WoW.com Guest Post: Why Eclipse is broken

Chris Dinwiddie is better known as Graylo from the popular moonkin blog Gray Matter. "I started my blog to talk about all the things in WoW that interest me. It turns out that my interests are fairly narrow, and my blog primarily focuses on moonkin raiding with the occasional aside to celebrate a boss kill or mourn a near miss."
Ahh, Eclipse. The talent the moonkin love to hate, or do we hate to love it? I don't know. It's a complicated relationship, but I would be surprised if there is another talent out there that stirs up as much emotion as Eclipse. On one hand, it is the savior of raiding moonkin everywhere. Without it moonkin would not be raid viable, and would be stuck with what some consider to be a very boring rotation. On the other hand, it can be a source of despair when you look at the WoL parse and face tough questions from a raid leader.
It is widely accepted with in the moonkin community that Eclipse is broken, but many moonkin don't seem to understand what the real issues are. In this post I will take a look at the common misconceptions about Eclipse, and highlight some of the issues that make Eclipse a broken talent.
Randomness is not the issue
The most common complaint against Eclipse is that it is too random. The critics will site examples where they would cast Wrath fifteen times in a row without an Eclipse proc to show that the proc rate on Lunar Eclipse is to low. This conclusion is wrong for two reasons.
First, random numbers are random. I completely understand how frustrating a run of bad luck is, but it is no different then any other class. A mage can cast Arcane Blast 15 times in a row without a critical strike as well. In the long run, these periods of bad luck will be balanced by periods of good luck, but people tend to remember the bad luck better then the good luck. For example, I can remember only one time where someone set me an email or a tell about their good luck (Grats Murmurs on you 65% Wrath crit rate on Rotface.), but I've seen and received countless messages about how someone was destroyed by RNG. This is not a problem with randomness. This is a problem with perception.
Second, the Solar Eclipse and Lunar Eclipse proc rates are fairly similar. Yes, Solar Eclipse is procs 100% of the time when you crit with Starfire, as apposed to 60% with Wrath, but Wrath is also a much faster cast time. In the time it takes you to cast one Starfire, you can cast 1.8 to 2 Wraths. I won't detail the math for you here, but at Naxx levels of gear the two proc rates are very close. At ICC levels of gear the Solar Eclipse proc rate is about 5% higher but that is primarily due to Starfire having a higher crit rate. If Blizzard wanted to equalize the proc rates, they could increase the Wrath rate to 70% on crit instead of 60%, but more then that would clearly favor Lunar Eclipse.
Eclipse is too powerful
The first problem with Eclipse is that it is now the 800 pound gorilla in the balance tree. For most moonkin Eclipse is responsible for about 14% of their DPS. To put this in perspective, most talents boost DPS by around 3%-5%. As you can see Eclipse is way over budget in comparison.
On the surface this may not seem like a big deal. It gives moonkin more DPS, and that is rarely a bad thing. However, if you look a little closer you will see it is a problem because it amplifies all of the other problems with the talent. If it was a just a minor DPS increase, then who would care that it has big limitations in some fights? So, unless the talent is designed perfectly, having a big talent also means having big problems.
Eclipse and scaling
The size of Eclipse not only amplifies other problems but it actually causes other problems as well. The prime example is how moonkin scale with crit rating. A lot of people assume that Eclipse increases the marginal value of crit rating for moonkin because Eclipse procs off of critical strikes. In reality Eclipse hurts the marginal value of crit rating relative to other stats more then it helps. This is due to diminishing returns felt as critical strike chance increases.
To put it in less mathy terms, increasing your critical strike chance increases the DPS of only part of your rotation. If you have a 50% crit chance and cast 100 spells, the 50 spells that would have crit anyway won't crit harder because your crit chance is increased to 51%. So, as you increase your critical strike chance, there are fewer and fewer spells that can be affected by increasing your crit rating. This decreases the marginal value of crit rating. When you have a buff like Lunar Eclipse that increases your crit chance by 40%, the marginal value of crit rating goes down a lot. In fact, the Lunar Eclipse buff is so large, that moonkin are able to cap their critical strike chance with normal ICC 25man gear, thus reducing the marginal value of crit rating to zero for Starfire while Lunar Eclipse is active. This is important, because it means other classes will improve with better gear more quickly then moonkin will.
