Jump to content


Photo

WrathCalcs 6.0


  • Please log in to reply
13 replies to this topic

#1 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 27 May 2014 - 08:01 PM

I've gotten pretty far with this since they announced the new Balance rotation:

http://us.battle.net...4-23-2014#druid

 

This is still a tool for assisting in theorycraft and planning how to play the new rotation. Since all of the numbers are surely subject to change, don't read too much into the details of talents or stat values, though the general patterns should be right unless the relative coefficients of spells change dramatically. I'm pretty happy with it as a model of what we know so far.

 

Without further ado. When you open the sheet (attached), you'll notice a few things.

--The gear selector and all its associated accoutrement have been removed. It was simply not worth the overhead of building and maintaining (or rather, of having Tecton maintain) an entire WoW gear profiler contained in this sheet, when many such tools already exist. The profiler was a bigger project than the actual DPS model.

--What I recommend is to use the Wowhead profiler: http://www.wowhead.com/list . If your either import your profile from battle.net or make a custom one, you'll see a column called "Item stats" that contains your stats from gear before any buffs. Copying these into the relevant boxes on the Main page should be all you need to do. I'd like to include a simple Excel widget that imports these from the Wowhead profile automatically--if anyone can help with this, please let me know.

--A huge upside of this is that the sheet is much smaller in size (most of the file size previously was the gear list). This means it runs faster. In particular, I turned on automatic computation of the data tables (on my computer it happens instantly), so you don't have to fiddle with pressing F9.

 

As far as substantive stuff, I'm quite happy with the model, based on what we know so far. I'll probably make another Balance thread soon for general discussion of the rotation, but here are a few details on how it works.

--It assumes you cast MF once per cycle and SuF twice per cycle (9.2 seconds apart, the minimum possible for 100% uptime).

--The Lunar half of the cycle is modeled for two cases: 1 Starsurge used and 2 Starsurges used. The Solar half is modeled for 0 and 1 SS used.

--It computes your average number of SS procs per cycle based on your DoT crit and haste. It then averages together the cases listed above, based on how often you have extra SS casts to use. It does account for the possibility of being forced to use SS suboptimally due to the randomness of procs.

--When using 1 SS in Lunar, it casts SS-SF-MF-SF at the highest part of the Lunar phase. With two SSs, it casts SS-SF-SF-MF-SS-SF-SF, with the MF centered at the peak.

--When using an SS in Solar, it cases SS-W-W-W-SuF at the highest part of the phase.

 

--Incarnation and FoN are unchanged from before (I'm assuming for now that Incarnation affects all damage). SotF is easy to model.

--L90 talents no longer add DPS.

 

L100 talents:

--Stellar Flare is not in since we don't know its coefficients.

--Balance of Power assumes that you can extend MF and SuF indefinitely. So, no GCDs to refresh, no DD damage, and a perfect Eclipse snapshot on both.

--Euphoria assumes you refresh MF and SuF once each in the 20s cycle.

 

--Readiness is assumed to affect only CA and Incarnation. There's a lot unknown about how this will work.

 

--There are no Glyphs in yet. The only Perks that are in are 20% damage to SF/W/MF/SuF. Also, the passive listed in the patch notes adding 100% to the MF DD.

Attached Files


Edited by Hamlet, 27 May 2014 - 08:12 PM.


#2 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 02 June 2014 - 04:04 PM

Update for a few things.

--Added Stellar Flare (coefficients seem to be placeholders).

--Updated for 40s cycle length and DoT duration.

--Updated for new Eclipse curve that has a brief pause at the top.

Few other minor things throughout.

 

 

Attached Files



#3 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 04 June 2014 - 04:33 AM

Revisions to the model.

 

 

Attached Files



#4 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 04 June 2014 - 10:10 PM

More named cells (or attached cell comments) on the Rotation tab please.  I don't yet have a good enough feel for this stuff to understand all that is going on.

 

Spells tab: It seems that for empowered casts, Bliz is more likely to code t/1.2 than t*0.8, but I don't know that for sure.

 

I'm thinking rotations are going to include rules approximately like

  • Don't cast MF/SuF in the last 5 seconds before CA becomes available.
  • Do cast MF during CA whenever the cast will clip at most one of MF or SuF.

Those rules almost guarantee at least one MF during CA (and on rare occasions two MF during CA).  That might mean that out-of-CA DoT casts will be a bit less (perhaps 7/8) than you'd otherwise predict.



#5 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 05 June 2014 - 01:26 AM

Yeah, I need to comment that better. If I still don't have alpha when I get home and wind up spending some time on the sheet I'll try to say more here.

 

Spells: Yeah, will just check what they did in-game when I can.

 

I didn't focus on details of DoTs in CA yet. It's odd because normally you MF once every 40s, and you only have one good opportunity every 40s. But when you CA, that results in 15s of MF downtime unless you MF an extra time (which will get clipped after 15s, or equivalent). The DPET of MF is (currently, and probably will be) enough that casting one for 15s duration is fine. So I think it's fair to assume one DoT cast during CA always, and no reduction in MF casts outside. Similar for Sunfire since there too you're normally casting enough to cover each rotation exactly. So the one cast during CA will give 15s of added uptime, but you don't actually cast one fewer outside (but there will be some edge effects due to snapshotting).



