Bokomatic

Skyhoof, what formulae do you use for calculating HEP? Where can I check them out ?
Thanks for info in advance ^^

------------------- SPREADSHEET REBUILD ------------------------------------------------
Hey Ho Mates!

I've rebuilt my healing spells spreadsheet.

At the moment it calculates Healing done,HPS,HPSC,HPM,MP5 for all valuable ranks of healing spells.
In addition average HPS is calculated for each spell over different duration fights.
New "mixed" spells conception was involved.
Added a little of new personal preferences.

New version can be downloaded here: RSSSS_v106.rar; RSSSS_v106.zip ( .zip are greater than .rar for 1.5 times)

Spreadsheet is made in MS Excel 2003, i don't know if it works in OpenOffice.
Spreadseet use 'iterations' function, for full functionaliy of spreadsheet you should to switch this function on in your Excel.
Check the Iteration option in menu Tools->Options->Calculation[Tab]–>Iteration[checkbox].

A little tip:

How to use spreadsheet:

- Only cyan cells are for input your character information. Don't modify gray cells please

1. Go to "<0>Talents" sheet and set up your talents



2. Set up your relics, if you use macros for swap relics when casting you can choose all 3 types.
If you use only one relic always - choose only one in spreadsheet too.



3. You can see below table calculated for your talent build and containing
both base values (without +healing)of HPS,HPM
and scaling coefficients for HPS,HPM (per +1000 healing)



4. Go to "<1>Stats" sheet and enter your paperdoll stats and external sources of MP5



5. Next you can setup some facultative settings



5.1 Human delays were embedded in the model for closer fitting to real cast process.
It isn't ping or latency (that problem has removed already from client-server transactions).
It's just human reaction times and percentage of misses. I do delays about 50-100ms per cast.

5.2 Setting nonzero Crit UltraOverheal will underrate crits, at least I think that crits overheal more than normal heals.
5.3 Setting nonzero Ancestral Bonus will overrate crits because of ancestral procs. I don't know what actual number should be here.

5.4 There is separate EarthShield CritUltraOverheal coeff and separate spellcrit chance for Earth Shield. It's caused by ES mechanics.

6. Now you can see the spell efficiency table Avg.Heal/HPS/HPSC/HPM/MP5/MP5edge.
Spellcrit, haste, human reaction delays, are taken in consideration.



HPSC means Healing per second casting, it's different from HPS only for our HoT spells, HST and ES.
MP5 means MP5 required for infinite spamming this spell by cooldown.
MP5edge is MP5 when you get OO5SR(spirit) mana ticks even when spam spell as often as you can.

7. Now you can specify your typical fight duration in minutes in cyan cells.
Spreadsheet will calculate your average HPS over fighs. Manapool, I5SR MP5 and OO5SR MP5 are taken in consideration.



This table helps us to choose what downrank is optimal for actual fight. You can find some sudden and interesting facts here ^^

Yellow texts indicate maximal HPS for spell family, its best spell for this fight duration.
Brown texts indicate HPS in 1% range of maximal, dark background indicates spells in 5% range of maximal.
Coloring was intended just for visual usability, nothing special but easier to understand.

OO5SR mana regen was calculated on the assumtion that delays between spellcasts are distributed exponentially,
detailed math can be checked out at "5SR%" sheet.

HPS in the last line doesn't correspond to "normal" spell. This is result for precalculated mix of 2 spells spending together full manapool/mp5.
This mix if very demonstrative for haste effects, when your haste increases "mixed" spell increases both HPS and mana efficiency.
Notice that "mixed" spells often are the most efficient in family.

8. If you scroll sheet right you can see what actually each "mixed" spell means.



There are displayed two spell names and weights,
if you cast these spells in this proportion you will have results equivalent to "mixed" spell.

That's all concerning description


I'm open for dialog. If anyone needs more explaination or doesnt agree with something just ask me.
Also I'm interested what you think about spreadsheet, it may be worth to add some specific here.
If I miss something or something is incorrect tell me please, I will add/correct it.

Skyhoof, i remember you said something about T5/6 set bonuses. Could you post here what actually these bonuses do? I'm going to include it.

 

Beta

Thank you, very useful indeed.

Spreadsheet is totally line with results that I have got while chain spamming raid and pretty much confirms that haste stacking isn't only way to go. Obtaining over 3k HPS is totally viable both ways at this point of game and mainly comes down what tactics or healing setups are used on boss encounters.

