Jump to content


Photo

[Balance] New 6.0 Rotation Discussion


  • Please log in to reply
30 replies to this topic

#1 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 03 June 2014 - 02:43 PM

First for reference, the patch notes where all the changes were announced:

http://us.battle.net...-4-3-2014#druid

 

I've started a thread for ongoing versions of WrathCalcs as I keep refining the model based on the new rotation:

http://forums.elitis...-wrathcalcs-60/

 

Finally, rather that restate all the basics so far, here are some posts with the analysis I've done so far:

http://iam.yellingon...onkin-rotation/

http://iam.yellingon...in-second-look/

 

Overall I'm happy with the rotation, but am still trying to keep thinking about things that have to be iterated. I'm sure more will become apparent as we have an alpha to play, but they've given us such nice and detailed notes that we can get pretty far already.

 

Some various points for potential discussion:

--As discussed in the second of those blog posts, a significant change is cutting a rotational spell by merging Moonfire and Sunfire into one button. The upside of this is clear; it will great freeing up a core button. And it generally makes sense since the goal will be to only use each DoT when it's getting a nice Eclipse buff. However, we should think through the ramifications of, for example, not being able to multidot with both dots, or not being able to apply both to a fresh target.

--Astral Communion has the potential to be very interesting, but rather complicated. I'll try to get around to posting some math I did on it, but it's summarized in the blog post. No rotational DPS benefit (good), but can be used to set up a high damage burst in a short fixed window when you need. I think the timing needed to judge when to start AC to get Eclipse where you want it at a certain moment (which is itself non-obvious) will take a lot of practice.

--Right now, AC cannot be used out of combat, and we can't set our Eclipse anywhere but the center for starting. My proposal would be to mirror the current behavior: Eclipse remains still (wherever it is) anytime you're out of combat, but you can use AC to move it even then. I welcome any thoughts on whether this would have any issues.

--As discussed in the blog post, Sunfire timing is weird and nonintuitive due to snapshot mechanics and Pandemic. Personally I'd like to keep the Eclipse DoT snapshotting, but so far I haven't come up with a way make Sunfire natural the way that Moonfire is.

 

There will be specific numbers/modeling issues too. I got into some of them in WC thread, and I think a lot of interested people might get some info just by looking at WC itself. Once we have an alpha to play (and in particular, reasonably final spell coefficients), there might be a lot more of the usual talk about optimizing numbers.

 

One specific point on planning out optimal play: with the current (likely placeholder) spell coefficients and revealed passives/perks, the DD component of Moonfire is too high. It actually beats out Wrath in DPET. I've of course ignored this (and not made rotations that just spam Moonfire as filler), but it does make it hard to answer certain questions of optimal DoT usage. When there's no actual downside to refreshing as often as you want, questions of how best to use DoTs during say movement, multi-target, or procs don't work well.



#2 Andcoup

Andcoup

    Glass Joe

  • Members
  • 1 posts

Posted 03 June 2014 - 07:20 PM

AC

-I really like the current system of AC so I would enjoy your suggested changes, however, I think it may be a weaker advantage than it is atm because there is no longer a starfall use gained from lunar.

 

Solar Flare

- This sounds good because currently it seems that almost all of the big stuff will be happening at the peaks of eclipses and just spamming wrath/starfire in-between but solar flare allows something to do in the middle.

 

Starsurge

-I think it being an instant cast will be very nice because it will be easier to time it with eclipse peaks without needing to think about the time it takes to cast.

 

Starfall

-I am so happy with the new system for starfall, whenever I am playing currently I always notice a lack of clutch AoE and I hope having the charge system and no target limit will allow for that.

 

DoTs

-The biggest problem I see with the new system at the moment is an almost complete lack of movement damage without spamming non-peak eclipse dots. Currently if it hurts to use dots due to snapshotting then you can use mushrooms instead, but mushrooms will no longer deal damage so that option is gone. This may not be an issue if the only benefit is procs and the damage is very low even at peak eclispe but I don't think this is the case with the current numbers.



#3 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 03 June 2014 - 10:14 PM

Worth noting that with Eclipse going to 40s, (and Sunfire to 20+s), only one of the three T100 talents will have a tough time maintaining Sunfire, at least for a Patchwerk fight.  Euphoria will have a 20s cycle to go with that 20s DoT.  Balance of Power will maintain that DoT for you almost forever.

 

Stellar Flare doesn't have either of those advantages, but it does have another always up DoT.