Eclipse interrupted
There is one issue with Eclipse that the moonkin community understands very well. That is the problem of losing Eclipse procs due to movement or other fight mechanics that prevent moonkin from DPSing.
Every class has some issues with movement heavy fights, but for most classes, their key abilities and procs aren't necessarily rendered useless by poor timing. For example, a fire mage can still use Hot Streak while moving because it makes Pyroblast an instant spell. Warlocks can still get the full benefit of Backlash and Backdraft, because the buff duration is much longer then the time needed to used the charges.
Eclipse is different. Let me use Professor Putricide as an example. Phase 2 of the fight is like a ping pong match where moonkin are constantly run back and forth across the room. Usually it is possible to cast a few spells on each end, but rarely do moonkin have the opportunity to take full advantage of a 15 second long Eclipse. Therefore, the most powerful talent moonkin is dramatically diminished on any fight with significant interruptions, and few fights very few let you stand still and DPS.
The good news is that Blizzard agrees. In the recent Developer chat on Twitter, the developer's commented that "losing an important proc because it happened at the wrong time (say just before you moved) is a problem." They also commented that they have some ideas on how to fix the issue in the next expansion.
Limited opportunities for Eclipse
The last issue I will talk about has to do with the cooldown on Eclipse. Since neither Eclipse buff can be procced while either of the buffs are active, the amount of time moonkin have to proc Eclipse in a fight is quite a bit less then the actual fight length. For example, if a fight is 5 minutes long, a moonkin can proc can proc Eclipse a maximum of 20 times. If Eclipse procs 10 times, the moonkin had only 2.5 minutes to proc those Eclipse buffs. On the other had, lets say an elemental shaman has an average cast time of 1.5 seconds on his Lightning Bolt. In that same five minute fight he can cast Lightning Bolt 200 times and all of them have the potential to crit no matter what happened before. To put it in more mathy terms, moonkin have a smaller sample size for their procs then most other classes.
A lot of players confuse this problem with randomness, but it is really a problem with the variability of results. All classes can have a streak of bad luck. However, when a moonkin goes without an Eclipse proc for an extended period of time, it has a much greater impact on his DPS then when a Shaman goes through a similar dry spell. The cooldowns on Eclipse mean moonkin have less time to rebound while the shaman still has the full 5 minutes. This issue with a smaller sample means there is greater variability in moonkin DPS because outliers are more probable then they are for other classes.
TLDR
Eclipse is broken as most moonkin believe, but not in the way that most of them think. Randomness is not the problem. All classes have to deal with randomness to some extent. They all have random procs or crits, and in the long run bad luck will balance out with good luck.
The problems with Eclipse are much more structural. First, Eclipse is way too powerful and dominates the moonkin DPS calculation. Second, the buffs are so large that they limit a moonkin's ability to scale with gear. Third, moonkin are unable to control when Eclipse procs, and are likely to lose a significant portion of the buff to movement or another fight mechanic. Finally, the cooldown on Eclipse limits a moonkin's ability to proc Eclipse, and makes moonkin DPS more variable then that of other classes.
Despite the wishes of some players Eclipse is likely a permanent part of moonkin DPS, and will continue to be based on random procs to some extent. The good news is that Blizzard realizes there is an issue, and have a plan to change Eclipse in the next expansion.
WoW.com has tapped some of the best and brightest writing talent around the Internet to contribute a guest post to our site. Interested in submitting to WoW.com? Keep an eye on our Twitter feed for topic call outs and be sure to register with Seed.com for submissions.
Filed under: Druid, Guest Posts






Reader Comments (Page 1 of 5)
Groth Feb 4th 2010 11:06AM
Do we love to hate it or hate to love it?
It depends on which procs first.
catharsis80 Feb 4th 2010 5:52PM
Groth? As in Grothania of Arthas? I wouldn't have asked if it weren't for your icon.
