The Moonkin Repository

I'm aware that most people are concentrating on Cataclysm but I was playing with this recently and thought I would share.

I think it was some comment from Hamlet that gave me the idea, although I don't remember which. The basic concept is that maintaining high Eclipse uptime is essential - and therefore, counterintuitively, you should always reapply DoTs only inside of Eclipse. Any time you do that outside of Eclipse (or near the end of Eclipse, in the case of Wrath and travel time) you are delaying the next proc.

Now the obvious counterargument is that while all of the above is true, you lose more dps by losing the benefit of Eclipse then you get from the higher uptime. So I turned to Simcraft to try and get a decent objective look. I used my current gear: http://www.wowhead.com/profile=20755359 and then modified the criteria it uses for casting Insect Swarm and Moonfire.

Apply each Dot at the Start of each Eclipse proc (Moonfire in Lunar, IS in Solar): 12351
Reapply whenever they drop: 12130
Reapply outside: 12206
Reapply outside for next eclipse (IE, only refresh IS if you're trying to proc Solar): 12133
Reapply at the end for the next eclipse (IS at the end of Lunar): 12163

Overall I found the results pretty surprising, but I've been trying them in raids and am seeing positive results (by which I mean nothing statistically relevant, but my dps has looked good compared to the rest of the raid and on WoL). So at least for now, no DoTs outside of Eclipse for me - if I can stand and nuke.

Yeah, this is something I'd been wondering about. I'm not going to have time to tinker with the spreadsheet for a few days, but it will be something to add in.

Funny thing is, I remember having this talk extensively a very long time ago (people probably remember, way back in 3.1, just like now, the big Moonkin TC issue was trying to figure out when to refresh DoT's). I remember comparing the lost time from asking in Eclipse to the increased cycle length from casting out of Eclipse and finding that the latter was better--that's why everything was based on "refresh when Eclipse ends" for so long. But a lot's changed since then. Also, for a long time, the 2T9 bonus and Glyph of IS made DoT's so strong that you generally refreshed them regardless of Eclipse, and this issue was mostly forgotten.

Will be interesting to look into more.

I think the one exception which is hard to simulate or calculate is the IS which helps aid the guesswork which goes into Wrath spam when you *think* you've procced Solar.

After a LunarE - I will IS immediately following the first non eclipsed starfire (the first eligible to proc solar)- this buys me a GCD to react, and potentially gain an extra Wrath in my Solar rotation.

My Moonfires though will go up only after Lunar procs - as the flighttime of Wrath doesn't allow for you to 'use' the GCD the same way, and you're better off gambling with NGrace procs.

Cheers for the insight anyway, always working to push my DPS further - I don't even know what our Cataclysm talents are yet, I'm sheltered like that

Nice to see that a simpler rotation won out. Trying to wrap one's head around when in an Eclipse a DoT should be refreshed is complex. Knowing whether you are close enough to the end of an Eclipse gets sticky; casting Starfire at 2.1 seconds left in Lunar makes more sense than at 1.7 seconds, but is it really possible to tell the difference between the two? Enter the necessity for the above mentioned IS trick. Anyway, it appears you have essentially tested both of these considerations with the "Reapply at the end for the next eclipse" case. I've fine-tuned that case in Team Robot before (example, reapply moonfire at 4 seconds left in Solar Eclipse), but the gain was never more than 100 dps compared to other methods (none of which included your winning case).

I've honestly spent a lot of time unlearning the "guess" rotation tricks - I make a conscious effort to not switch early, or to cast a dot when I "think" Eclipse is about to proc.

The problem is that while those do work a good portion of the time (and it's really, really awesome when they do) it really kills your dps when they don't. And unfortunately there's a trend to remember the really awesome moments and glaze over the average - I'm guilty of it, I know that.

Also the "IS trick" gives 12229 dps as far as I can tell (casting Moonfire at the start of each Lunar still. That could be wrong (it's hard sometimes to tell if simcraft is actually doing what you want it do) but I'm pretty sure this logic is solid:
actions+=/moonfire,if=!ticking&buff.lunar_eclipse.remains>0
actions+=/insect_swarm,if=!ticking&!eclipse&trigger_solar

cast IS if it's not ticking, we're not in Eclipse and we want to trigger a Solar. Since the logic should still make it cast a Starfire before that, this should fit the criteria.