#4 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 04 June 2014 - 03:32 AM

That's a good point. I've mostly been thinking about the rotation with no L100 talents (because they're so preliminary right now). Also, Stellar Flare is the only one of them I really like so far, so maybe I don't like the idea of relying on L100 talents for this.

 

I touched on talent issues a bit in the post. One in particular I was thinking about is that the Eclipse-speed effect of Euphoria is, on its own, DPS-neutral or slightly negative. So they have to tie some unrelated bonus into the talent to make it worth taking. But at that point, why have the Eclipse shift in the talent. I think it's better placed as a Major Glyph, so you can choose Eclipse speed independently of other factors. I'm not even entirely excited by the idea of Eclipse speed changing, since I'm not sure when exactly you'd want it, but I can see it coming up at times. So make it a choice you can play with anytime to see if you can get some advantage from it, while using whatever DPS talents you want.



#5 Caaowl

Caaowl

    Glass Joe

  • New Members
  • 6 posts

Posted 04 June 2014 - 08:38 AM


DoTs

-The biggest problem I see with the new system at the moment is an almost complete lack of movement damage without spamming non-peak eclipse dots. Currently if it hurts to use dots due to snapshotting then you can use mushrooms instead, but mushrooms will no longer deal damage so that option is gone. This may not be an issue if the only benefit is procs and the damage is very low even at peak eclispe but I don't think this is the case with the current numbers.

 

Also, in PVP we would be the only class that is unable to refresh both dots. Even if we get back the separate buttons, we will be off of our single-target peak in PVP since one of the two dots won't be buffed (this is not an issue any other classes have to deal with anymore) -- but with only 1 button, the first 20 secounds of any encounter we not only lose half of our DOT damage, but also half of our shooting stars procs.

 

(Dear Hamlet, I promise to stop making this point in all your threads at this point, thanks! :))


Edited by Caaowl, 04 June 2014 - 08:39 AM.


#6 Qaajn

Qaajn

    Von Kaiser

  • Members
  • 43 posts

Posted 04 June 2014 - 06:45 PM

I understand their desire for Euphoria, and I am personally also very intrigued by it. The faster eclipse cycles would make you switch spells faster and give more of a "flow" in the rotation. Taking only 20 seconds for a full cycle, it would be only 10 seconds from one peak to the next (naturally), so there would be quite freqent shifts between spell use. This would be a skill in its own, more so then for the slower cycles, as you would have to go more on feeling with switching spells, if you could fit in another SF before wrath would be better, and so on. I don't know if that's how the starsurge charges works now, but if it doesn't, I would very much like that you got one charge for each cycle. I realise that this would give Euphoria a base SS use double that of the other talents, but I think this would fit well with the theme for the talent; you do everything faster, and be easy to balance to not be stronger then the other talents.



#7 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 04 June 2014 - 09:35 PM

I agree that Stellar Flare has the best "feel" to it for PvE.  Being able to do something useful when Lunar and Solar are both weak seems like a big plus, and should make the rotation more interesting.  I'd say the same for PvP, except that Flare is not instant cast.  Risking a both-schools lockout for a DoT cast sounds like a poor choice (unless we can kick butt as a Bear or Cat for six seconds or so).

 

I'd originally (incorrectly) read that with Euphoria, SS gave 20% haste for six seconds.  That would make that talent much more competitive, since your existing DoTs (which don't snapshot haste), would speed up for a little bit.  Euphoria also has the advantage that if you clip a DoT at low energy (because of movement, for instance), you don't have to wait nearly as long to repair the damage.



#8 Caaowl

Caaowl

    Glass Joe

  • New Members
  • 6 posts

Posted 04 June 2014 - 10:16 PM

(unless we can kick butt as a Bear or Cat for six seconds or so).

I have always hoped (not always secretly) that DoC becomes like so... I love being a hybrid,and it would be neat if it was a tad more than "you can heal better now chickin". I wish we could dish out a 70-75% DPS while "catting", and maybe become tad more durable while bearing  Something like a permanent, weaker HotW. 

 

Ah well, one can dream...



#9 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 05 June 2014 - 05:52 AM

Might be of interest:


#10 Qaajn

Qaajn

    Von Kaiser

  • Members
  • 43 posts

Posted 05 June 2014 - 10:17 AM

Might be of interest:

 

First impressions:
I found it very interesting to look at. I noticed that you mostly used a second SS during lunar, do you think that is worth more then using a single during solar?

Since there are two dot refreshes during solar, and none of them is exactely at the peak, I imagine it is easier to fit 3 wraths closer to the peak then 2 starfires. Using a starsurge right before lunar peak, I noticed that you never fit in the second starsurge before you were well down the lunar slope, atleast at 70% or lower. An unbuffed starsurge also hit only marginally harder or as hard as a regular starfire.

