Monday, January 30, 2012

My MoP Wish List - Part 2: Release and Moonkin Issues

Last week I posted part 1 of my MoP wish list focusing on a Gem Overhaul.

Here is part 2 of my MoP wish list, and most of it will focus on the things I would like to see for the Moonkin play style, but the first item is an exception.

So, lets get started.

A 2012 2Q or 3Q Release:

Oh god do I hope for a late 2Q or early Q3 release of Mists of Pandaria. I am really not sure I can keep playing Dragon Soul for 10 more months. I don't think anyone would oppose an earlier quicker release of the new expansion, but from what I've seen a lot of players are clearly skeptical of Blizzards ability to do so.

Yes, according to the spreadsheet that was leaked a couple of years ago, the planned release time frame for MoP was 2Q 2012, but lets be honest. Blizzard has been saying for years that they would like to release patches and new expansions more quickly, yet it still took them 2 years to release WotLK after TBC, and it took them 2 years to release Cata after WotLK. We also have yet to hear about Mop Alpha and Diablo 3 is way behind schedule.

That said, I'm optimistic. MoP shouldn't have some of the challenges that previous expansions had like completely reworking the old world and judging from the amount of information that has already been released, it seems like Mop is already farther along then most would expect before Alpha testing. That said, if we don't hear about alpha testing soon, I think we are looking at another holiday release.

A Simpler Moonkin:

During the WotLK expansion, we Moonkin complained and complained about Eclipse and the general state of Moonkin DPS, and the funny thing is part of me looks back on that time with envy. To be fair, I'm sure my some of my hindsight is tinted with rose colored glasses. The old model had very real and very big problems, but it wasn't the complicated mess that we have to day where players have to make choices on the fly with limited information.

The current moonkin DPS model is way to complicated in my opinion and I don't find it to be that fun. As far as I know, no other spec is expected to make the number of judgement calls we have to make just to do average DPS. What other class has to clip their DoTs frequently to make sure they are fully buffed? Which other spec, has to decided if it's better to cast a cooldown based Nuke now or later?

The good news is that according to the information already released about MoP, things should be improving. From what I can tell the changes that have been proposed so far, hit the mark of what we need. The changes to out DoTs will simplify them, but won't make them mindless and the change to Starfall ends the debate of when to cast it.

There are still things that need to be addressed but I like the way Blizzard is headed.

A Real Movement DPS Solution:

I have to give Blizzard credit for trying, but I think everyone would agree that the first attempt was a failure. At it's peak, Lunar Shower was used more as a Nuke then a movement DPS solution, and now at the end of Cataclysm it is almost universally ignored. Blizzard has said they want to revamp Lunar Shower for MoP, but the question I have is into what?

Here's the problem as I see it. All movement is very disruptive to casting and has a significant impact on DPS, because all movement will interrupt your cast. This is particularly bad for Moonkin since Starfire has the longest cast time of all the core nukes from all five caster DPS classes (yes, I know pyroblast is longer, but it's not really a spammed nuke from what I understand). This means that on average Moonkins will lose more casting time due to movement then other caster classes because our Nukes more time to cast on average. On top of that, our core DPS mechanic Eclipse is based upon us completing our casts. Therefore movement also means that we will transition between Eclipses more slowly again reducing out DPS in multiple ways. To be clear, having to move is bad for all DPS specs, but from what I can tell, Moonkins are one of the hardest hit when movement occurs.

The other half of the problem is that most movement isn't persistent and only lasts a few seconds at most. As a result, the solutions I've seen Blizzard implement so far have been a bit underwhelming. Take Lunar Shower for example, for it to have it's full impact it needs to be fully stacked which would take 3 seconds at the absolute minimum, and most movement would be over before it's fully stacked. The other example is the Shaman spell Spiritwalker's Grace that allows Ele's to cast while moving for 15 seconds. It's definitely, a nice thing to have for those long preplanned repositioning movements, but with a 2 minute cooldown, it has less of an impact on the short and spontaneous movements, that tend to be more of the norm in my experience.

