Archived

This topic is now archived and is closed to further replies.

Hamlet

[Resto] TreeCalcs

188 posts in this topic

Getting a thread going for this. Short post for now, probably more when I have time. Here's an update with some stuff I was tinkering with (all on the talents/tables page). It's not dressed up yet, just some new numbers, and in general I have to decide how to present everything.

These are the two things TC wasn't that great at analyzing:

1) Effect of changes in total mana pool on longevity.

2) Correlation between mana consumption and added healing.

Slourette had done some interesting math on (2) a little ways back in the Resto thread. He uploaded a version of the sheet with a whole added framework for total healing done over a fight of fixed length. I liked where it was going, but I was hoping not to have to add a bunch of new machinery to the sheet. I think what I have here works pretty well. Briefly, it varies the % spell usage of Nourish vs. Rejuv to see the effect on mana consumption and on HPS. From that it computes the HPS gain from added MP5, assuming you vary spell usage to hold mana consumption constant. I added another column next to the scaling factors for stats, for now, to show what happens if you take this into account. They're pretty similar to the "sustained healing" stat weights people have been throwing around.

For (1), I just compute the MP5 equivalence of increases in starting mana pool, based on the current time to OOM. Those values are currently added to the MP5 scaling factors for Int/Furor/HotW/etc.

I think this all pretty good, but I want to find a way to nicely present everything without adding more clutter to the page.

Share this post


Link to post
Share on other sites

On the "Gear Setup" sheet, I'm getting Error 502 in all the equipment slots (i.e. A56). I imagine it has to do with using OpenOffice as a search brought up similar issues with this program in the Balance druid spreadsheet thread. Any solution to this, or do I need to hunt down a copy of Microsoft Excel?

Share this post


Link to post
Share on other sites

I think there's a mostly-working fix in WrathCalcs right now. It should be moved to TreeCalcs as well once it's all done.

Share this post


Link to post
Share on other sites

Yeah, sorry guys, I'll get this updated with the OpenOffice stuff as soon as possible.

Share this post


Link to post
Share on other sites

You shouldn't need to use tables for the HPS/MP5 computation.

For the HPS/MP5 you get by using extra mana to convert Nourishes into HT

HPS/MP5 = (HT_HPET - Nourish_HPET) / (HT_MP5 - Nourish_MP5)

Using HPET and MP5 values from the Rotation page, I'm matching your table results when I use HT or Rg fillers. I'm getting values that are slightly off when I use a Rj filler. I don't know why the Rj numbers don't match.

Edit:

For many (most?) resto's, Rejuv without symbiosis is more efficient than Nourish with symbiosis. If that is true for you, and if you are mana constrained (and have enough targets for Rejuv), you should only cast Nourish to refresh Lifebloom. The mana you spend on any other Nourish would be better to save for a Rejuv cast.

Otherwise, it is better to spend time not casting (bandage or melee). This shows up in column N of the main tab, where "Cast Delay" shows an increase in HPS.

If you don't use Nourish as a filler, your HPS/MP5 value is just Rejuv's HPS/MP5 (about 1.63, for the spreadsheet as downloaded).

Share this post


Link to post
Share on other sites

Those are both good points.

The slight discrepancy with Rejuv is because is changing the Rejuv frequency changes Revitalize proc rate (and if I implement it, Nature's Bounty uptime). Not sure if that's worth having another table in there. It is only one table though.

The spell usage issue will be interesting as soon as the patch hits. Right now you barely have to stop casting Rejuv between trinketed MTT and the very low cost, so I don't think about this (esp. since I'm playing Balance most of the time). For now I'll just add in the new tech with the current model, but it shouldn't be too hard to adapt if we change the spell usage model in some way. It would very easy to change the filler option from "Nourish" to "do nothing." I could add in slightly more refined Rejuv/Rg/HT filler options too, but I figure most people are using HT/RG outside of Clearcasts so rarely that it's not worth trying to estimate.

I think I will have to add a third scaling column for everything, to give the "combined HPS+MP5" score, since I don't really want to eliminate either the current HPS or MP5 columns.

Share this post


Link to post
Share on other sites

Update with the main page dressed up a little to take the above into account. I also change the individual spell stat scaling weights to something that might be more informative--it's raw HPET gain for each spell, with the user-input Mastery usage percentages taken into account.