 

Checking in more detail:

Rechecking opening, 5.2k SS and 4k buffed SS during lunar peak with what seem like trinket procs for both. Due to this, I believe that you should consider shifting your starsurge use 1 spell earlier during the lunar cycle to fit two fully buffed SF along with the moonfire during the peak. Though you seem to hit a 4.5k SS -before- you reach the lunar peak, about halfways up. What's up with that, it gets boosted by both solar and lunar or just a huge damage range?

Going towards the solar peak, you get a double trinket SS crit for about 9.6k very close to the peak. That would make it "only" 4.8k with the second trinket proc before crit. I would have expected more at solar peak and double trinkets. With the second trinket up then, your wraths hit for 3.2k each, which would place them at about the same DPS as starfire, assuming equal mastery.

About to enter the second lunar, you launch a wrath that hit for 1.9k very close to the switch. Coming out of the second lunar, you land a SF for 3.2k right before the switch. Both were no SS buff or trinkets. It seem to place them again close in DPS. The SF seemed to be a bit during lunar still.

Looking at moonfire ticks placed right at the peak, and sunfire right as a fresh solar is entered, the eclipse damage difference between lowest and highest seem to be 25%,  with 726 damage moonfire and 581 damage sunfire. It also appears as if moonfire/sunfire snapshot trinkets, as their periodic damage doesn't alter with those procing or fading during the spells duration.

At the second solar peak, you land 3 unbuffed Wraths, hitting from 2.1k-2.6k, which seem to indicate about the maximum damage range of wraths. With the damage difference of minimum and maximum of solar of 25%, this would place midpoint wraths at 1.7k-2.0k, which seem close to what we have observed.

Going up to the third lunar, an unbuffed SF hit for about 3.8k close to the peak, a buffed one hit for 5k what seem to be right after the peak.

At the third solar peak, with 1 trinket up, the SS hit for 4.6k, followed by 2 wraths hitting during the peak, for 3k and 3.2k respectively, with the third wrath right after it for 3.6k. All affected by that trinket.

Going out of the third solar, a wrath hit right before it is left for about 1.9k.

At the forth lunar, 2 buffed SF seem to hit at or very close to the peak, both hitting for 5.1k. At the same time, we have two SS during the same cycle, both hitting outside the peak with no trinkets, with one for 3.8k and the other for 4.5k, which still seem to indicate quite the damage range.

 

Considerations:

I could check the numbers further indepth, but at this point I don't think that would give us much. Some conclusions we can draw from this is that it seems that the damage range of SS is quite big, and that our dots as of that patch snapshot trinkets. Combining the fact that SS doesn't seem to hit considerably harder then a SF with that SF and wrath seem close in DPET along with the desire for a moonfire at the lunar peak, as of now I would suggest that instead of the base SS at lunar peak, try it at solar. You can more consistenly fit all the buffed spells at maximum benefit and not disrupt the desired rotation. To further strengthen this conclusion, SS in this itteration had a casttime, which would keep it at or close to the peak while allowing the wraths to reach maximum damage.

Another thing to note is that spells seemed to check their eclipse value on cast completion and not spell hit. It is somewhat hard to judge as we only have traveltime on SS and wrath, but it seemed to indicate so from wrath casts, and is also what we would expect.

Personally, I quite like the flow of the rotation with a casttime SS. It gives a nice feel to the rotation, although when you press keys doesn't change and when damage happen relative to other things only shift back and forth 1 GCD, I still liked the look of it, traveltime spells look better to me shot from standing then jumping.




 


Edited by Qaajn, 05 June 2014 - 10:18 AM.


#11 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 05 June 2014 - 12:42 PM

Incidentally, I think my peak Eclipse bonus in that video is 69.54%. 301 mastery rating at L90 is 13.087 points of mastery. And mastery still seems to be giving 1.875% per point (patch notes suggest 1.5%). So (13.087+8)*1.875 + the base 30%.



#12 Qaajn

Qaajn

    Von Kaiser

  • Members
  • 43 posts

Posted 05 June 2014 - 02:48 PM

Incidentally, I think my peak Eclipse bonus in that video is 69.54%. 301 mastery rating at L90 is 13.087 points of mastery. And mastery still seems to be giving 1.875% per point (patch notes suggest 1.5%). So (13.087+8)*1.875 + the base 30%.