As I'm sure you can tell, I'm not happy with either of the solutions above, but to be honest I don't have a great suggestion either. One idea I have liked, (but can't remember where I heard it) was what if they allowed casters to cast while moving but at a slower rate. For example if you start casting a 3 second spell but move for one second during the middle it would take 4 seconds in stead of 3, because that middle second took twice as long. It sounds like a very interesting idea, but I'm sure the technology needed to do something like that would be very complicated, and possibly prohibitive.

A Moonkin Execute:

How many of you have been in this situation? You're fighting a boss, and doing a respectable DPS. You take a quick look at the DPS meters and lets say for arguments sake you are number 5 in a 25man raid. Then, we get to the end of the encounter with just 25% health left on the boss and by the end of the encounter you're 10th instead of 5th. Does this sound familiar to anyone else?

DPS can be very hard to judge mid-fight especially when you have a mechanic like eclipse that by definition makes your DPS streaky, but I'm tired of getting trashed in the "execute phase" of a fight. Moonkin need some sort of ability or buff that allows us to do ramped up DPS at the end of a fight just like most of the other DPS specs in the game. That said, it needs to be done with care, so that we aren't once again crippled by complexity.

Tuesday, January 24, 2012

My MoP Wish List - Part 1: A Gem Overhaul

Quick Note: This is part one of three(ish) posts I am planning on making regarding what I would like to see happen/change in the upcoming Mists of Pandaria expansion. I say three(ish) posts, because what I had originally planned to be just one post has grown in to four so far. Last weeks post on the current epic gem situation grew out of his idea, and became too big not to be it's own post. Today’s post was also meant to be apart of another post but grew to big as well. Who knows, maybe this idea will keep me posting into March.

With the Mists of Pandaria expansion getting closer I've been thinking a lot about what I would like to see in it from multiple points of view. While we already know more about MoP then we have known about the other expansions at this point in their development, we are also still at a very early stage and have an opportunity to shape it's development, with some good suggestions. So, that's what I'm trying to do in general with these posts. Not all of the suggestions are serious or incredibly significant. At least one of them doesn't have a snowball's chance in hell of actually happening, but the are things I would like to see. So, on to part 1.

A Gem Overhaul:

Last week, I made a post about how I felt the Cataclysm implementation of epic gems has been a failure, and it looks like I am not alone in that sentiment. In that post I made several suggestions that Blizzard could implement to ease the situation, but none of them addressed what I consider to be the biggest underlying cause of the situation. The unbalanced demand for the different gem types, due to how the gems are currently organized by color. One of the changes I would like to see in Mop is for Blizzard to reorganize the system to make the other gems more desirable. Here are a few scenarios thought of or heard from others:

Current Scenario:








RedYellowBlue
IntellectMasteryStam
AgilityHasteSpirit
StrengthCritHit
ParryResilienceSpell Pen
ExpertiseDodge

I'm listing the current scenario, because despite it's flaws, it also has some significant advantages.
The advantage of this system is that it isolates the primary stats from most of the secondary stats. It's actually very smartly designed, from a stat accessibility point of view. Think of every Primary/Secondary stat combination a player might want and almost all of them are possible in this scenario. Parry is fine as a red because DPS don't want it. Spirit and Spell Pen are fine as blue because tanks won't want those. The only accessibility issue you could have with this design is that melee can't realistically gem for Expertise in this system and tanks can't gem for hit, both of which I think are minor points.

As I said last week, the main problem with this scenario is that it creates a huge imbalance in demand for the different gems. 90+% of the players in the game want red gems, and the other gems are at best situationally desirable.

Suggestion #1: Gem Redistribution








RedYellowBlue
IntellectAgilityStrength
HitMasterySpirit
CritResilienceStam
ParryDodgeSpell Pen
ExpertiseHaste

The advantage to this scenario is that it keeps primary stat gems in the game and it spreads them around the different colors to create a more balanced demand. The red gems will still be the most popular since there will likely be more Int users in any group then there are Agi or Str users, but the green and blue gem type drops won't be the trash that they are today.

