Jump to content


Photo

[BETA] Balance Discussion


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

#21 dreslav

dreslav

    Glass Joe

  • Members
  • 7 posts

Posted 15 November 2010 - 07:36 PM

The last L85 sheet I uploaded was here:
Summing up. Here is, I think, the optimal socketing flowchart.
Assuming blue gems are better than yellow (i.e. Spirit is worth more than haste, and the hit cap is not a problem):
1) Let N be the total number of sockets you have. Let R be min (number of red+prismatic sockets,(N-1)/2).
2) Put R red gems into red/prismatic sockets.
3) If there are still red/prismatic sockets left over, put purple in them.
4) Let Y be min(R, number of yellow sockets).
5) Put Y green gems into yellow sockets.
6) Put R-Y blue gems into blue sockets. //you're now at red/blue parity and have at least one blue or yellow socket left.
7) Put one blue or green gem into one socket. Meta requirement now satisfied.
8) For each remaining pair of blue/yellow sockets, socket one blue and one orange gem.
9) If yellow are left, fill with half green and half orange. If blue are left, fill with purple.

Hmm, this came out worse than I thought. I'm going to see if I can clean that up into simple rules. And this isn't even getting to the possibility of ignoring socket bonuses.

EDIT: the above isn't totally correct. I'll try to make a good writeup.


Forgive me for my ignorance, but how does the accounting work for the meta requirement? Specifically, does it take two greens to equal one red (i.e. two half blues required to offset the whole red), or does a green count as a full blue gem and a full yellow gem, and thus a purple equal a full blue and full red that offset one another (which is the same as the current wrath model)? From reading this, it seems as though it's the latter indicating no change from the current regime, in which case I think your guide is, indeed, completely correct.

For clarity, I may combine step #3, #8, and #9 into a final catch-all step: "All remaining red fill with purple. All remaining blue slots fill with purple unless there are remaining yellow slots. For each empty yellow slot that has an empty blue slot available, socket one blue and one orange gem. For each pair of empty yellow slots, socket one orange and one green. For a single yellow, fill with green."

#22 Cdin

Cdin

    Von Kaiser

  • Members
  • 75 posts

Posted 15 November 2010 - 08:14 PM

Forgive me for my ignorance, but how does the accounting work for the meta requirement? Specifically, does it take two greens to equal one red (i.e. two half blues required to offset the whole red), or does a green count as a full blue gem and a full yellow gem, and thus a purple equal a full blue and full red that offset one another (which is the same as the current wrath model)? From reading this, it seems as though it's the latter indicating no change from the current regime, in which case I think your guide is, indeed, completely correct.

For clarity, I may combine step #3, #8, and #9 into a final catch-all step: "All remaining red fill with purple. All remaining blue slots fill with purple unless there are remaining yellow slots. For each empty yellow slot that has an empty blue slot available, socket one blue and one orange gem. For each pair of empty yellow slots, socket one orange and one green. For a single yellow, fill with green."


Hybrids count as two full colors. Socketing a purple gem is the same as socketing one red gem and one blue gem, not 50% of each.
www.GrayMatterWoW.blogspot.com - My Moonkin related Blog.

#23 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 15 November 2010 - 09:09 PM

I put a detailed writeup here: http://elitistjerks....equirement.html

If you skip to "conclusion," you can see the correct and simplified version of my flowchart above (basically, as concise as I can make it without sacrificing optimality). It should be true for all classes, so feel free to spread to anyone who's discussing this same topic.

#24 Draugir

Draugir

    Glass Joe

  • Members
  • 9 posts

Posted 16 November 2010 - 10:08 AM

This obviously doesn't work in a general gear set, but I gemmed a theoretical "best in slot" set (mostly using 372 ilvl tier + offset leather) both for the new meta requirement and according to current (WotlK) gemming. Using the dps values from Wrathcalcs Beta I found it to be a 150 dps increase to use a Chaotic Skyflare Meta. The 2 things that are making this occur are that, as Roywyn said, Int is so much better than any other stat, and that with an abundance of mastery on gear, it was easy for me to reach hit cap through reforging all mastery (which makes using hit gems basically a gain in mastery that you don't have to reforge).

#25 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 20 November 2010 - 04:12 AM

After playing with the new L85 WrathCalcs (posted on the other thread, try out the new item selector), I'm pretty confident that Chaotic will be the best meta.

#26 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 20 November 2010 - 10:42 PM

I'm still seeing Wild Mushroom as better DPET and Wrath and Starfire (by quite a lot, if you can really plant down 3 Mushrooms in 1.5 seconds). I haven't played Balance at 85 much on the beta--am I missing something here? Or is it really beneficial to plant down 3 Mushrooms every 10 seconds no matter what you're doing?