That seem to fit quite well with my napkin math then. You've lost half the peak mastery at the midpoint, which would put it at about 35%. That would give us 1.7/1.35=1.26 damage increase going from the midpoint to the peak, close to the 25% I calculated based on your dots.

Edited by Qaajn, 05 June 2014 - 02:50 PM.


#13 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 05 June 2014 - 03:45 PM

I was trying to use SS whenever I could during peaks, it definitely wasn't perfect. Fitting 2 into one side of a cycle is hard with how short they are; 40 seconds will be a lot nicer. I found that it was easy to let SS cap. It's possible that a cap higher than 3 will be good, especially as we get procs more often with higher stats (don't remember exactly what my haste/crit were).

 

One issue is that you really don't want to SS on the way down. That was never ideal because you're heading into a valley, but sometimes you want to if you're about to cap or suddenly capped from procs. The way Empowerment works now, if you SS on the way down, you can't use the Empowerment before it's time to switch spells, so it's a waste. That could possibly be fixed so that the Empowerments aren't two separate buffs, but might be tricky because they have different numbers of charges. Maybe more important to ensure the SS proc rate is never going to get too high that you can't make good decisions about when to use it.

 

I'd originally (incorrectly) read that with Euphoria, SS gave 20% haste for six seconds.  That would make that talent much more competitive, since your existing DoTs (which don't snapshot haste), would speed up for a little bit.  Euphoria also has the advantage that if you clip a DoT at low energy (because of movement, for instance), you don't have to wait nearly as long to repair the damage.

 

I think it was haste in the initial version. That's a good point on the faster Eclipse. It's maybe another reason it shouldn't depend on a DPS talent, but be a choice you can make independently. I don't like how hard multidotting/target-switching is looking right now, but that would be an interesting option to help with it.

 

I haven't actually checked since he tweeted the revised cycle times whether hastening Eclipse is still a loss. It might be a minor positive due to being able to maintain a 100% Sunfire (or still a negative due to narrower peaks making it harder to Starsurge efficiently). Probably a minor issue either way; I don't see how it would be a significant increase.


Edited by Hamlet, 05 June 2014 - 05:39 PM.


#14 Tecton

Tecton

    Soda Popinski

  •  Patrons
  • 3,951 posts

Posted 06 June 2014 - 02:43 PM

My main concerns about the rotation are (and I'm sure some of these will be remedied when I get a chance to try it out myself):

How static the single target rotation is potentially going to feel. The nature of it just asks us to math out the best point to DoT, the best point to Starsurge, etc. With the reduced proc rate of Shooting Stars, the ability to store 3 charges of Starsurge and the removal of Starfall as a rotational spell I worry that the reactive nature of the spec will be gone. Not really sure how to resolve that without us ending up back in SS machine-gun land, though. Right now the results you'd get from a spell predictor addon would not be great, it definitely seems that this kind of thing would turn out much higher DPS after 6.0 changes come in. That may be the desired direction from some of the comments, though, so I guess that's understandable.

The AC problems have been highlighted above, but I think it's worth reiterating that not being able to pre-set our Eclipse position is going to get very annoying, very quickly and there doesn't really seem to be any great reason for us not to be able to do this. The other issue is that using AC to line up for combat events will probably devolve to having a BPT type mod tie into your boss mod and say "AC for X seconds now". Not going to be a problem for high-end druids, but I'd imagine the majority will struggle.

The DoT changes present an issue in that we're going to mangle our DPS if we overwrite them while moving, and we need to use Starsurge to keep our normal rotation competitive. Obviously, encounter design is going to play a big part in this, but it does make me a little nervous for things like Proving Grounds/Challenge Modes and the inevitable heavy movement encounters again.

Overall, I think the changes are very positive, and I'm sure things will evolve and change but I'm a little worried about ending up as a turret-style class who's DPS falls into a hole as soon as you need to dodge mechanics/etc. Granted, I've not played it yet and I'll be able to give more useful feedback once I see how it works in reality and raid testing.

#15 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 06 June 2014 - 04:37 PM

It is a little static, I felt that way playing it for those few minutes. Funny background--I felt the exact same way when the current Eclipse was first revealed in Cata beta (and it was a fixed number of casts between each Eclipse, basically like now). I gave that feedback that we needed a proc of some kind of else you'd /castsequence the whole thing, and they added Shooting Stars :P .

 

All in all I'm happy to trade the current Shooting Stars, where you just react instantly when it procs, for this one where you plan out when to use an added charge. Hopefully the feeling of constantly having to think about when you're going to use your next Starsurge (and what to do when you get a proc) is enough to make it not completely static. You do have to watch SS charges pretty closely it seems, if at any point you might be trying to save 1 for the next Eclipse, potentially getting 1 from a proc, and getting 1 sometime soon from the ordinary cooldown.

