 |
02/21/13, 11:29 AM
|
#361
|
|
<Druid Trainer>
|
RealPPM is just something we'll have to get used to if they want to go this way rather than making trinkets effectively into short cooldowns. You're no longer more likely to get one during your opening timer stack than at any other time, in particular. I'm not sure how you conclude that you're only in Eclipse 1/5 of the time (did you mean 1/2?). You should be most of the time. Also getting one outside Eclipse isn't horrible anyway. The DoTs aren't Eclipsed, but the Starsurges push you into it so you often get a similar benefit.
I've been saying the whole time that a proc doesn't actually result in "chain Starsurges" (especially to people who keep speculating about horrible PvP effects). You get a few extra Starsurges over the course of the DoTs. That's still a pretty nice DPS burst.
A bigger point I want to make about the trinket: the "positive feedback loop" is not as dramatic as people make it out to be. In reality we really only expect the super-DoTs to be slightly longer than normal DoTs (say 20 seconds over a 14 second base, as opposed to a normal average of ~16). Yes, you're casting a lot of Starsurges, but those Starsurges have a normal crit rate. You always chaincast a bunch of nukes during your DoTs and have a chance of extending them. The only added benefit here is that you might get ~1 extra cast in since Starsurge is faster than Starfire, and that when you do get a crit it extends both DoTs more often.
Point being: the benefit of this trinket is a bunch of extra DoT crits and a few extra Starsurges (4-5 isn't bad at all) over the course of ~20s every time it procs. That's a lot of damage. It's not based on the idea of some insane feedback loop of extending the DoTs with nonstop Starsurges.
--------------
The big issue I alluded to above is that you need to have a pretty dramatic UI alert for it. You absolutely need to recognize the proc with 2 GCD's to spare to cast either Moonfire/Sunfire or (if you're lucky and they're up), Incarnation, CA, Moonfire. This may even necessitate a /stopcasting if you started a Starfire or something half a second after the proc (can't be avoided due to reaction time).
They're clearly trying to make it so that trinkets will benefit players who pay attention to the procs (see Re-Origination also). This is a slightly a extreme example though, since you need to be aware of it within about 1 second every time. An ordinary weakaura along with your other buffs might not be enough, you need something huge like you'd use for an important boss ability. It should be fine as long as you do that though.
Last edited by Hamlet : 02/21/13 at 11:39 AM.
|
|
|
|
02/21/13, 3:58 PM
|
#362
|
|
Piston Honda
|
I know you guys are probably over hearing my failed attempts at trying to model 2/3 target fights in WrathCalcs, but I gave it another shot today so hopefully I'm a little more accurate than dividing tick rates... I apologise in advance if it's wrong and I've wasted your times (again).
I first tried brainstorming the changes so I could calculate the damage gained off to the side.
Gains:
2 dots (1 uneclipsed, 1 eclipsed)
additional SS procs
additional dot extentions (one target, due to additional SS procs)
additional chance for trinket proc (I decided to not include this, hard to follow Hamlet's modelling of the Unerring trinket)
Losses:
2 GCD's
Both gains and losses are doubled when considering a full eclipse cycle, and because of the speed at which we cycle eclipses there should be more than enough time to dot two targets and still have faux-100% NG uptime. Dendritus and I spoke about this yesterday, in that NG only has to be active for 30s + SFcasttime + Wrathcasttime for the full benefit to be had. Thus, both a haste and crit reforging achieved this (longest cycle length of 32.9s). So I figured these should be the only changes had when including another target. I used the rotation tab for all calcs.
So for additional DoT damage I just added the two dots from the Lunar eclipse (D35 & D36) for the eclipsed and uneclipsed dots. Additional SS procs should be the exact same as normal SS procs as it's just a duplication of current DoT ticks, so I copied D23 and multiplied by SSAverEclipse to find additional SS proc damage. DoT extentions only applied to one set of DoTs, and they should be extended by (add'tnl SS procs)*TotalCrit*2 and then the damage from these ticks should've been (ticks)/MFTickRateNG*MFTickDamage, I hope. I think it would be more accurate to use the Solar damage numbers for the DoTs and add that to the Lunar damage instead of just doubling, but they looked close enough, I'm sure I'd have larger errors somewhere.
Anyway, that should have approximated the gains from a half-cycle, so doubling that came to just over 3 mill additional damage over a full eclipse cycle. Adding that to the Final Damage line (H37) and including "+4*NGGCD" to H30 (Final Time) made Subtotal DPS increase by a fair amount (up to 252k or so).
I hadn't factored in CD usage, so I tried to emulate that with double SS procs in I32 and double DoT damage in I35 and I36. I didn't mess around with the Incarnation damage, as I'm not sure if we should DoT the second target during Incarnation and before CA is popped.
Anyway after all that, for two targets I got 263107.81 DPS with a crit build and 266043.01 with a haste BP build, which both use the Unerring trinket. If we were to swap out the Unerring trinket and replace with Cha-Ye's, a crit build yields 246888.49 DPS and a haste BP build yields 246638.29 DPS.