Haste consumes your mana also faster making MP5 more important stat to have. Mostly if you have a shadow priest it isn't an issue but I would not personally recommend people totally skip MP5, specially if fights last longer than 6 minutes (and SP could die also, what then?).

So, a little bit haste, some +heal and then drop of MP5 with nice balanced mix and you have a great shaman.

Side note, did break today first time 2.9k HPS on brutallus, resulting over 1million healed and was wearing only 6% haste.

 

Pitbuller

With new HEP values "the best" gear look pretty solid.
Unbuffed values 2742 +healing, 131mp5 and 326 haste.
Also t6(chest + bracer + belt + shoes) is best of slot and it's easy to get bonuses.
There is warcrafter toon.

Slow, slower, shaman weapon.

 

Telesto

Since I, like many others, have been speculating alot in haste lately, I would very much like to see where you would consider [Battlemaster's Alacrity] to be amongst the trinkets. personally I would say it has alot to do with gear, since this trinket is obviously not for the pre-kara, or even the pre-BT, geared shaman. But then on the other hand, I see a trend raising amongst end game shaman, where this trinket is being prefered over your "best of the best trinkets". and of course not to forget the extra health you get from it, which is a very nice addition, due to the fact that raid damage(and thereby the need for stamina/health) is increasing rapidly along side instance difficulty. what are you guys thoughts?

 

Beta

It's nice trinket and considering using it on short fights, but the thing with CH is that, the most effective part of your heals are additional jumps. Are people gimping +heal too much and loosing then most of chain heal potential? What is the difference when you can cast 0.15s faster with +2300 vs +2700 heal buffed what comes to those additional jumps (CH4 and CH5)? In most cases primary cast produce some over healing, jumps not so much regardless how much +heal you have.

I'm not saying that haste is bad, it's excellent no doubt about it, but when focus has turned to recommend ZA amulet over Sunwell one I see that haste is valued maybe a bit too much over other stats.

About that trinket, I would value it like HEP numbers indicate, 80 or 88 HEP witch make Memento better purely healing wise.

 

Skyhoof

I reversed engineered the HEP values. Let's say we all agree that X, Y and Z are the best items in the game. Well, how would you have to value their stats to arrive at the conclusion? Then, I take those values for something like +healing and plug them into Binkenstein's Shamstats to derive the other values and to make sure the values are correct in relationship to one another. It's mainly a lot of trial and error.

I can't seem to unzip the spreadsheet at work. I'll try at home tonight. I wouldn't worry about the Tier 5 set bonuses. They really don't do much for our healing. However, the Tier 6 bonuses have a huge impact on both mana usage and healing output.

 

Bokomatic

Hmm... so we just need HEP equivalents for each stats. Right? Do you mean this here?

I checked Shamstats and found there calculations of HEP equivalents for each stat. Do you use these equivalency coefficients directly or make some euristic precalculations before?
Actually i don't understand clearly sentence 'take values for something like +healing'. What does it actually mean?

I've added link to zip archive to you, afaik rar is better but isn't so common as zip.
What does T6 bonus actually mean? (sry, i'm only in the beginning of T4 ^^)

 

Mishael

The T6 set bonuses are incredible. 2pc is 10% reduced mana cost for Chain Heal and 4pc is 5% more healing from Chain Heal.

 

galzohar

In my opinion that method is horribly flawed. HEP values are supposed to let you know which items are better, reverse engineering items that you think are the best to make the HEP values is not the way to do it. It's almost as bad as using the itemization cost of stats as your HEP system... Just the fact blizzard gives an item a higher budget and more stats doesn't automatically make it better. Some items are just made to be bad and just becuase they drop in BT/sunwell/whatever isn't an automatic sign of their goodness.

In general, looking at how much each stat actaully increases your ability to heal, using your current gear as a baseline, is the way to go. Yes it's rediculessly complicated but there's no way around it if you want a better HEP system than a "we just looked at the best items and made an HEP system that will make you choose those as the best items".

 

Coztomba

Very nice work Bokomatic. I'd be interested in seeing effect of 10 Storms 3 piece also with maybe a field where you can input how much healing / mp5 you loose with the 3 pieces that would deduct from your normal plus heal, or something along those lines. Really good work tho!