Unfortunately this scenario has a huge problem. It blocks a lot of Primary/Secondary stat combinations that players will want. Just one example is that Mages and Warlocks would be giving up a lot to gem Hit since it shares a gem color with Int. While it would balance out the demand for the gem types, I think this is to high a price to pay.

Suggestion #2: No Primary Stats







RedYellowBlue
HasteMasterySpirit
ParryResilienceHit
ExpertiseDodgeSpell Pen
Crit

As Rohan suggested in the comments of my last post, it might be a good idea to eliminate primary stats from the system all together. Currently, primary stats are universally every players first choice when trying to pick a gem or enchant for their gear. How players value secondary stats is much more diverse however. Some prefer haste and others want mastery. If implemented this suggestion should lead to more balanced demand because of the more diverse interest in secondary stats. That said, I'm sure there will still be favorites and least favorites, but hopefully it wouldn't be as extreme as the current situation.

The down side to this suggestion is what it means for profession bonuses. The profession bonus for both Jewelcrafting and Blacksmithing is based upon gems, and altering which gems are available will have an impact on which professions are most desirable. However, this concern is manageable. The JC only gems could still provide primary stats and the BS gem sockets could be converted to enchants that provide a set amount of stats rather then a socket to be defined by the user.

Suggestion #3: No Pure Secondary Stats








RedOrangeYellowGreenBluePurple
IntellectInt/CritAgilityAgi/HitStrengthStam/Parry
Int/MasteryStr/HitStamStam/Exp
Int/hasteStam/DodgeInt/Spirit
Int/ResilAgi/ResilInt/Hit
Int/Spell PenAgi/ExpStr/Resil

This is by far my most radical suggestion, but I have to ask the question. What's the point of having pure secondary stat gems anyway? As far as I know almost no one uses them, because the primary stats are so much better. So here's the radical suggestion: Get rid of them.

The table above gives you and idea of that I'm thinking about, but by no means includes all the combinations that would be necessary. Each of the Primary stats would have it's own primary color gem, with Strength and Stamina sharing blue. The hybrid gem colors would provide two stats just like they do now, but they would always be a combination of a primary stat and a secondary stat, and the secondary stat wouldn't be tied to a particular color like they are now (ie: Resilience =/= Yellow). Also, if having no pure secondary stat gems is a real deal breaker, who says you can't have pure stat gems in hybrid gem color.

The advantage of this system is that it should broaden the demand for the different types of gems and should allow for every primary/secondary combination desired by players.

The disadvantage to this system is that it is less intuitive than the current system and since it's a fairly radical change, it may confuse a lot of players. I also worry that it may not broaden the demand for other gems enough. Reds would still easily be the most popular gems with blues and yellows coming in second and third. The hybrid gems on the other hand could still lag significantly behind.

TL-DR:

While the current gem structure has it's advantages, it's led to an extremely lopsided market for gems in WoW where Reds are valued significantly more then all of the others. There are several ways that Blizzard could modify the system so that the gem valuations are more balanced across the different types.

1. Blizzard could just reassign a lot of the gem types to different colors but this would cause situations where some players couldn't get a hybrid gem combination that they found desirable.

2. Blizzard could also remove primary stats from the gem structure and reassign some of the secondary stats to different colors. This should balance the demand for gems, but would cause issues with profession bonuses.

3. In a fairly radical move, Blizzard could also eliminate pure secondary stats gems because very few people use them anyway. This would allow Blizzard to create hybrid gems that weren't as reliant on gem color and could allow for all the combinations a player might want.

Friday, January 20, 2012

Epic Gems in Cata = A Failure


Quick Announcement: I'm trying to be more active on twitter. If you want to follow me, my ID is @graylo.

Last night my guild killed Heroic Hagara for the second time and not one, but two [Girdle of the Grotesque] dropped. As I'm sure most of you know, this is the BiS moonkin belt, and I was quite happy when I was awarded one of them. Then I had a revelation.