If my alterations are correct (I'm hoping), it shows that the Unerring trinket favours the haste breakpoint by quite a lot, and the additional SS procs do fall behind when we consider the 100% crit rate on our DoTs (which would devalue crit during that period). However, when the Unerring trinket is replaced, a crit build is only slightly, but definitely is, ahead of the haste build.
So again, I apologise if I wasted your time with incorrect maths/logic, and I recognise there's a couple of flaws with what I've modified, but I feel like this is a pretty good approximation of a 2-target fight, and shows that the haste breakpoint isn't worthless or not worth reaching in a T15H BiS scenario. Maybe this would help if Tec/Hamlet were going through to make WrathCalcs support 2/3 target fights.
Oh, and the reason I didn't mess with 3-target was because NG would definitely fall off, and then it's getting into more complicated stuff so I figured 3 targets would just extrapolate itself from a 2 target fight.
Anyway, criticise me as much as possible, I'm not a programmer at all, just trying to figure this stuff out before 5.2 hits.
|
|
|
|
|
02/21/13, 4:20 PM
|
#363
|
|
<Druid Trainer>
|
Don't have time right now to think through whether you've included all of the +'s and -'s to total damage that you'd need to account for to get a reasonable estimate. I do think it will be eventually be more accurate and also easier to build the DoTs into the actual rotation, rather than trying to think of everything you'd need to add and subtract manually. Looking down the rotation real quick, what I'm imagining I'd do is:
1) In line 11, add extra casts to account for added DoTs. Lines 12-13 should flow correctly as they are.
2) In line 19, add extra DoT ticks to correspond to unextended DoTs being cast (ignoring all the machinery in lines 14-18 about nuke crits since they only affect the main target). To add the new DoT's perfectly, will have to add a new line that splits up the new ticks to each phase of Eclipse cycle, much like the above lines.
3) In the rotation time block, only change should be to line 29, to account for added casts. Everything else should flow from the updates to line 19.
4) Add the actual damage for the new DoTs (remember, unextended) to lines 35 and 36. If I want to be fancy, account for Lunar Shower.
5) Update Celestial Alignment by changing lines 19, 29, and 35/36 similarly.
|
|
|
|
02/21/13, 4:31 PM
|
#364
|
|
Glass Joe
|
Originally Posted by Hamlet
I'm not sure how you conclude that you're only in Eclipse 1/5 of the time (did you mean 1/2?). You should be most of the time.
|
I'm not concluding anything, I'm just saying, for 10 procs, I got only 2 during an eclipse. But that's not my point here, maybe some others got 100% procs during an eclipse. You said that I should be most of the time in eclipse time. In fact, nobody atm can manipulate the proc. Thus, we can't choose the right moment.
Originally Posted by Hamlet
Also getting one outside Eclipse isn't horrible anyway. The DoTs aren't Eclipsed, but the Starsurges push you into it so you often get a similar benefit.
|
It can be pretty long going on Lunar Eclipse without luck (meaning procs SS). Also if your DoT's arent' Eclipsed, like I said, it's a 30% procrate. It's not a fully spamming one button. I even had the same impression with or without the buff at some times.
Originally Posted by Hamlet
I've been saying the whole time that a proc doesn't actually result in "chain Starsurges" (especially to people who keep speculating about horrible PvP effects). You get a few extra Starsurges over the course of the DoTs. That's still a pretty nice DPS burst.
|
I'm agree without, excepted that I doubt about this "nice DPS burst" every 3 or 4 minutes versus an intellect proc with a stable procrate. I can easily imagine the firemage delaying his combustion during an eternity for nothing
|
|
|
|
|
02/21/13, 4:34 PM
|
#365
|
|
<Druid Trainer>
|
The proc frequencies/variance/whatever on UVLS and the otherwise best trinket, Breath of the Hydra, are identical. Both are 0.5 RPPM.
|
|
|
|
02/21/13, 4:36 PM
|
#366
|
|
Piston Honda
|

Originally Posted by Hamlet
Don't have time right now to think through whether you've included all of the +'s and -'s to total damage that you'd need to account for to get a reasonable estimate. I do think it will be eventually be more accurate and also easier to build the DoTs into the actual rotation, rather than trying to think of everything you'd need to add and subtract manually. Looking down the rotation real quick, what I'm imagining I'd do is:
1) In line 11, add extra casts to account for added DoTs. Lines 12-13 should flow correctly as they are.
2) In line 19, add extra DoT ticks to correspond to unextended DoTs being cast (ignoring all the machinery in lines 14-18 about nuke crits since they only affect the main target). To add the new DoT's perfectly, will have to add a new line that splits up the new ticks to each phase of Eclipse cycle, much like the above lines.
3) In the rotation time block, only change should be to line 29, to account for added casts. Everything else should flow from the updates to line 19.
4) Add the actual damage for the new DoTs (remember, unextended) to lines 35 and 36. If I want to be fancy, account for Lunar Shower.