#edit: along with 2pc/4pc T6

 

Beta

I got the following HEP's for haste rating by using spread sheet:

2.2 for +2000 heal
2.5 for +2500 heal
2.8 for +3000 heal

I did compare CH5 1, 2 and 3 target HPS to gain these. A bit late tho and those are rounded up but someone should maybe confirm are they even close (I could honestly totally calculate it all wrong, not so good at math).

I double checked em by comparing how much +heal is needed to gain same HPS values and each time it matched, so does haste really scale better the more +heal you have or did I do something wrong?

 

Mishael

Absolutely it does. Conversely, +healing scales with the haste you have. Increasing either stat will cause the HPS increase gain from the other to increase.

 

Bokomatic

Ok. Concerning 3pcT2, 2pcT6, 4pcT6.

Current formulae for CH are:
Mana = BaseMana * TidalFocus - RelicMana
HealJump0= ( BaseHeal + SpellCoeff * DownrankCoeff * ( Healing + RelicHealing ) ) * Purification * ImpCH
HealJump1= 0.5 * HealJump0;
HealJump2= 0.5 * HealJump1;

For my relic & 0/5/56 talents it's:
Mana = BaseMana * 0.95 - 20
HealJump0= ( BaseHeal + 0.7143 * DownrankCoeff * Healing ) * 1.10 * 1.20

Changes for 3pcT2:
HealJump1= 0.65 * HealJump0;
HealJump2= 0.65 * HealJump1;

Changes for 2pcT6:
Mana = 0.9 * BaseMana * TidalFocus - RelicMana

Changes for 4pcT6:
HealJump0= 1.05 * ( BaseHeal + SpellCoeff * DownrankCoeff * ( Healing + RelicHealing ) ) * Purification * ImpCH

Do I write formulae right? If you confirm them I embed them into spreadsheet.

Actually i'm not sure concerning:

Mana = ( 0.9 * BaseMana * TidalFocus ) - RelicMana
or
Mana = 0.9 * ( BaseMana * TidalFocus - RelicMana )

 

Mekias

2 piece Tier 6 is:

Mana = (( 1 - 2pcT6 - TidalFocus ) * BaseMana ) - RelicMana

2pcT6 = 10% = .1
TidalFocus = 5% = .05

Mana = (( 1 - .1 - .05) * 540) - RelicMana
Mana = (.85 * 540) - RelicMana
Mana = 459 - RelicMana

I haven't tested the RelicMana (Totem of Living Water) portion since 2.4 but it used to lose some of it's potency with 2pcT6 and TidalFocus. Now it's supposed to be a flat 20 mana off at the end of the calculation.

 

Bokomatic

I've tested Mana = (( 1 - TidalFocus ) * BaseMana ) - RelicMana. It's right.

I don't sure about Mana = (( 1 - 2pcT6 - TidalFocus ) * BaseMana )

Get a simple example from 1st page of this thread.
BaseMana=435;

Mana(TidalFocus) = ( 1 - TidalFocus ) * BaseMana = ( 1 - 0.05 ) * 435 = 0.95 = 413.25 (as in the table)

(1) Mana(TidalFocus+2pcT6) = ( 1 - 2pcT6- TidalFocus ) * BaseMana = ( 1 - 0.1 - 0.05 ) * 435 = 0.85 * 435 = 369,75 (372 in the table)

(2) ManaCH4(TidalFocus+2pcT6) = ( 1 - 2pcT6 ) * ( 1- TidalFocus ) * BaseMana = ( 1 - 0.1) * ( 1 - 0.05 ) * 435 = 0.95 * 0.9 * 435 = 371,925 (372 in the table)

So ( 1 - 2pcT6 ) * ( 1- TidalFocus ) * BaseMana seems right.

I don't know where Daidalos/Skyhoof got these data tho.

 

Daidalos

I came up with the calculations after basically guessing in what order the bonus would be applied and then checking it with in game tooltop display since I had 2 pc t6.

Restoration Shaman calculations:http://spreadsheets.google.com/ccc?k...IWv3sPmcYnEJUg

 

Skyhoof

Actually, I didn't explain that very well. Yeah, only reverse engineering would be horribly flawed. First, you run Sham Stats, which mathematically determines stat weights using the equations that Binkenstein created. However, the number outputted depend on the numbers inputted (healing, crit, etc). So, I run several different scenarios through the spreadsheet and then average the results. Even then the final numbers require some tweaking, which is done through reverse engineering a few items that are a very close call.