Lissanna Feb 4th 2010 11:13AM
Part of the problem is that they had to make it a significant portion of our DPS because otherwise we had to deal with the same mechanic problems for not enough reward (so people wanted to just starfire-spam anyway). It used to be something on a couple minute cooldown that wasn't a very big portion of our DPS, so we accidentally got it buffed to the horrible place that it's at today, because it's just not a very good/functional design.
Murmurs Feb 4th 2010 11:43AM
Well, I suppose we are mostly likely to blame for it. I mean, honestly, we did start the chain of events that lead to the current version of Eclipse. The changing of the cooldown itself was of little significance; that change was needed as it made Eclipse worth something. On it's original cooldown, Eclipse was an utterly worthless talent. The problem was that we steered Blizzard away from the original values.
All of the values were originally 15%. This was fine, but, for Starfire, it was far too RNG heavy. Although 15% additional Crit is a lot, it just wasn't making the talent all that worthwhile since, at that time with gearing factors we knew about, you could very easily get RNG screwed with no gain from Eclipse. In fact, I recall plenty of times in Beta where I went an entire Lunar proc without a single Starfire Crit.
So we made them buff Starfire to 30%. This, of course, ended up unbalancing the Wrath portion to the point that, aside form the Nature's Grace factor, it was a huge loss in DPS value to get a Solar porc. So it too had to be buffed to 30% in order to equalize it. And so on it went.
Simply put, it and when Blizzard nerfs Eclipse, the Starfire effect cannot remain as a Critical Strike increase - it must be changed to something else. If not, then too low and the effect isn't noticeable enough - too high and the talent falls back into where it is today.
Astemus Feb 4th 2010 11:48AM
I think what they should do to fix eclipse is remove the cooldown and duration, and instead, make it a permanent aura that switches at a certain number of casts. Spam wrath at the start because you're sitting on solar eclipse, then it procs and swaps to lunar eclipse, so we start spamming starfire. This would maintain the increase we need to stay competitive and eliminate the problems that exist today, as well as remove the randomness of losing eclipse.
Obviously, the effects would have to be scaled down to account for the 100% uptime, which would solve the last problem of poor scaling at high gear levels.
Astemus Feb 4th 2010 11:52AM
Or possibly make the switch a random proc based on crits as it is now. That way we are still dealing with the proc mechanic to keep it interesting.
Lissanna Feb 4th 2010 11:59AM
In the end, the "fix" has to be adding at least 1 (pref. 2) new single-target spells to moonin's rotation and removing Nature's Grace from the picture to make haste scale better with wrath. Eclipse on it's own isn't the real problem. The real problem is that our 4 spells don't make an interesting rotation on their own, and Eclipse was just a way of trying to make us press those 4 buttons without giving us anything in our rotation we can actually control the timing of, besides Treants.
When people say "pop your cooldowns", I just stand there with nothing to pop unless my treants are up. I have so many random procs between talents & gear that my DPS swings totally all over the place depending entirely on how much I have to move relative to when the procs go off.
Lissanna Feb 4th 2010 12:00PM
basically, Eclipse is a symptom of a much bigger problem for moonkin, and it just doesn't quite work on it's own.
Tyler Caraway Feb 4th 2010 12:26PM
I find it a little amusing how often I hear that, Liss. I mean, I agree with the statement, but, honestly, how many other classes use as many buttons as we do? Arcane Mages use two, Fire Mages use three, maybe four, Destruction Warlocks use four, Shadow Priests use four - five if you count the one application of SW: P - Elemental Shaman use three to four depending on Chain Lightning. We really are no different than any other class out there. Very, very few classes use a wide plethora of abilities.
We need a new spell because, well, we haven't actually gotten one that wasn't a long cooldown button in....ever.
Graylo Feb 4th 2010 1:31PM
I wrote this post to highlight the real problems with Eclipse. In no way do I think Eclipse is the only problem in the moonkin spec at the moment. I completely agree that Nature's Grace is broken, and Starfire and Wrath are to similar. However, I wrote this post to focus on Eclipse.