TreeCalcs 110203 PTR.xls

Share this post


Link to post
Share on other sites
but I figure most people are using HT/RG outside of Clearcasts so rarely that it's not worth trying to estimate

From the view of 10 man hc encounters you really do often need to chaincast HT on tanks (at least with our setup) - so it would be a nice option for me, if you implement it.

Wyrmbraker is nealry pure HT spam in the beginning. Maloriak there are some situations where you need to HT. OK, Chimaeron is a special encounter to healers. Buton the encounter we are working on, Chogall, it's also necessary to use HT quite often.

Share this post


Link to post
Share on other sites

Thanks for the Sheet, just two questions:

- to remove a reforgement from an item, I just choose nothing in the "into" box ?

- I sometimes receive 2x Dark Intent (the buff stacks), could you add this as an option in the buff-section ? Just curious about the new breakpoints then.

- Does "Innervate Someone eles" takes into account that you switch Ina with another Druid ? Or is this option missing ?

Share this post


Link to post
Share on other sites

Thank you again for this work!

I think there is problem on the ratio INT/MP5 (MAIN tab, O7)

Intelligence gives more mp5 than spirit?

+1 INT -> 1.08 MP5

+1 SPIRIT-> 0.66 MP5

In your formula, you have integrated "longevity", which explains this bizarre result.

Maybe he should add another column "longevity"next to MP5.

Personally column "Combined' isn't interesting information.

Share this post


Link to post
Share on other sites

lolofora,

I agree that the short "time to oom" for the downloaded sheet means that the mana contribution from Int is overstated. However, since we don't have a lot of gearing opportunities to trade Int for Spirit (outside of trinkets) the error may not matter too much.

If you wanted to know the relative mana contribution of Int and Spirit on a five-minute fight, I think you can get good numbers by adjusting "casting delay" (M11 on the main tab) until OOM goes to five minutes (for the sheet as downloaded, casting delay = 0.32), and then hitting F9 so that tables get updated.

I argue that the "Combined" column is interesting

- It gives treecalcs a quantitative way to measure the relative value of mana and throughput.

- In LK, mana was more of a hard-cap stat. If you didn't have enough, it was the best stat. Once you had enough, it was worthless. In Cata, either statement is often false.

As an example, we can use TreeCalcs to determine how long a mana-constrained fight needs to be for Genesis to be better than Furor. I'd encourage discussions that improve that quantitative model. Even with flaws, I find that capability interesting.

Share this post


Link to post
Share on other sites

Hmm, yeah, I should put some better gear in the default sheet. You do have to turn Rejuv frequency down to get reasonable longevity post-patch though, as Rejuv cost is going back up, WG expenditure is going up, and MTT gains are smaller. We'll have to use Nourish (or almost equivalently, nothing) a significant portion of the time. Until I change the model somehow, adding in some casting delay is a good way to simulate doing nothing, but it really changes very little from just casting Nourish.

I'll add in another option for 30% Innervates (swapping with another Druid), but other than that I can't think of any ways it's underestimating longevity.

Share this post


Link to post
Share on other sites

Is it possible to add empty value to "Reforge from" combo boxes for all slots, as currently it's not possible to reset reforges? You can select empty value from "Reforge into" combo box but stat you are reforging from will still be reduced.

Share this post


Link to post
Share on other sites

I'll be catching Treecalcs up in the next week or so, including the option requested.

Share this post


Link to post
Share on other sites

Maybe I am just blind, but I have been inputting my gear into TreeCalcs, and I do not see the option for Blazewing's Furious Kilt under the Leg section.

Share this post


Link to post
Share on other sites

Might have missed one. Depending on how comfortable you are with Excel, you can go to gear list page, insert a row in the "Legs" section, and fill in the info using the same template as the other items. Once it's filled in the list, the dropdowns will update automatically. Otherwise, I'll try to remember to put it in at some point.

Share this post


Link to post
Share on other sites
PTR Lifebloom and Efflo changes. Haven't done anything with Tranq. Assuming no mastery double benefit for Efflo.

Based on my testing on the PTR, it is acting as it always has. It does NOT double benefit from mastery. It still benefits due to SM benefiting, but the "HoT" itself doesn't.

Secondly, the current Efflo "HoT" can crit per tick, but I've been unable to get the new one to crit at all.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.