And the "winning" logic:
actions+=/moonfire,if=!ticking&buff.lunar_eclipse.remains>0
actions+=/insect_swarm,if=!ticking&buff.solar_eclipse.remains>2

The >2 is just so it doesn't apply IS right at the end sometimes. Note that in both cases we aren't applying Dots at the start - you should still do it but since it's equal for all rotations I didn't bother figuring out how to make Simcraft do it.

Another oddity is I'm getting really high scale values for crit using this rotation. Simcraft isn't super exact on those so I'm going to investigate it further before posting more.

Edit: Checked the scalings for multiple runs: (10,000 iterations, 500 second fights, full raid buffs minus FM).
SP : crit : haste
1 : 0.6817 : 0.4478
1 : 0.6804 : 0.4490
1 : 0.6823 : 0.4521

By comparison, using a "standard" rotation we get something like this:
1 : 0.6313 : 0.7038

I'll need to play around with gear sets and make sure I'm not seeing an artifact due to an exact haste value, but I doubt it. I'd guess that because the rotation focuses on high Eclipse time crit gets a slightly higher value. I'm not sure why it kills haste scaling though, that seems odd to me.

The difference is overall pretty small but is enough that I'm debating going to Potent gems for yellow socket bonuses.

Note that if you are casting DoTs at the beginning of Eclipse, you are only casting them when NG is up (NG is up at the beginning of Eclipse).

Casting either DoT when NG is down tends to really extend NG downtime. IS can't give you NG, and if MF gives you NG, the first 1.0 to 1.2s of that NG are wasted.

Also, if you aren't careful, you aren't casting DoTs until your first Eclipse, meaning your Idol is getting off to a slow start, meaning low crit could really impact the start of the fight.

I've seen good results in AskMrRobot, requiring NG before casting a DoT. Note that at the start of the fight you'll probaby have Starfall, and that will give you NG for for your first DoT(s).

Simcraft gives me 95% uptime and 97% effectiveness on Nature's Grace using that method. More standard ones give 92%/93% uptime and then 96% effectiveness, so while minor that does seem to be a good point.

If I were to cast starfire until I saw solar proc, I'd cast one starfire too many every time. Is that how you were simulating it? I queue an Insect Swarm after the last Starfire of a lunar Eclipse since the proc is likely and it gives me a second to see if solar Eclipse actually procs (I guess some raid buffs are missing some times?)

I'm guessing you mostly mean for the solar -> lunar transition, since that's not predictable. I've always at least one Wrath too many due to the latency of seeing the Eclipse proc, but at least Wrath is fast. (I confess I'll throw an early Moonfire if lunar Eclipse is being stubborn, but that's more for the psychology of it than any real dps benefit.)

I hadn't considered Smo might be talking about the start of the fight... I like starting with both dots: You start low on threat, you can cast them as you run into position, and once you start nuking all your procs are probably active (trinkets, enchants, idol*5.) A bit less damage on the tank while the healers are positioning, even.

No, I mean for both transitions.

The problem with the "guess" rotation of casting IS after the first Starfire is what happens the (roughly) 40% of the time that starfire doesn't crit. What simcraft is suggesting is that you lose more dps from that then you will gain back the rest of the time. In other words (made up numbers ahead) lets say you gain 150 dps when it works but lose 250 dps when it doesn't... the net result is that doing it loses you damage.

So yes, it looks like you should always cast that extra spell and then apply the dot.

Interesting, because that's another thing that I checked way back when I first came up with the Eclipse-guessing stuff. Can't find the EJ post where I broke down the possibilities for swapping to SF on a W NG proc, but it was a long time ago.

Notably, it's never been modeled in WC though.

I'd love to see what Wrathcalcs says if it's modeled - I trust Simcraft but independent observations are always useful.

My guess is that it's related to losing NG on the cases where Starfire fails to crit, which leaves you in a really bad situation. Straight up Starfire spam avoids that to a large extent.

There won't be any work in WC for another week or so. But then I will try adding in fixed DoT refreshing inside of Eclipse.

Eclipse-guessing is trickier. I can try to come up something, but it probably won't include the subtleties of NG.

I was under the impression that, if the last cast of Starfire during Lunar Eclipse landed after Lunar Eclipse was over, it would still have the 100% crit chance but be able to proc Solar Eclipse. Is one of those two not correct? I'm pretty sure my last Lunar Starfire has procced Solar Eclipse a lot more than 60%.

Starfire does not get the Eclipse benefit if it ends after Eclipse expires. It never has. This is different from the WiseEclipse phenomenon that took place during the 3.2 patch cycle.