The GotG has two Red sockets and gains a prismatic socket once you apply the belt buckle. Since it is a BiS item I would expect it to be gemmed with epic quality gems even if they are a bit expensive, because I'm not going to replace it. Then I have to consider the reality of the situation.

The mean price of Queen's Ruby's on my server is 7,000g. Sure the price varies a little bit, but even the lowest I've seen them on my server is still 5,500g and that is in no way common. In short, I have to spend at least 16,500g just to gem my belt the way it should. You can buy epic items for less then that.

It's clear to me that the current system is broken. So, how did we get here, and how can we fix it?

A Quick History:

In The Burning Crusade (when the JC profession was introduced) epic gems were mined from special mining nodes in the Black Temple and Mount Hyjal raid instances. Eventually they could be purchased with badges.

Pros:
  • It supplied epic gems directly to the main group of players interested in having epic gems: hardcore progression raiders.

  • Allowed the guild to allocate gems based upon need.

Cons:
  • PvPers had extremely limited access to the epic gems, and but they were pretty much required at the higher levels of competition
  • Aquiring the gems was a 25man effort, but single players could cause havoc with ninja looting and favoritism

In Wrath of the Lich King, epic gems were prospected from uncommon ores. Alchemists could also transmute superior gems in to epic gems once a day with some additional mats.
Pros:
  • Epic gems were highly accessible for all players.

  • Players controlled both the supply and demand sides of the gem market.

Cons:
  • Epic gems were so accessible that it crippled the market for blue quality gems.
T
his is pure speculation but here is what I think happened. In TBC, Blizzard introduced epic gems as a reward for high progression, but in the end weren't accessible enough. First, things first the guilds who mined them horded them for their own needs. On top of that, obtaining the gems was a 25man effort, but one or two people cause big problems by ninja mining the nodes or a GM playing favorites with the gems. There were other issues as well, but I think all of them revolve around the epic gems not being accessible enough.

So in WotLK, Blizzard created an entirely new system that was completely driven by the ingame economy. Players determined the supply by mining/prospecting ore or doing transmutes, and as a result Epic gems were probably too accessible in Blizzards mind. During WotLK, epic gems were fairly cheap. Players completely regemmed gear within a week or two of their introduction and the market for blue quality gems was crippled. As a result Blizzard felt the need to create a new system for Cata which leads us to......

The Current Problem:

The current problem can easily be broken into two sections: Supply and Demand.

Demand side:
The problem with gem demand, is that it is heavily skewed towards red gems at the moment. DPS and Healing are the two most common roles played in the game at the moment and almost all of them favor the stats provided by red gems (Int, Str, & Agi). On top of that, Blizzard likes to give players their best stats in the last tier of an expansion. As a result, the T13 year, has significantly more red sockets then previous tiers which reduces the demand for hybrid gem colors like purple and orange.

In short, 90% of the players in a raid want red gems, but only 17% of them will get one on average (assuming the distribution is equal). As a result, red gems are extremely expensive, but the other five gems are relatively cheap. Here are the mean prices on my server as an example.



RedBlueYellowGreenOrangePurple
7,000g599b1385g999g2060g1325g

Ultimately this is a problem Blizzard created 2 years ago when the started designing the Cataclysm expansion, and can't really be fixed until the next expansion even if blizzard wanted to.

Supply side:
The Cata system for epic gems does have several positives. For example, every player who kills a boss on normal or heroic levels has a chance to get an epic gem. The supply isn't controlled by the guild, and players have a choice to use it them selves or sell it for gold.

This system does eliminate some of the core problems of the previous two systems, but it's too random and inflexible to meet the players needs in a reasonable way. It creates a surplus of all colors except red, and an extreme shortage of red gems which increases prices to unreasonable levels.

The good news is, that changing the supply should be easy, and Blizzard has several options if they so desired to make a change.
  1. Change Drop Rates: There is no rule that says all gems have to have an equal chance to drop from the Geodes. I bet it wouldn't be difficult for Blizzard to modify the drop rates. They could make epic gems more common on the more common Geode, or they could change it so that red gems are favored over the other 5 options for both types of Geode. My guess is that Blizzard could easily implement this with a hotfix.

