This is close to truth, but the way spell queueing works is that if you have no haste or NG and a cast time of 1.5, your casting-interval will be 1.5 because you can place the spell in queue before your last one finishes casting.

Interesting observations.

So for Oceanic players like me who commonly suffer 300 ms or worst, even with 2pT8.5 bonus, it is still more beneficial to cast SF despite solar eclipse, while the opposite is true for players with >150 ms(cast Wrath if Solar, cast SF for anything else).

I'm not so sure about that. Initially I did think this but the more I check my logs the less true I find this.

Example:World of Logs - Real Time Raid Analysis (First guild kill of Anub, my dps is all over the place I'm afraid but the first Anub section before he burrows at the start is what interests me here).

The first large dps spike is a Bloodlusted Lunar Eclipse, so ignore that but look at the following three dps spikes after that. They reflect a Solar, Lunar, Solar Eclipse cycle and all three are extremely close in terms of damage and this isn't due to trinket procs on the Solar Eclipses.

And this is for a Moonkin with a measly 2174 odd SP and 19% crit unbuffed and being haste soft capped (actually crit was lower for the fight, was wearing [Pants of the Soothing Touch] during the fight). I imagine for a better geared moonkin with more crit we'd see a better return on Solar Eclipse cycles.

I don't follow. The first wrath should suffer the same 200ms latency as the second wrath. So as long as you fire your wraths at a 1s interval from your client then the server-side should receive both wraths at a 1s interval.

This post did make me think, however, of changing my personal soft-haste cap based on my latency. I don't think it's necessary anymore. Thoughts?

There are too many factors in a boss fight that can influence your DPS and one sample can not be taken for proof. But there is a much simpler way to test the theory: go to a level 80 training puppet and measure your average DPS by only casting Wrath at it (ignore solar eclipses). Then repeat the same experiment by only casting Starfire at it and compare both DPS.

Wrath has by math already more DPET than Starfire because it scales better with spellpower. But I would be surprised if you discovered that you do more DPS with Wrath than with Starfire against a training puppet.


Ok well ... actually you might because Lag isn't usually high around puppets.

A pitty that we can't revert to the old way of a common eclipse CD. The best way to go was simply switching from solar to lunar rotation. I can only testify that my Raid DPS considerably improved once I switched from solar to lunar rotation back then.

I completely agree that one sample isn't conclusive proof of anything. I was merely showing you an example in a simple phase of a boss fight (i.e. assigned to Anub in the first phase before he burrows is as simple a tank and spank as we have in Crusader's Coliseum). During the fight the dps spikes during Lunar Eclipse procs was fairly consistent when Anub was up. It's highly unlikely that the the lunar eclipse spike at the start after the bloodlust was an extreme outlier with this being the case which means that it's quite likely that lunar and solar eclipse procs at my gear levels are fairly consistent in terms of damage.

What kind of a drop were you seeing with solar eclipses?

I recall my raid DPS going up about 15-20% when I forced myself to use Lunar rotation back then. Even with the training puppets my Lunar runs were generally higher dps than my solar runs.

I don't want to carve those numbers in rock either. I only know that I was generally displeased by my raid performance compared to other raid members until someone suggested I should try to follow some guides instead of using my own mathematical conclusions. All guides I've read at that time favored the lunar rotation. So I did like a dozen mock fights at the training puppets with both rotations and discovered myself that I had it easier to achieve high DPS when using lunar.

I guess we all know how hard it is to nail down moonkin DPS to a specific numer with all the RNG involved in our mechanics.

In the end, isn't all of that moot, since the 3.2 changes favor using both?

It makes a difference because if Solar Eclipse is much lower in terms of dps than Lunar Eclipse for someone then if there was an unlucky string of non-crits after Lunar they could find themselves in Solar when Lunar Eclipse's cooldown was over and in which case they probably should cancel Solar Eclipse to get back into Lunar asap.

Personally I'm not seeing this dps drop in practice, so I'd be inclined to only cancel Solar Eclipse in the above situation during Bloodlust where Lunar is far superior to Solar, but this mightn't be the case for everyone so it's arguably an open question where it really does come down to you whether it's better to end Solar Eclipses early or not.

I think it has some merit in Bloodlust phases. Personally I wouldn't dream of using it outside of Bloodlust. The idea behind using outside of Bloodlust is to maximise the number of Lunar Eclipses in a fight, I've not worked out the maths to see if this is worthwhile though. With a large enough disparity between Lunar and Solar dps it probably would be recommended but I don't think that disparity is there outside of Bloodlust.


Edit: To expand on the above:

We want to maximise the number of Lunar Eclipses during the 40 seconds of Bloodlust for obvious reasons. Ideally we can fit one and two thirds in if we proc Lunar with our first Bloodlust Wrath cast. Then once Lunar is off cooldown we need to get it back up asap to maximise dps over the phase. Every extra second spent in Solar past this point is potentially a second we could have spent in Lunar. In order to maximise average dps in Bloodlust phases we should cancel Solar unless there's only a few seconds of Bloodlust left to go. Assuming an early Lunar proc during the Bloodlust phase we should almost always cancel Solar once Lunar is off-cooldown. It's only in situations where we go into Lunar late (i.e. close to 10 seconds into Bloodlust) that this is a dps loss. Where we proc Lunar just before Bloodlust is cast then the argument for cancelling Solar becomes even stronger.

The same logic then applies to someone who for whatever reason sees substantially lower dps during Solar phases. On average, given enough off a difference, it could be a winner, though you'd have to model it in order to work out how much of a difference would make it worthwhile.