On another topic, I want to disagree with Tyler/Murmurs for a second (big surprise. :P). I don't think we are to blame for the current state of Eclipse. Talents are not valued in binary terms. They are not "good" or "bad". There is tons of gray in the middle. When Eclipse was first seen in the Alpha it was a horrible talent. You, I, and several other people did the math to show that it was true. They revised it several times and we eventually got to a talent that was acceptable at the start of WotLK. Blizzard could have stopped there, but they didn't.
I don't remember any prominent theorycrafter calling for them to buff Eclipse in 3.2. There were a lot of people calling for buffs, but I don't think anyone wanted them to happen to Eclipse. In 3.3, we needed a buff again, but both you an I called for something other than an Eclipse buff. This issue is all Blizzards. They didn't want to do anything to risky or boring to buff us so they kept going to the well that they new a lot about. Now that well is dry.
Lissanna Feb 4th 2010 1:30PM
Yes, but - Blizzard wants us using a lot of buttons and wouldn't let us just starfire spam all day to happiness. If they want us to push more buttons, the best way to do that is to give us buttons to push. :)
Tyler Caraway Feb 4th 2010 1:42PM
To 3.2 I agree with, but we are at fault for the 3.1 change which boosted Wrath to 30%. There were many complaints that the Solar portion of Eclipse wasn't strong enough, and since nerfing Lunar was out of the question, Solar was buffed as the only logical choice. In that respects, we are at fault.
The 3.2 change to allow us to alternate Eclipse procs, though, was the largest change we'd ever seen and was what solidly placed Eclipse as being one of, if not the, single most overpowered talent in the game. The massive amount of DPS that Eclipse accounts for is absurd, and to that end, I do agree that I don't think we ever - at least not of any serious note - made suggestions to support that change. By that point I think most people were already on the 'nerf/destroy Eclipse' wagon.
And, of course, the 3.3 change was an expected icing on the cake.
Lissanna Feb 4th 2010 2:13PM
It's my fault because every time I complain about Eclipse, it gets buffed instead of "fixed". We can just blame me.
Actually, if it gets buffed again after this post, I blame Graylo.
Thomas Roach Feb 4th 2010 11:25AM
Perhaps they could make the damage increase of Wrath 20%-30% instead of 40%, but make it an instant cast spell? I have found that my haste gets Wrath down to around 1 second anyways. For a Solar Eclipse to grant instant cast Wrath would just give us some mobility, without increasing our cast speed. This would benefit fights in PvE like Putricide and might make Moonkins less than terrible in PvP.
Hollow Leviathan Feb 4th 2010 11:59AM
Making spells instant cast just means that haste rating will no longer benefit your dps as well as crit rating, and just make boomkin dps scaling with gear even worse.
To avoid bad timing ruining your dps, why not make the eclipse give a buff with charges on it that expires over a very long period, a minute or two. Such that they can control the DPS gain from the talent with the amount of charges the buff gives you, while giving you the control to have to move and then use those charges on your own time.
Branalia Feb 4th 2010 3:57PM
What if... Lunar eclipse changed Wrath into a five second channeled spell affected by haste that would shoot off one wrath bubble every second until the channel ends? Then we could squeeze in extra wrath balls under the GCD but at the cost of movement.
Branalia Feb 4th 2010 3:58PM
And yes, I meant Solar Eclipse, not Lunar... but there is no edit button! Whee!
Aedor Feb 4th 2010 11:29AM
Coming from a raiding Boomkin who loved topping the charts in Black Temple and has the 2nd highest gearscore in the raid (under the tank - over 3K on wowheroes), it's painful to always be 3rd to 5th in every fight in ICC now. Something must be down, I don't want to have to shelve my favorite toon for my Spriest or Retadin.
Aedor Feb 4th 2010 11:30AM
10 man btw.
Lissanna Feb 4th 2010 11:35AM
the problem with being lower on DPS meters isn't just the fault of Eclipse, but is a bigger problem with moonkin design as a whole.