Jump to content


Photo

TreeCalcs: WrathCalcs, respecced


  • This topic is locked This topic is locked
42 replies to this topic

#1 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 07 October 2009 - 08:33 PM

Ok, so since I'm playing Resto most of the time these days, and I already had this great Moonkin spreadsheet sitting around, I retooled it for Resto purposes. For those of you who've never used WrathCalcs, the focus is on providing accurate marginal values of stats/talents. The basic idea is to have accurate numbers on what all our spells do. Obviously we're not always playing to maximize some numerical output like DPS are, but good spellcasting decisions rely on knowing how much each of our spells do, and perhaps more importantly, good gearing decisions require knowing what each stat does for all of our spells. For now, it's focused on raw HPET of all healing spells, but I can refine that if it becomes important for anything.

Input stats/setup on the front page. Next to each stat/talent/set bonus, etc., there will be marginal HPS/MP5 benefits. For stats, it shows the benefit of having 1 more of that stat. For talents and other bonuses, it shows a) if you have no points in the talent, the value you would gain from one point, B) if you have points, the value you currently gain from those points (i.e. the amount you would lose by dropping them).
All of these computations are derived from a spellcasting cycle you pick in the bottom-right corner. For now I lazily just put in a Rejuv5-WG one to get going, but I'll add more as I think about useful spellcasting scenarios to model.

Since healing, unlike DPS, isn't quite as focused on the singular output from one rotation of spells, the second page becomes a bit more important. It lists all spells, the HPS/MP5 of each, and the amount which that HPS/HPM of each would increase from added points of spellpower/haste/crit. The top box on this page will also contain a bit more info on spellcasting cycles once I come up with them.

Any suggestions are welcome. Basically, the framework is there to model any combination of spells--tell me which ones you think would be useful to add. Also, ignore all the residual Moonkin stuff; it will just get worked out as I replace it.

Updated version lives here now:
http://elitistjerks....sion/#TreeCalcs

#2 Tecton

Tecton

    Soda Popinski

  •  Patrons
  • 3,946 posts

Posted 08 October 2009 - 12:13 PM

Excellent idea, will be useful to get more into the guts of resto rotations.

Perhaps add options to analyse the effect the 3.3 changes will have on our rotations (new glyph for hasted Rejuvenation, one less Rejuvenation tick per cast, tier set bonuses)?

Also, shouldn't Healing Received & All Haste have an HPS value attributed in Raid De/Buffs? Focus Magic is showing as a negative HPS value if you don't have the buff, I think that should be a zero value with extra HPS being attributed if you do have it? Just a quick look during my lunch break, will have a closer look tonight.

#3 Erdluf

Erdluf

    Great Tiger

  • Members
  • 960 posts

Posted 08 October 2009 - 12:43 PM

Looks pretty good. I didn't realize Tranq benefited from EmpRej, but logs confirm you are right.

With 4t9, Insightful should provide some HPS (21 Int, and int is showing .32 HPS with 4t9).

There is now an 8% BoK (Drums). Probably an issue on Wrathcalcs, too.

If tabs like Regrowth would show min/max non-crit, min/max crit, and min/max Living Seed, it would be easier for people to compare the spreadsheet with logs, and find bugs (either in the spreadsheet, or in their input to the spreadsheet).

Glyph of Healing Touch is additive (subtractive) with GoN. So HT formula should be
... (1+.02*GoN-if(GHTActive,.5)) ...
Reference is here.

For tank healing rotations, I like to assume that on average, Rj and Rg is refreshed a bit late (so perhaps the Rg is cast, on average every 28s, and Rj is cast, on average, every 19s, currently). The only way to not be late is to sometimes not be casting (or to clip a HoT).

If using a Lb3Slow, I assume that the initial cast (after the bloom) is a bit late, and that one of the two "refresh" casts is an entire tick too early.

If someone has Glyph of Regrowth, the tank cycle should be 8 ticks and roughly 25s.

In your first post in this thread, you misspelled "haste" as "hit."

#4 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 08 October 2009 - 05:35 PM

The 3.3 Rejuv-haste idea wouldn't do anything in this model, since it doesn't change the HPET of Rejuv casts. The 2T10 I can put in once we know the numbers. 4T10 will basically just a be a multiplicative buff to Rejuv--doesn't change too much materially.

Right now the All Healing Recieved raidbuff doesn't do anything since it assumed you have it from your own Tree form. I'm going to take that out; then you'll be able to see the individual effects of ToL mana reduction, iToL, ToL aura, and MSS more easily.

Erdluf:
The meta/Spark lines only show the proc/bonus on the item, not the passive stats gain. Basically, things are always set up so the stat values in the top-left match those you should see on your Armory--makes it much easier to use. So the "Insightful" line right now only shows the gain from the proc (also, I think I have to fix the way the "Ember" one works).

Right now I'm approximating slightly on HoT reapplication times being instantaneous. It's the same assumption that's always been made in WC--it would probably complicate things a lot to try to estimate the gap. I'll consider adding half a GCD to each. LB stacking already accounts for clipped tick--I should add it somehow to Regrowth.

#5 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 08 October 2009 - 10:26 PM

Updated--fixed BoK, Ember meta, LB rolling, Tree form, Glyph of HT.

#6 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 13 October 2009 - 10:34 PM

Added a version reflecting known 3.3 PTR changes.

#7 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 17 October 2009 - 09:54 PM

Updated--put Rejuv tick back in, added a simple model for T10 set bonuses and T10 idol.

#8 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 24 October 2009 - 02:49 AM

Updated 3.3 sheet for newest GOTEM functionality.

