also would you mind, if I edited your spreadsheet for my model? Perhaps, we can combine the two models and have a toggle.

Lazare, the point at which healing supersedes mp5 is when your mana regen in a raid setting exceeds your mana consumption. As such if you are casting Gheal rank 3 or lower this is a very easy to do. However if you are casting Gheal rank 7 then it would take quite a lot of mp5 before diminishing returns start kicking in. With that being said, I think you maybe right in 2.3 when the new meditation kicks in.

Edit: A few more replies came in while I was typing.

To the poster above me, Even with 2k +healing Gheal rank 1 is almost twice as efficient as Gheal rank 7 with other ranks being spread in between.

 

galzohar

As I said, if I did the numbers wrong please point me at correct ones, as you've obviously seen them if you're so sure.

anyway if downranking = more efficinecy, hp/s can be simply translated into more efficiency due to more downranking - again assuming you're not needing bigger max rank heals.

 

Jayde

Another thing in regard to regen.. I see people note that it has "diminishing returns." However, I don't believe the math (especially if you are looking specifically at healing pool) really shows that at all.

Until you can chain-cast GH7 forever, you will always get use out of regen from my perspective. Second, until you completely cap out on infinite sustainability, regen has increasing returns per point on healing pool rather than the other way around. With my current +healing stats, at 100 mp5, adding an additional 100 mp5 increases the healing pool of GH7 spam by around 8k. At 1000 mp5, adding an additional 100 mp5 increases the healing pool of GH7 spam by around 454k.

Of course there is a cap here in regard to how much you have to heal, however given ~10 minute duration fight I believe that cap for non-stop GH7 spamming is somewhere around 1200 mp5. Of course, the realistic cap is much lower but it certainly isn't anywhere near the 300 buffed mp5 mark that most people run around with in raids.

Just my 2 cents here, but I don't feel that regen particularly becomes diminished, other than the fact that, at the end of the day, you do have to get your HPS to the correct target regardless of how you do it. (This is, of course, why I personally compare up vs. downranking vs. healing gains per cast to weight regen.)

Irise:
If you grab the newest version and come up with something in the Custom bit and send it my way, I could probably include it as an option. I also would like to add all the Priest spells, but need a bit of time to get that running. On my to-do list!

 

expired321

Sigh...last post im making here unless something is specifically directed at me. My point is obviously not getting across, so I'm just gonna say this. Jayde it's great and all that you made an attempt at a spreadsheet but if you don't know how to do it then your simply going to confuse people and throw them off. In your latest version the "GH7 Healing Pool" options math is completely wrong, not even going to get into it, also you messed up your buff's section (specifically the consumables which are not being calculated), and the rounding is still not correct for intellect and spirit (i recommend experimenting with your gear before stating something as fact). Lastly as a fair warning to anyone reading this, know that the formulas, as well as the functionality of this spreadsheet are wrong so use it at your own risk.

EDIT:
THANK YOU for proving my point that your calculations are entirely wrong. Anyone that knows anything about holy priests knows that these numbers could not possibly be correct thus proving that his system for generating them is fundamentally flawed.

 

galzohar

-healing pool" concept is wrong imo. You're trying to get as much healing done over X minutes, not "Z healing until you go oom". Changing to this concept greatly affects your perception on +healing VS mp5, as with your concept mp5 is increasingly good as the more you have the longer you can cast and thus even more mana you have, while in a real fight it'll last X minutes and the mp5 will give exactly Y mana, or Z healing regardless of how much mp5 you already have. The diminishing returns are that Z healing per Y mana goes up the more +heailng you have, and therefore the balance between +heal and mana leans towards +heal the more mp5 you have and towards mp5 the more +heal you have.

*looks like someone else also disagrees with the healing pool concept (posted while I was typing), keeping my post here though as it shows how to correct this rather than just saying it's wrong.

 

Jayde

Consumables need to be toggled "Yes" on the main page to take any effect in the calculations. By default I turned this off.

Anyhow, thanks for the few things that you pointed out so I could correct rather than vauge bits about how "things" are wrong. That really isn't very constructive or helpful. No, I have not spent 200+ hours on the sheet as you have, nor have I had the opportunity. That doesn't mean that being non-constructive/condescending is going to be terribly helpful or worth posting.

