Thursday, October 29, 2009

Blue Post: Need Before Greed and Other Things

This slipped my radar at first, but thanks to Lissanna's excellent post on Restokin.com it came to my attention.

The Blue Posts:

  1. The Need Before Greed loot system will be the unalterable default looting system for pick-up groups in the Dungeon System and has been updated.

  2. Need Before Greed will now recognize gear appropriate for a class in three ways: the class must be able to equip the item, pure melee will be unable to roll on spell power items, and classes are limited to their dominant armor type (ex. paladins for plate). All items will still be available via Greed rolls as well as the new Disenchant option should no member be able to use the item.

  3. Group Disenchanting Option: In addition to rolling “Need” or “Greed” on items, players now have the option to elect for an item to be disenchanted. “Disenchant” works exactly like “Greed” except if a player wins the “Greed” roll, they will receive the disenchanted materials instead. Players who choose “Need” will always win the item and will always beat those that choose “Greed” or “Disenchant.”
A lot has already been said about these changes on the forums, and I want to try and cut through the hysteria.

The Actual Impact:

In a literal sense this change means that Moonkin will not be able to "Need" cloth when using the random group feature coming out in 3.3. A moonkin can "Greed" on cloth using the tool, but their rolls are grouped with everyone that just wants the item for enchanting mats or vendor gold.

In actual play this change will probably have very little impact. Who is running instances for gear any more? Most of the people using this too will be doing so to get emblems, get rep, or get achievements. The only really problem I have with the actual system is that it treats "Greed" and "DE" on the same level. Therefore an item that is a real upgrade for a new moonkin may get automatically DEed so that another player can have a shard in the bank. Other then that, the impact will only be felt by people who are leveling.

The Symbolic Impact:

The reason I have a problem with this change is the symbolic impact. Whether Blizzard means to or not, this change will be interpreted as a big flashing sign that says: "MOONKIN SHOULDN'T WEAR CLOTH!!!"

I am trying to look at it optimistically. First, this will probably have little or no impact right away. Thoughtful guilds like mine will still realize that there are valid reasons for Moonkin to wear cloth in some situations. Thoughtful Moonkin will pick their battles wisely and try not to rock the boat to much. Guilds that currently prevent moonkin from rolling on cloth will continue to do so. The status quo will be maintained.

Another reason to be optimistic is that there are massive changed coming to gear itemization in the next expansion. Moonkin shouldn't have to worry about Spirit any more and hopefully Blizzard will be better about providing gear with the necessary stats like Hit Rating. I may be dreaming but we may actually get to a place where spell leather is equal to cloth. If that happens then the leather vs cloth debate will change quite a bit.

That said, the pessimist in me is worried. Blizzard has gotten a lot better about providing quality spell leather, but there are a couple of facts that make me think that spell leather will never equal cloth.
  1. If you exclude Ele Shaman, there is one spec that will want to use DPS spell leather. There are 3 classes that are restricted to cloth.


  2. Most raids will run with only one Moonkin. Two at the most. On the flip side, most raids will have 6 or more cloth DPSers.


  3. In WotLK, cloth drops are twice as frequent as spell leather drops.

Blizzard has said many times that they don't like seeing items useful to only one class getting sharded week after week when there are possible upgrades for other classes. This is why they don't like to have idols drop from bosses. This means that there will always be better cloth options because there will always be more cloth options. If you only have one non-tier leather piece with Hit Rating you might be forced to use it, because you don't have other options. Moonkin will be back to the same cookie cutter gear we had in TBC.

A Failure of Logic:

Blizzard has acknowledged that this change was likely to be controversial, but they feel it is necessary to protect people against ninja's because it will be harder for players to protect themselves with cross-realm LFG. I understand the reasoning, but I question its logic. Blizzard explained it this way:

Ultimately, our logic went like this: If a Holy paladin loses a great non-plate upgrade to another player, one out of five players might feel bad. If a Holy paladin rolls Need on every type of armor that drops, then four out of five players might have a bad experience. The jerk potential seemed worse than the lost-an-upgrade potential as far as whether or not players buy into using the new dungeon tool.
On the surface the logic seems reasonable. Upsetting 1 person is better then upsetting 4, but I want to ask one question. How big of a problem are Jerks/Ninjas?

I've played WoW for a long time and I've done a lot of pugs. I've met my fair share of assholes, but I would say the vast majority players are reasonable and aren't going to "Need" every single drop.

For arguments sake, lets say that in 1 out of 10 instantance runs there is a jerk that decides ninja loot. Then lets say that in 5 out of 10 instance runs there is a Druid, Shaman or Pally that loses a legitimate upgrade, because they couldn't roll need or another player wanted a shard. Who knows what the actual rate for these to events will be, but I think everyone can see how the 1 vs 4 argument is misleading.

To use Blizzard's example, only the Holy Pally will be upset that he can't roll on the non-plate item, but he may be upset every time he runs that instance because there will usually be someone that hits the DE button without checking to see if it's an upgrade for someone that can't roll need.

I think we would all agree that ninja's suck, but someone getting screwed out of a legitimate upgrade is going to be much more common.

Other Things:

Nature's Grace:
I agree Nature's Grace is a big part of the problem here. We changed Gift of the Earth Mother for pretty similar reasons. However, Nature's Grace is a Really Big Deal. We honestly keep talking about changing it, but we're also paranoid about screwing it up because Balance dps is dependent on it. NG is probably an overbudget talent, but in this case if we nerfed it, we'd have to mess with Balance in a lot of other areas to compensate. We'll almost certainly do it at some point, but I worry about messing with it for 3.3.
Translation: "Nature's Grace is getting nerfed in 4.0."