#9 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 31 October 2009 - 01:39 AM

Apparently the 2T10 is datamined to be "reduces WG decay by 30%." I put in my best guess as to how that will work numerically. I'm pretty sure it's a pathetic set bonus regardless of how precisely it works.

#10 Kirbie44

Kirbie44

    Piston Honda

  • Members
  • 152 posts

Posted 02 November 2009 - 04:09 PM

When I plug in Ember to the meta, it gives me like a 11.94 HPS increase. I understand that the 25 SP from this is figured in the above SP part, but how does the 2% int figure in as HPS? Or what part is altering it. Chances are I am missing something big, but I didn't dig through the spreadsheet TOO much, as it is a little easier to ask how you set it up.

EDIT: I do not have Lunar Guidance on the spreadsheet.
Artificial Intelligence usually beats real stupidity.

#11 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 02 November 2009 - 04:23 PM

When I plug in Ember to the meta, it gives me like a 11.94 HPS increase. I understand that the 25 SP from this is figured in the above SP part, but how does the 2% int figure in as HPS? Or what part is altering it. Chances are I am missing something big, but I didn't dig through the spreadsheet TOO much, as it is a little easier to ask how you set it up.

EDIT: I do not have Lunar Guidance on the spreadsheet.


Do you have 4T9? Then the int gives a bit of crit. Can't think of anything else offhand.

#12 Kirbie44

Kirbie44

    Piston Honda

  • Members
  • 152 posts

Posted 02 November 2009 - 04:35 PM

Do you have 4T9? Then the int gives a bit of crit. Can't think of anything else offhand.


entia non sunt multiplicanda praeter necessitatem

Aye, that is exactly what it is. But isn't that intellect already figured into the base stats at the top of the page?
Artificial Intelligence usually beats real stupidity.

#13 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 02 November 2009 - 04:41 PM

entia non sunt multiplicanda praeter necessitatem

Aye, that is exactly what it is. But isn't that intellect already figured into the base stats at the top of the page?


Yes. But for all talents/bonuses that add to your "armory stats" (Ember, iMotW, Lunar Guidance, Living Spirit), the sheet jumps through some hoops to still accurately figure out how much you're getting from the talent.

#14 Tupsi

Tupsi

    Glass Joe

  • Members
  • 9 posts

Posted 05 November 2009 - 01:53 PM

Hello

I am trying to figure out what and where to put into the MP5 cell (C9) as inside it has a formula saying

=195*15/(45+10*'Rejuv Calcs'!C15)

(3.3 sheet has)
=21 + 241*15/(45+10*'Rejuv Calcs'!C15)

so I would assume that "195" or "241" are you counted MP5 gear numbers and I would have to put mine at that place replacing only these numbers and leaving the rest of the forumula intact?

You might think about moving that forumula to another cell and leave C9 just for the raw number for confused users like me in the future :-)

#15 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 05 November 2009 - 07:03 PM

Hello

I am trying to figure out what and where to put into the MP5 cell (C9) as inside it has a formula saying

=195*15/(45+10*'Rejuv Calcs'!C15)

(3.3 sheet has)
=21 + 241*15/(45+10*'Rejuv Calcs'!C15)

so I would assume that "195" or "241" are you counted MP5 gear numbers and I would have to put mine at that place replacing only these numbers and leaving the rest of the forumula intact?

You might think about moving that forumula to another cell and leave C9 just for the raw number for confused users like me in the future :-)


No, I was just estimating my own MP5 return from Sif's/Show of Faith. Just put into that box whatever is your static MP5 value from gear bonuses.

#16 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 05 November 2009 - 09:07 PM

Fixed a bug with value of Genesis.

#17 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 05 November 2009 - 09:20 PM

Something feels off about the math though. If you count up the losses from SP, crit, haste, INT, and SPI for losing 4T9 from a situation of 3500 SP, 400 crit, 500 haste, 1100 INT, 1100 SPI, you get a loss of 2515 HPS. But the spreadsheet only calls this a loss of 1700 HPS for dropping 4T9. Whatever the exact truth is, I think the point stands: 4T9 is not to be dropped lightly.


The latter computation is correct. Think about what you're doing in the former--it doesn't quite make sense. The marginal value for, say, spellpower is the value of one point of spellpower in your current setup. It is not the average value of each point of spellpower from 0 up to your current value. When you change something, say T9, you can't obtain the total change in the contribution from spellpower by multiplying the change in the marginal value by the total spellpower. You'd have to integrate the change in marginal value over the whole range of spellpower from 0 up to your current value.

I know this explanation isn't totally clear, but I think you're facile enough with calculus see what I'm saying.

#18 Erdluf

Erdluf

    Great Tiger

  • Members
  • 960 posts

Posted 30 November 2009 - 06:36 PM

treecalcs_091105_3.3.xls

Spreadsheet is showing 38 MP5 for . In 3.2, MP5 items got buffed. Wowhead is now showing 45 MP5.

WTB: A spreadsheet function to update gear stats from online databases.

#19 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 30 November 2009 - 06:47 PM

Hey, if anyone wants to design one of those fancy frontends for both WrathCalcs and TreeCalcs (they're basically the same sheet) where you have dropdowns for everything and stats are automatically imported, etc., that would be great! I know that's the norm these days and these sheets aren't at all user-friendly in comparison. But it's a lot of development work I'm probably never going to get around to doing myself.

#20 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,540 posts

Posted 10 December 2009 - 06:04 PM

New version up with Ephemeral Snowflake and a few other minor things.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users