As I mentioned before, I'm more than happy to correct any formula errors.

Not really sure what you mean, honestly. It should be clear that regen--as with any % based reduction of a primary stat--will have increasing returns on longevity. This is the same as the fact that a 1% mitigation increase while tanking provides more value when you have 98% than when you have 0%. I'm not saying it always pays off in regard to what you would have to do to reach that point--nor does my formula assume as such.

 

Bendyr

I agree with Jayde that there are a lot more mathematical resources for other classes compared to healers, probably because it's much more difficult to model healing except for specific fights.

No one is expecting this spreadsheet to be completely authoritative, I think it's a great resource for new priests. Obviously the formulas can be improved slightly, but it's a nice effort.

One suggestion- most of the other classes' gear spreadsheets (i.e. shadowpanther) provide external links to the items, which would be kind of nice.

 

Jayde

I have the IDs for the item in the sheet, so I should be able to link to WoWHead. Will try to add that in my next version. Thanks for the suggestion.

 

galzohar

Again, longevity is a pointless value as it assumes you're not healing at all after that period is over, which is not true. It also assumes you can keep healing regardless of when the fight is over, which is also not true.

 

Lazare

Okay, first:

I think your HP pool calculations are completely off. I'm not entirely sure where the error is, but my spreadsheet says that you should be seeing something like a GH7 healing pool of 300k. You're calculating it as 80k in 15 casts, which only takes around 10k mana.

That doesn't sound right. In a 5 minute boss fight, 200 mp5 alone provides more mana than that...


Imagine you have the mana to cast a single heal. Each point of +healing you do is applied to that single heal, which means you get around 0.86 HP healed per point of +healing (if we assume GH7).

Imagine you have the mana to cast a hundred heals. Each point of +healing you do is applied to all 100 heals, which means you get around 86 HP healed per point of +healing.

The way I see it, this means that every point of mana you have increases the value of your +healing. In other words, mp5 makes +healing more valuable - which is another way of saying that mp5, expressed in terms of +healing, has a decreasing marginal benefit.


That's an artifact of your approach to calculating a healing pool. You're basically calculating how fast it takes you to go OOM, then multiplying by HPS, I think. As regen approaches spell cost, the time to go OOM approaches infinite - but boss fights aren't infinitely long. And as I calculate it, there are diminishing returns to regen, because +healing becomes relatively stronger.

At 1800 +healing and 300 mp5:

• An additional 100 +healing adds ~7k HP healed.
• An additional 100 mp5 adds ~48k HP healed.
• Therefore, mp5 is ~6.9 times better than +healing.

At 1800 +healing and 400 mp5:

• An additional 100 +healing adds ~8k HP healed.
• An additional 100 mp5 adds ~48k HP healed.
• Therefore mp5 is ~5.9 times better than +healing.

At 2000 +healing and 300 mp5:

• An additional 100 +healing adds ~7k HP healed.
• An additional 100 mp5 adds ~50k HP healed.
• Therefore mp5 is ~7.2 times better than +healing

My calculations don't show that regen does less as you stack it - it does exactly the same. However, +healing does more - which has the same net effect is an AEP system.

In other words, I don't see how we can give the same weight for additional mp5 to someone with 1800 +healing and 150 mp5 as we can with someone with 1800 +healing and 450 mp5. The impact of 1 additional +healing for the latter person should be much higher - or it is in a healing pool calculation.


To reiterate, however, I'm quite impressed with the spreadsheet. (Although I do wish you'd made much greater use of named ranged. Needing to work out what "Calculator!$H$5" means makes debugging much harder, in my experience.)

 

Jayde

"That doesn't sound right. In a 5 minute boss fight, 200 mp5 alone provides more mana than that..."

My method to calculating healing pool at present is casts until OOM while also considering regen with the given in-combat regen + OO5SR% regen amounts. This is similar to how TheoryCraft/DrDamage calculate things in tooltips so it's what I went with initially.

I intend on putting some fight duration stuff that will cap things somewhat, but that would be a two-stage process and take some time to implement. It's on my list, but it's a bit harder to do. Wanted to get a basic version done before I looked at fancier features.

"Although I do wish you'd made much greater use of named ranged. Needing to work out what "Calculator!$H$5" means makes debugging much harder, in my experience."