#6 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 05 June 2014 - 05:06 AM

Rotation tab:

 

--Shaded stuff is unused right now.

--There are 6 columns of computations. 2 for the Lunar half and 3 for the Solar half, based on how many SSs are used. They'll be averaged together later. 1 column for CA.

--The first block of rows in each column should be pretty easy to parse. How much time in each case do I spend on each spell: DoTs and SS, then empowered nukes, and the remainder to umempowered nukes.

--Basically all of the rest is estimating how much mastery bonus to allocate to each spell.

 

--Moonfire/Sunfire: Bonus is hand-set based on when the cycle we expect to use them. Sunfire is fudged a little since it's snapshot with two different bonuses--I compute both bonuses and then average them with the higher one being weighed 2/3.

--SS and buffed nukes. First assume that these spells are cast in a window that's evenly centered around the peak. The mean bonus in that window is computed by integrating the Eclipse bonus over the window (row "Mean bonus in window"). Then that average bonus is applied to SS and empowered nukes. There's one added detail in Solar--Moonfire is removed from the average before it's applied to SS and Empowered Starfire (tiny difference in reality, hardly worth worrying about).

--Unbuffed nukes. These get the "remainder" of the total mastery bonus that's not used by other spells. It's done this way, rather than computing a bonus directly, to ensure we're never cheating by applying more (or less) total Eclipse bonus to all spells than is actually available during a cycle. So compute the integral of bonus*time over the whole cycle, and then compute the total bonus*time used by spells so far (row 54 right now), and then subtract to find how much bonus is left to apply to unbuffed nukes.

 

--After this, all spells' damage is added up with their respective Eclipse bonuses.

--Then, the total amount of SS actually available per cycle (based on the Shooting Star proc rate) is used to see how much of each Lunar/Solar column to include in the final average.



#7 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 12 June 2014 - 08:22 PM

--Changed Eclipse amplitude to 105

--Added Versatility (using same rating conversion as crit, I assume it will be lower eventually).

--Increased SS proc rate to 5%. The high amount of SS causes a few modeling issues. It's generally okay, but I'll refine that some if the SS rate stays high.

--Added "mastery attunement"

--New Starsurge coefficient.

 

Attached Files


Edited by Hamlet, 12 June 2014 - 08:23 PM.


#8 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 20 June 2014 - 02:30 PM

I'm not sure of the exact mechanics, but assume that the Starsurge timer is set to 30s whenever you drop below full (or it just went off and you weren't full), and also that Starsurge remains an extremely powerful spell (substantially more damage than our other nukes, and much higher DPET).

 

It would mean we'll almost never want to be holding on to more than one charge (for fear that shooting stars would reset the timer).  That means that we will often be using Starsurge at "poor" times (bad spots on the eclipse cycle, or wasting Empowered charges).

 

I had been thinking we would open with something like

   CA MF Flare SS SF ...

but perhaps it will be closer to

   CA MF Flare SS SS SF ...

and at higher crit levels, at least the first SS might go before Flare.



#9 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 21 June 2014 - 03:39 AM

Nothing special in today's spellinfo, but added the set bonuses (kind of--didn't separate CA and Inc yet, so 4P is affecting both):

 

---

 

That's right about SS--any time you spend at 3 is waste. It's not a huge waste to momentarily hit 3 (if you spend one right away, you only lose the 1 GCD or so of charge time). But it's a reason to stay low to avoid it. One reason I don't like the much higher SS proc rate they're putting in--less opportunity to hold SS, and more closely resembling the current version where you have to use procs right away.

 

I don't think you'd want to waste the Empowerment from your first SS, but haven't looked with any detail.



#10 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 21 June 2014 - 05:06 AM

I don't see a link for 140620 version.

 

Looks like 140612 version is missing 5% base crit (all classes, but some, like ferals, get another 10%) and 3% Versatility raid buff included in MotW.



#11 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 22 June 2014 - 07:00 PM

Whoops, here's the update. It does have base crit and the new raid buffs.

 

 

Attached Files



#12 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 03 October 2014 - 03:52 PM

http://iam.yellingon...com/wrathcalcs/

 

Updated--really not that much major has changed since the last time except for lack of DoT snapshotting.  But got talents, set bonuses, etc. up to date.



#13 Tarm

Tarm

    Glass Joe

  • Members
  • 5 posts

Posted 04 October 2014 - 03:41 AM

Thanks for the WrathCalcs update! I always enjoy playing around with these.

I noticed a small issue with using the Item Stats column from Wowhead to fill in the pink boxes.

WrathCalcs expects you to input your unbuffed total spell power in the box, but in the Item Stats column, it only gives you the spell power that is directly on your gear, i.e. only the spell power from your weapon.

You could change the formula in C7 of Basic Calcs to be just "=Spellpower + ExtraSP" so that we can use the spell power value from the Item Stats column. Alternatively you could just make it more clear that the value in the spell power box should be "your spell power from gear plus your intellect from gear".


Edited by Tarm, 04 October 2014 - 03:54 AM.


#14 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11576 posts

Posted 05 October 2014 - 12:31 AM

Thanks--I should change it in one of those ways for next update.






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users