  2. Eliminate the Geodes: Blizzard could eliminate the geodes all together and allow players to buy the gems directly, instead of buying a random chance at an epic gem.

  3. Enable Gem Transmutes: If I remember correctly there were some transmutation patters for gems in the 4.3 ptr, but they never made it to the live release. This would allow blizzard to maintain the randomness of the geode, but still give players a way to modify supply, even if it's with a weekly cooldown.

Conclusions:

As much as I would like them to, I doubt Blizzard will make a change at this point. Prices will come down as people are getting less gear upgrades, and more gems are being created. That said, when I talk to players, this epic gem situation is a common source of frustration. Since we are entering the lull between expansions when burn out among players is high, I think it would be a good idea for Blizzard to take a look at this issue and try and reduce some of those frustrations.

Thursday, January 12, 2012

FWSC: Upgrading from 4T12


In my last post I addressed the value of the 4T13 set bonus because before it was buffed in PTR patch 4.3.2 it was widely regarded as very weak and several people asked me if it was a set bonus people should strive for or just ignore it and use other gear.

I think I answered that question sufficiently last week, but I was surprised by how many of you asked when we should upgrade from 4T12, and weren't even concerned about 4T13 yet. I had assumed (somewhat wrongly) that upgrading to 2T13 was a given based upon my experience and where items fell on my gear list. After thinking about it a little bit, I realized it's not that easy.

In the past, upgrading gear between tiers had usually felt fairly clean to me. The values of the set bonuses did very some, but that was usually small and the difference was made up for with improved stats. However, with this tier, I didn't take into account the significantly weaker set bonus and I didn't realize that the ilevel upgrade was as small as it is. So, lets stop making assumptions and start looking at some simulations to see what the results actually are.

My Methodology:
  • All of the results generated below were generated out of SimulationCraft using the simc-430-6 build.
  • The gear sets used in the simulations were based upon my own. It was primarily a mixture of Heroic T12 gear and Normal T13 gear, but I did not have a single piece of the T13 Moonkin tier set due to some extremely bad luck with guilds. I chose these gear sets because it was clearly a realistic set up and was easy for me to upgrade the gear.
  • iLevel gains were minimized to avoid confusion. For example, my T12 helm was from Normal mode, so I chose the T13 LFR helm as the upgrade in my analysis.
  • I made a few changes to the default rotation used by SimulationCraft. The main change is that I had it cast Starsurge on cooldown rather then only in Eclipse.
  • All results are from a Patchwerk style simulation.


The Results:

First lets look at the values of the different set bonuses. These will be a little different then what I presented last week, because I'm using a different gear/glyph set up and a slightly newer build of SimulationCraft. While there are differences, I don't think they are significant.




4T12Only 2T12No Set BonusOnly 2T13Live 4T13PTR 4T13
% Change2.05%1.71%1.75%1.23%2.54%
Avg DPS34,93134,22833,65234,24234,66435,111


As you can see, both T12 set bonuses are better then the T13 bonuses that would replace them (at least until 4.3.2 is released). When we trade 4T12 for 2T13 currently we are giving up a 2.05% damage increase for a 1.75% damage increase, for a net loss of 0.30%. When trading 2T12 for 4T13 it's even worse with a net loss of 0.48%. The 4T13 set bonus does get significantly better when 4.3.2 is released, but we are not there yet, so lets see if improved ilevel can make up for our weakening set bonus.

This list represents a series of simulations I ran to see the value in transitioning from 4T12 to 4T13. It includes both the set bonus changes and the ilevel changes.

>






iLvl RankSets UsedSlots UpgradedAvg DPS% Change
04T12None34,922n/a
122T12/2T13S,G34,952-0.06%
122T12/2T13S,L35,1510.51%
182T12/2T13S,L,H35,2990.42%
244T13 LiveS,L,H,G35,3050.02%
244T13 PTRS,L,H,G35,7551.27%


Column Definitions:
iLvl Rank: Represents how much my total ilevel increased in each simulation.
Sets Used: Indicates which set bonuses were used in the simulation. If the 4 piece is used, then the corresponding 2 piece was used as well.
Slots Upgraded: S=Shoulder, G=Gloves, L=Legs, H=Helm
Avg DPS: The average DPS from the simulation
% Change: The percent change in average DPS when compared to the previous iLvl Rank.