Yes, using your current gear as a baseline is probably the ideal way to evaluate a new piece of gear. However, it's not that complicated using Shamstats. It's what the spreadsheet is designed to do. However, many people want to be able to look at two pieces of gear and have a rough idea of which is better -- which is what an HEP system does. It's not perfect but it's a good general guideline.

 

Bokomatic

I've added 3pcT2, 2pcT6 and 4pcT6 bonuses in spreadsheet.

Download: RSSSS_v109.rar (77kB) or RSSSS_v109.zip (112kB)

If you want to enable bonus you should check 1 instead 0 at <0>Talents sheet.

 

Skyhoof

The Wiki for Resto Shaman is now live and available at the
Theorycrafting Think Tank: Shaman: Restoration

Thanks to all who helped with editing and refining the article, especially Daidalos, Malan and Binkenstein.

The Theorycrafting Think Tank is designed to be the definitive source of information. Post there only if you have a correction or update.

If you have a question, or want to discuss an idea, you can still post in this thread. However, you should also feel free to create a new thread in Class Mechanics to discuss some specific aspect of Resto Shamans.

I have removed most of the information in the first post of “How to Heal Like a Pro,” since it duplicates what is in the Wiki.

 

Mekias

It's easily tested in game. I just tested Totem of Living Water and apparantly it wasn't changed for patch 2.4. Here are my numbers:

Rank 5 w/ ToLW = 494
Rank 5 w/ 2pT6 = 459
Rank 5 w/ 2pT6 + ToLW = 442

Rank 4 w/ ToLW = 394
Rank 4 w/ 2pT6 = 369
Rank 4 w/ 2pT6 + ToLW = 352

So currently it's:

( 1 - 2pcT6 - TidalFocus ) * ( BaseMana - RelicMana)

 

Daidalos

Hmm I thought it had changed in 2.4 I'll try to remeber to check when I get home.

Restoration Shaman calculations:http://spreadsheets.google.com/ccc?k...IWv3sPmcYnEJUg

 

Bokomatic

I rechecked ToLW, you are right, spell doesn't cost 393=513-20, it costs 394.
I cant check T6 bonus ((
Ohh i so want to be able to check it ))

One comment to your formula and your data.
There a several exclusions, i.e.
Rank 4 w/ 2pT6 + ToLW =( 1 - 0.10 - 0.05 ) * (435 - 20) = 352.75
Rounded it is 353. Your result is 352.

First i thought that there is some better formula, but all of them that i could imagine give us worse results.
So i deduce this formula is right, but game round down manacosts. In this case your formula completly fits to experimental data.

P.S I've put this formula (with rounding down) in spreadsheet and corrected link in previous post.

 

Karede

On Sunday we made our first attempts at the Council. We ended the night with them at 12% hitting the enrage timer, largely because the priest was able to get 8 CoH off healing them up for just over 800K in total health and we had lost a few DPS about half way in. As is clear, we were having some trouble with our interrupts, especially through the BoP.

I was wondering if it's advisable to have a resto shaman (we raid with 2 or 3) on interrupting duty with a rank 1 Earth Shock along with their healing duties. I mean, it seemed pretty clear to me that the key to that fight is interrupting the heals. But at the same time, I worry about having one of the resto doing that, as I feel like we (at least I know that I was) spend a good amount of time moving to avoid the Blizzard and Flamestrike. With the very limited range of shocks, it seems like it might not be feasible to have a healer trying to interrupt that.

Anyone have any experience with this?

 

Daidalos

Looks like I need to take another look at my spreatsheet again and fix it too . I'm off by 2-3 mana atm I'll try to confirm in game and fix this when I get a chance.

Restoration Shaman calculations:http://spreadsheets.google.com/ccc?k...IWv3sPmcYnEJUg

 

Daidalos

I do it all the time. The resto sham cannot be the only interrupting (not enough hit) he can however be helping if another is out of range or gets resisted. This shaman should be given a slightly easier healing assignment since he will be spending a little less time healing. Make sure he has stopcasting and focus interupt macro for this and its not that hard.

Restoration Shaman calculations:http://spreadsheets.google.com/ccc?k...IWv3sPmcYnEJUg