5) Update Celestial Alignment by changing lines 19, 29, and 35/36 similarly.
|
Yeah, don't bother looking through my numbers, your answer's much more eloquent and straight-forward. I'll leave this stuff to you guys haha, as I said I'm not very good with this stuff - I'll stick to napkin mathing. :p
|
|
|
|
|
02/21/13, 4:47 PM
|
#367
|
|
Moonkin Hatchling
|
A sample size of 10 procs is absolutely too small to draw any real conclusions about anything. Furthermore, it's a 51% proc rate on Shooting Stars within a DoT tick period, since each DoT has a 30% chance on crit to proc and both DoTs then have 100% crit chance. There is a chance that fight mechanics will be more punishing for this trinket than others (e.g. a wasted proc due to a period of boss invulnerability, etc.) because it is only a 4 second window wherein we can utilize it, but you should be keeping a variety of trinkets available for use anyway should you encounter mechanics like that.
Last edited by Lazerdollarz : 02/21/13 at 5:17 PM.
|
|
|
|
02/21/13, 5:09 PM
|
#368
|
|
<Druid Trainer>
|
Originally Posted by Slippykins
Yeah, don't bother looking through my numbers, your answer's much more eloquent and straight-forward. I'll leave this stuff to you guys haha, as I said I'm not very good with this stuff - I'll stick to napkin mathing. :p
|
See here: [Balance] WrathCalcs
Think it's not going to be quite so simple for n=2, but worth doing. N=3 is basically just a multidot spam rotation.
|
|
|
|
02/21/13, 7:25 PM
|
#369
|
|
<Druid Trainer>
|
As an aside, with all the talk about RPPM trinkets, I thought of something else--I'm more prepared to say HotW is likely to be better than NV in 5.2. HotW was slightly ahead on paper in 5.2, but NV's interaction with the Incarn/CA combo and (at at least ones in the fight) all your trinkets and procs put it noticeably ahead in practice.
But now--half of NV doesn't interact in an above-average way with anything at all (not even Eclipse), since you have to use it in lockstep halfway between Incarnations. And the other half of NV no longer interacts with trinkets (or Jade Spirit, since 5.1). The intangible benefits of NV now just amount to a boost from Bloodlust and a DPS potion once during the fight.
|
|
|
|
02/21/13, 8:27 PM
|
#370
|
|
Great Tiger
Night Elf Druid
Echo Isles
|
Of course NV has about 1/3 uptime (and that uptime damage is above-average), and provides 25% of your damage as healing, so you end up doing nearly 9% of your total damage as healing (with no dps-cast-time penalty). If the HotW->NV DPS penalty is only around 2%, that could be a worthwhile tradeoff if you have a fight where healers are challenged.
|
|
|
|
|
02/21/13, 8:42 PM
|
#371
|
|
<Druid Trainer>
|
Yeah, I forgot to mention the big bump to the healing utility compared to 5.1 (I've mentioned it other times). Only issue is that's it's hard to evaluate for a reason I point out a lot in Resto discussions--the value of healing is heavily discounted when it comes at times that are either random or fixed (i.e. not used intelligently for healing purposes). If it comes at an easy time, even if it's effective, it just amounts to saving them a little mana. There will definitely be times when the hard healing and the hard DPS part of a fight are at the same time though (maybe people are used to watching for this already after using NV all through the first tier).
|
|
|
|
02/21/13, 11:06 PM
|
#372
|
|
Von Kaiser
Troll Druid
Stormrage (EU)
|
I'm also wondering about the issue of which DoT should be placed first.
I always cast uneclipsed DoT first and the eclipsed DoT second because of Lunar Shower and that's what this guide suggests.
But some top Moonkins (Zoomkins and Zamey) have said that they always cast the DoTs in the same order. If I understand their reasoning correctly, that is because of how the DoTs line up together. If you cast them in the opposite order every eclipse, you end up clipping some of them I suppose?
Another thing I am wondering about is the cast priority in multi-target situations.
Is it: Eclipsed DoT > Shooting Stars > Uneclipsed DoT
Or is it: Shooting Stars > Eclipsed DoT > Uneclipsed DoT
I have seen both answers suggested by posters here and elsewhere.
I am assuming that whichever order is correct, you should make sure that the Eclipsed DoT is up on all mobs before casting the instant Starsurge, which will make you leave Eclipse.
|
|
|
|
|
02/22/13, 1:41 AM
|
#373
|
|
<Druid Trainer>
|
If there's any issue of clipping whatsoever, you cast the Eclipsed DoT first because it's the one that can get extended. Offhand I can't think of any reason for alternating like that since again, the two DoTs are different durations due to extensions.
|
|
|
|
02/22/13, 2:16 PM
|
#374
|
|
Glass Joe
Tauren Druid
Doomhammer (EU)
|
The proc frequency changes based on the ilvl of the trinket:
541 ilevel = 112.88% proc multiplier
535 ilevel = 106.74% proc multiplier
528 ilevel = 100.00% proc multiplier
522 ilevel = 94.56% proc multiplier
502 ilevel = 78.49% proc multiplier
463 ilevel = 54.57% proc multiplier
This counts for both Unerring Vision and Rune of Re-Origination.
Source:
PTR Class and Set Bonus Issues, Part III - MMO-Champion BlueTracker
|
|
|
|
|
|