What this Table Tells Us:
  1. Upgrading isn't bad: In all but one occasion, upgrading my gear in the simulation resulted in higher average DPS. The one time it didn't increase DPS, the decrease was just 0.06% which is really insignificant. And remember, these simulations were done using the minimum ilvl upgrade possible.
  2. How You Upgrade Matters: The one time average DPS when down was when I upgraded using the two weakest tier slots (shoulders and gloves). When upgrading with the larger tier pieces (Legs, Chest, and Helm), the change in Avg DPS was more significant.


TL-DR:

As long as you use some common sense while upgrading your gear, you shouldn't see any significant DPS losses from going from 4T12 to 2T13 and beyond. That said, depending on what loot system your guild uses, it doesn't always make sense to go for an upgrade right away. If you have a choice, focus on upgrading the larger slots like Legs and Helm, before the smaller slots.

Wednesday, January 4, 2012

FWSC: A Look at Set Bonus 4T13


Since the release of patch 4.3 I've received several questions about the Moonkin 4 piece set bonus, and had been working on this post for a couple of weeks. Unfortunately, it is now a little late given new information from the recent PTR build, but I would like to go ahead and get it out there to answer any questions that might be lingering.

Fun With SimulationCraft: The Moonkin 4T13 Set Bonus

When the set bonuses were reveled on the PTR months ago, the criticism of the 4T13 set bonus was pretty quick and pretty loud. I don't remember seeing a single theorycrafter or analysis that suggested it was a strong set bonus, and I like a lot of people thought it should and would be buffed. However, it quickly became clear that Blizzard wasn't going to buff it (at least before the patch), and the question became: should we use it?

Since the 4T13 set bonus is being change on the PTR to also buff Starsurge's damage by 10%, the answer to that question is a bit obvious now. Of course we will want the set bonus when the 4.3.2 patch is released. But lets take a look at the numbers anyway. Who knows when the new patch is going to be released some of you might find this info helpful in the mean time.

My Methodology:



  • All of the results generated below were generated out of SimulationCraft using the simc-430-4 build.

  • Most of the results were created using the Moonkin Heroic T13 build that comes with the program, but a few of the results were created using a Normal mode Moonkin T13 build that I created. I did this to answer a specific question that will be come obvious later.

  • SimulationCraft obviously isn't updated yet for changes that occurred last night. So, I estimated the benefit of the new 4T13 set bonus by increasing Starsurge's DPS on the old 4T13 results by 10%. I think this is a reasonable estimate, but it may be a little generous since that assumes the buff is multiplicative and it may not be. Either way, I don't think it changes my conclusions and I don't have a way of getting more accurate results at the moment.

  • Rotation Change 1: The default SC Moonkin Rotation has you detonate Wild Mushrooms at the start of the fight when not in Solar Eclipse. I added a condition to only detonate it when Solar Eclipse is up. It was a tiny improvement.

  • Rotation Change 2: The default SC Moonkin Rotation has you only cast Starsurge while you are in Eclipse. I changed it to cast Starsurge on Cooldown. I can see delaying a Starsurge cast a little right before Lunar Eclipse, but it doesn't make sense to me at other times and my results agree with my logic. Let me know if I'm missing something.

  • All results are from a Patchwerk style simulation. I did run the simulations on helter skelter, but it didn't change the results significantly.


The Set Bonus Valuations:

From what I can tell most moonkin are using Glyph of Wrath for most fights in Dragon Soul because Glyph of Starsurge doesn't make sense with a few exceptions. That said, I ran the results using both glyphs for a comparison.