Edit: aha, WM planting still gives a GCD of 1, not 0.5.

Edit 2: Also, DoT's recompute crit rate in real time, so you can't abuse 4T11 by refreshing DoT's with 100% crit.

#27 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 21 November 2010 - 03:44 AM

Can an existing DoT consume the 4t11 proc? That would be ugly.

#28 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 21 November 2010 - 04:33 AM

No, just Moonfire direct hits.

Hmm, which I guess means you might want to cast 3 nukes before your NG/Eclipse Moonfire each half-cycle. Assuming you're in a stable 1-Moonfire rhythm, that's fine. If not though, maybe it's not worth worrying about.

#29 Cdin

Cdin

    Von Kaiser

  • Members
  • 75 posts

Posted 23 November 2010 - 09:45 PM

Edit 2: Also, DoT's recompute crit rate in real time, so you can't abuse 4T11 by refreshing DoT's with 100% crit.


Does this mean that DoT ticks are buffed by 4T11 for the 8 or less seconds it's up even if the dot was cast before the buff was procced?

I never got a chance to test the tier sets in Beta.
www.GrayMatterWoW.blogspot.com - My Moonkin related Blog.

#30 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 23 November 2010 - 09:48 PM

Does this mean that DoT ticks are buffed by 4T11 for the 8 or less seconds it's up even if the dot was cast before the buff was procced?

I never got a chance to test the tier sets in Beta.


All I tested was that if you cast the DoT when you have a 100% crit chance, you still get non-crit ticks after the buff expires. I'd be pretty confident that it works the other way too though, yeah.

#31 Zanarhi

Zanarhi

    Glass Joe

  • Members
  • 4 posts

Posted 23 November 2010 - 11:26 PM

Dots cast before gaining the buff will get the bonus to crit while the buff is up.
Periodic dot ticks do not consume any charges.
If starfall is used at the start of lunar the charges will be eaten very quickly and will not affect many dot ticks at all.

#32 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 29 November 2010 - 10:56 PM

Anyone else notice that, in the recent WrathCalcs sheets (including the most recent one where I think Tecton tried to put in what looked like BIS T11 gear), haste is showing better than hit?

That could certainly change some things--I'm going to have to think about it some. Among other things, I want to make sure I actually take hit into account correctly in all the calcs, since it was a nonissue for so long. But assuming that is the case: well, I suppose it doesn't change an incredible amount since the two best stats are still haste and hit, which is what matters most for gear selection and reforging. But it means that you wouldn't reforge haste to hit even if you were below cap (this virtually never comes up, now that I think about it), and you would reforge hit to haste even if it took you under cap.

#33 Tecton

Tecton

    Soda Popinski

  •  Patrons
  • 3,951 posts

Posted 30 November 2010 - 10:58 AM

Yeah, I've noticed haste pulling away from hit as you gear out of heroic blues and into T11 gear as I've played with different gearsets (I normally just try to set the gear back to something sensible before I save it for people to work with, there's been no effort to make it 100% BIS if anyone is looking at it in that sense, as an aside).

From what I can see, hit right is basically being treated as Damage * Total Hit % for each spell, yeah? If so, that's probably undervaluing hit because of NG/Starsurge/Eclipse energy lost? I know I'm undoubtedly missing something here, though!

#34 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 30 November 2010 - 07:13 PM

Yeah. I think I just forgot to include it in some of those things when I was working on the calcs during beta, since it never came up. I'll take a look soon.

#35 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 30 November 2010 - 10:55 PM

Even though I haven't updated the sheet yet, I wrote a post on hit being at least generally equal in value to haste:
http://elitistjerks...._important.html

It might be bit elementary for readers of this forum, but the basic point is worth emphasizing even here--between hit no longer being dominant, and reforging, the hit cap doesn't really matter anymore when picking items.

#36 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 02 December 2010 - 12:09 AM

Yeah, I've noticed haste pulling away from hit as you gear out of heroic blues and into T11 gear as I've played with different gearsets (I normally just try to set the gear back to something sensible before I save it for people to work with, there's been no effort to make it 100% BIS if anyone is looking at it in that sense, as an aside).

From what I can see, hit right is basically being treated as Damage * Total Hit % for each spell, yeah? If so, that's probably undervaluing hit because of NG/Starsurge/Eclipse energy lost? I know I'm undoubtedly missing something here, though!


I put in the only refinement to hit I can think of, which was to discount all energy gains based on miss rate. It actually bumped up the value of hit quite a bit--not enough to change the things I posted yesterday, but enough to make hit stronger than haste at least through T11.

Giving up 100 hit (1%), lengthens the whole cycle by around half a second.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users