I agree and will go back and fix this soon. The sheet grew to be substantially bigger than I originally set out for it to be. So a bit of an error in planning on my part. I will fix this after the next couple tweak versions.

As for the healing/regen debate, I would rather prefer to have it in the other thread, but I will say I can see your point but also see the other side of it. Upranking costs more mana and has lower efficiency, but ranking up for a higher coefficient is worth looking at. At extreme +healing values, gaining more Mp5 "could" be a better way to increase things due to being able to uprank to a higher coefficient spell.

Many Priests stick to lower rank spells due to longevity reasons (as we cannot chain-cast high rank spells forever) however as your +healing increases the HPS between the ranks also increases due to a lower coefficient. This makes upranking (and thus needing more regen) a comparatively more efficient way of increasing your healing with extreme +healing compared to with low +healing. I assume that when someone picks a "Downrank Target" they are picking it due to longevity limits (e.g. "I need to spam a heal the whole fight and I can spam GH2 and make it to the end") and thus we can use that timeframe/manacost target as a scaler for looking at the relative scaling of the +heal/regen values. If you have enough Mp5 to chain-cast GH4 forever, I would not select GH2 as the Downrank Target.

(Edit: Perhaps I could look into a way of auto-detecting this to make it more intuitive? e.g. looking at your OO5SR time to assume you are casting X% of the time, and figure out what the highest rank spell you can cast for that % of the Y minute fight before going OOM is and select that automatically as the downrank target? This is more or less what I do when I select the downrank target for myself, but I can see how this may be unclear. I will also add casts/time to OOM in the healing results section to make this easier to eyeball.)

I can see why not everyone will agree with that methodology though... which is why I won't attempt to bash it through anyone's skull as the One True Correct Way. (Of which I'm not sure there is for the sake of healing, simply due to the dynamic, case-by-case, and unpredictable nature.)

---

Back to the spreadsheet though, if you have any different ideas for calculating healing pool or spot any cells with errors, I will try to fix/implement changes ASAP.

 

Jayde

Found the issue with healing pool calculation with non-zero OO5SR figures. (Forgot to reference a cell that would increase the cast time to compensate for the delay needed to get OO5SR for that percentage of time... so the mana/sec consumption was a bit off due to the fact that to get OO5SR for 30% of the time, you would have to delay the cast of your next spell past the normal 2.5s cast time.)

Fixing those issues as well as adding a duration cap that will cap out casts after X seconds, as well as zero out the further value of regen in healing pool weight modes if you reach the cap for your selected rank.

Should be done with all the modifications in about 4 hours as I have to raid this evening.

 

Jayde

Have uploaded v1.4. Most of the notes are edited into the OP, but have modified a lot of things based on suggestions. Should be a little cleaner now and I believe the new healing pool fixes should be a lot more accurate when considering OO5SR% periods.

(I still stop healing pool calculation at OOM point. I will look into extending the calculation to include the effect of regen after the OOM point--however that's a bit more complex to add so I'll have to think of the best way to do it. Given the impact of the correct OO5SR regen though, the OOM point is a lot higher now. However, I presume from practical experience that if you ever do go actually OOM in a fight, you're probably up shit creek and someone will die. So this seems fair for the moment.)

Lazare, I've started adding name references for stats/settings and started converting a lot of the formulas over to use them. Should be a little easier to debug/understand at a glance now.

Bendyr, I've added links to wowhead for the items on the gear comparison sheet.

The addition of duration should make a few more things clear, and it will state when you have reached an infinite threshold. Please note, when looking at downranked healing pool weight method Mp5 can become extremely high weighted as you approach an infinite mana scenario with high +healing. (If you think about it logically, the difference between being almost infinite and infinite is actually quite large towards the end.) If this happens, you most likely should look at choosing a less efficient uprank for comparison. That said, the value of mp5 will become zero if you reach the casting time limit or infinite cast scenario itself.

 

Kinien

I made a pretty dumbed down spreadsheet for making gear decisions and calculating spirit vs mp5, it kind of pales in comparison of yours.. If you want to take a look and see how i work mine here ya go.

http://home.comcast.net/~kinien/Hosting/gear.xls

OR for 2.4 download

MP5-modified.xls