 

One thing to remember about AC is how fast it is. It currently takes 16 seconds to go through a whole loop with AC, but in the new system it's 7.5 (or 10 once they change it to 40s). So if you don't mind a second or two of inefficiency (which is usually fine in the period before the burst phase starts), it might not be so bad. 5-7 seconds before the start of the window you can try to AC until wherever your target point is and generally get it about right.

 

Am still concerned about movement and rampup, I should try actively brainstorming ideas more on those.

 

e: Putting the last two issues together, it does matter that you can AC while moving now. That's a pretty complicated decision to make on the fly though (whether you should or shouldn't, every time you move). Definitely interesting and should help, but isn't a complete movement solution.


Edited by Hamlet, 06 June 2014 - 04:42 PM.


#16 Caaowl

Caaowl

    Glass Joe

  • New Members
  • 6 posts

Posted 06 June 2014 - 07:45 PM

e: Putting the last two issues together, it does matter that you can AC while moving now. That's a pretty complicated decision to make on the fly though (whether you should or shouldn't, every time you move). Definitely interesting and should help, but isn't a complete movement solution.

I can totally see that to be the new boomkin pvp "rotation". Dot everyone up, run away while channeling AC, rinse and repeat.

Sounds like fun! (?)



#17 Tecton

Tecton

    Soda Popinski

  •  Patrons
  • 3,951 posts

Posted 06 June 2014 - 07:55 PM

I'll admit I'd forgotten AC can now be used on the move, that certainly helps a bit in some situations. It'll be interesting to see how things develop, I think we've got several more changes coming in the next build?



#18 Erdluf

Erdluf

    Great Tiger

  • Members
  • 968 posts

Posted 06 June 2014 - 08:05 PM

https://twitter.com/...437667963150337, Stellar Flare mastery bonus will be the product of both current Lunar and Solar bonuses.

 

So If I have 60% Mastery, my MF could get a multiplier anywhere from 1.3 to 1.6 (can't cast MF on Solar half), a 23% relative spread.

 

At worst, Flare would have a bonus of (1+.6)*(1+0) = 1.6, and at best (normally) 1.3*1.3 = 1.69.  Essentially that curve is mostly flat and low.  It has a peak that is somewhat pointy, but not very tall.  The main reason to cast Flare at the perfect time is not to give it a big bonus (<6% more than its weakest bonus), but to avoid casting anything else at a time when that "something else" has a low bonus.

 

However, during CA, Flare's bonus would be 1.6*1.6 or 2.56.

 

At first that 2.56 seems like a lot, but perhaps not.  During CA, a MF cast applies both MF and SuF DoTs, effectively doubling the DoT damage it would be doing (at least for a while).  CA's damage increase (in my example) is 2.56/1.69, or about a 50% increase.   MF probably got an even larger benefit.

 

Bottom line,the quadratic scaling is a concern, but if at 60% Mastery, non-CA MF does about the same damage as non-CA Flare, then Flare's CA scaling isn't going to be "crazy high" until we see Mastery numbers substantially over 100% (third tier of the expansion might be the earliest that happens).



#19 Hamlet

Hamlet

    Mike Tyson

  • • Guide Author
  • 11,567 posts

Posted 07 June 2014 - 01:00 AM

That was basically my analysis of StF. That's also how it works in the sheet already. It's still quite a big benefit to cast at the midpoint--even though the value of StF is flat, the opportunity cost of the 1.5s spent varies quite a bit between midpoint and peak.

 

The CA interaction is good. It may well be good to MF twice during CA, even though that adds some clipping.



#20 Caaowl

Caaowl

    Glass Joe

  • New Members
  • 6 posts

Posted 07 June 2014 - 04:59 AM

That was basically my analysis of StF. That's also how it works in the sheet already. It's still quite a big benefit to cast at the midpoint--even though the value of StF is flat, the opportunity cost of the 1.5s spent varies quite a bit between midpoint and peak.

 

The CA interaction is good. It may well be good to MF twice during CA, even though that adds some clipping.

 

Do you think it could make sense to StF twice during CA? 

Alternatively: from what Mastery does it make sense to do that?

(probably this is a question best asked once the numbers are set)

 

I have no idea how they think they will balance StF and Balance of Power against Euphoria. I agree it would probably work best as a glyph instead (and maybe they could bake in faster cast/lesser damage to the main nukes, to get us back to the feel we had a long time ago in a galaxy far far away)






2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users