Glyph of Wrath:


























">
No Set BonusOnly 2T13Live 4T13PTR 4T13
% Change1.8245%0.8344%2.2134%
Avg DPS44,72445,54045,92046,548


As you can see from these results the Live version of 4T13 buffs Moonkin DPS by about 0.8344%. This is easily one of the weakest Moonkin set bonuses in the history of the game. However, that doesn't mean it's useless either. While a 0.8344% buff isn't particularly strong, it's not something to be ignored either. I've seen plenty of players expend a lot of effort for a smaller bonus.

Glyph of Starsurge:


























">
No Set BonusOnly 2T13Old 4T13New 4T13
% Change1.7154%1.2554%2.6435%
Avg DPS44,48045,24345,81146,439


These results might give us a little clue as to why Blizzard was hesitant to buff the 4T13 set bonus before the patch went live. I wouldn't be surprised if Blizzard was expecting some level of Synergy between the 4T13 set bonus, Glyph of Starsurge, and Lunar Eclipse. These results show us that that expectation isn't completely baseless since the 4T13 set bonus is significantly more valuable when using Glyph of Starsurge instead of Glyph of Wrath.

That said, the results also show that Glyph of Starsurge isn't always the best choice even if it does synergize well with the 4T13 set bonus. These results are from a single target patchwerk simulation and Glyph of Wrath out performed Glyph of Starsurge in every simulation. Glyph of Starsurge would be better in multi target situations, but only two of the Dragon Soul raids fit that description in my opinion. In three of the fights it may be useful, but somewhat iffy, and in the other three fights Glyph of Starsurge would clearly be bad.

The Alternatives:

Before coming to any conclusion on the usefulness of the 4T13 set bonus we have to look at the alternatives. It's not enough for the set bonus to be bad or sub par. We have to have a better alternative to replace to declare the 4T13 useless. Luckily answering this question proved to be easier then I expected it to be.

Heroic Gear Level:
At the T13 heroic gear level the answer is easy. There literally are not any ilevel equivalent alternatives. The only head, shoulder, chest, hands, and leg option we have at the 410 ilevel are the tier pieces. To be fair, you could use the resto tier instead of the moonkin tier but that only seems to make sense in the chest slot. The other option is to use lower ilevel items, but doesn't make sense either when you look at Normal Mode gear.

Normal Gear Level:
At the normal gear level there are some non-tier options for four out of the five slots, but none of them are a clearly better choice. In most cases you would be trading one secondary stat for a lesser secondary stat like trading Haste for Mastery. The other issue is that the non-tier items have a lot spirit on them that could be difficult to get rid of and balance around. In short, if you ignored the set bonus in this way, you would be losing the set bonus without any significant stat gain. Obviously this is a DPS loss.

Using 2T12:
The final option would be to use two items from T12 so in hopes that the 2T12 set bonus is better for your DPS then the 4T13 and better stats from a higher ilevel. Anyone who has played this game for a while knows this isn't unheard of. In ToC during WotLK all moonkin continued to use the T8 items because the 4T8 set bonus was extremely over powered. A more recent example is that before patch 4.2 was released may moonkin were going to continue to use their T11 gear to keep the 4T11 set bonus, until Blizzard announced a nerf to the set bonus.

The question though, is 2T12 that good? The answer is, that it's not.

Using a 397 ilevel gear set I ran two simulations, swapping out the chest and legs of the Normal level Deep Earth moonkin gear for the chest and legs of the heroic level Obsidian Arborweave. The simulation with a full 4T13 set bonus had an average DPS of 37,697. The simulation with 2T12/2T13 set bonuses had an average DPS of 37,661.

While it is clear that the 2T12 set bonus is better then the 4T13 set bonus, it is not so much better that we can ignore the ilevel difference and continue to use the 2T12 set bonus without a DPS loss. Since the two numbers are so close it's not a big loss, but given that there is a clear advantage to upgrading to the heroic level tier gear it doesn't make sense to stick with the T12 level items.

TL:DR

The live version of the 4T13 set bonus is one of the weakest I've seen in my years of playing a moonkin. However, due to a lack of alternatives, there is no reason we should ignore it, especially considering the buff it has received on the latest PTR build.