After reading all of the discussion, I agree with Ghostcrawler's comment quite a bit. If Nature's Grace wasn't so powerful we wouldn't be clipping the GCD, but it's such a powerful talent that nerfing it is more difficult then it sounds. Given that we are so close to an expansion it is probably a lot easier to alter the talent and the spec as they are making the massive changes that come with every expansion.

Twisted Nether Blogcast:

In case you missed it, my appearance on the Twisted Nether Blogcast round table is now available. You can find it here. I am also told it is available on Itunes.

Wednesday, October 28, 2009

Moonkin Tools: SimCraft

SimulationCraft

What: SimCraft is a simulation based executable program. I think it was orignally created by Dedmonwakeen on the EJ forums and is not a dedicated Moonkin tool. It currently supports most of the classes in game and Starfox is the current Moonkin developer for the project.

Where: The project's home is in Google Code which you can find here. This is where you would find the downloads, source code, and issue logs. Most of the discussion regarding SimCraft can be found on the EJ forums. The general discussion can be found here. The druid specific discussion can be found here.

How to Use: The developers are working on a GUI interface, but it still in alpha at this point. Therefore, I recommend that you stick with using the text files and such. For the record, I am a Windows user. I don't have a clue as to if or how this would work for any other platform. Here are some basic instructions on how I got it to work.

First, down load the files and put them in their own directory on your computer. There are two files you want to look at first. They are the READ_ME_FIRST.txt and the Examples.txt files. These are probably the best sources of documentation you will find right of the bat.

I assume you will want to download your toon from the Armory or WoWhead. To do that you need to create a text file. You can do this using Notepad. To download from the armory you text file needs to have one line that looks like this:

armory=Region,Server,Player save=Player_Version.simc
For graylo I created one that looked like this:

armory=us,garona,graylo save=graylo_base.simc
When you drag the file you created over the SIMC.BAT file it will download your profile and save it to a simcraft file that can be modified for future runs. Then drag the .simc file you just created over the SIMC.BAT to get the DPS calculation. You can also drag that same file over the SCALE_FACTORS.BAT file to get stat weights as well as the DPS information.

There are several variations to this process. For example you could download your information from a WoWhead profile instead of the armory. You can review the Example.txt file to see more of the possibilities.

Pros:

  • Fewer Assumptions: Formulation tools like Rawr and WrathCalcs have to make a lot of assumptions to get results. For example, they calculate what the average Eclipse uptime will be. They calculate average cast times. They also make assumptions like DoTs are always up so that you benefit from Imp IS. These are not bad assumptions, and the tools are not bad for making them, but you can see that they might skew the results a little.

    Simulations make fewer assumptions. They model DPS by setting up a spell priority and and giving the rotation a time limit. DPS is then calculated based upon the the random procs that occur and the actual buffs and Debuffs you receive. If Fairie Fire is not on the target you do not get the 3% crit chance. I think this results in more realistic numbers when it comes to total DPS questions.

  • Alt Friendly: Like Rawr, SimCraft covers most of the classes and specs in the game. If you get comfortable using it on your moonkin, you can transfer your knowledge to try and use it to evaluate your alts as well..
Cons:


  • Statistics: The big problem with simulations is that you will never get the same results twice. Therefore you have to run them multiple times and use averages. Simcraft for example will run you simulation 20,000 times by default to come up with your spec's DPS. When you do the Scale Factors it runs it 20,000 for each stat. Therefore, it runs the simulation about 140,000 times

    That sounds like a lot, but it actually happens pretty quickly. The issue is in the variance. SimCraft is great at measuring large changes like what happens if I change my rotation or which set bonus is better. It is bad at measuring small changes like what happens if I add one more Spell Power or Crit Rating. As a result, to provide the stat weights SimCraft has to make large changes in stats. This brings up diminishing returns issues. It's also not good at evaluating minor gear changes.

  • Unuser Friendly: The project has some decent documentation, but lets face it. Creating text files and doing drag and drops will be intimidating for some users. Also, making changes in the simulation is not easy. If you want to try a different rotation you have to figure out what the triggers are and modify the text files. You can change your gear manually, but it's probably easier to download a new profile. The program also assumes you use all available raid buffs. If you want to turn one off, you have to turn them all off and then turn the rest back on. Hopefully, when they get the GUI interface running this will eliminate most of these issues.

  • Black Box: Again, unless you understand C++ SimCraft is a just a Black Box that you put numbers in and have numbers spit back out to you. If you're like me you like to know how it got from point A to point B, but that is not possible unless you understand the language.

    This isn't a huge issue, because a lot of smart people are reviewing these tools, but I like to be able to research an issue when the results seem strange to me. SimCraft does not allow me personally to do that easily.
TL:DR

SimCraft is great because it provides a view of your DPS that the other tools can't provide. It lets random events occur as random events, and impact your DPS as they would in an actual raid situation. This makes it a great tool for evaluating spell rotations, idol procs, and set bonuses. However, due to the very nature of statistics it can't provide exact results that are consistent every time you run the simulation. It is not necessarily the best tool for evaluating small gear or stat changes.

SimCraft also has some very big issues with how users will interact with it. Unless you are computer savvy it may be very intimidating to create the files necessary to use the tool.

Tuesday, October 20, 2009

A Return to the Twisted Nether Blogcast


Fimlys and Nibuca over at the Twisted Nether Blogcast have asked me to participate in a round table discussion this Friday. The primary topic of discussion will be patch 3.3 and how it affects us, but I'm sure the conversation will drift some. The round table will feature myself, Matt from World of Matticus, Stoneybaby from BigHitBox, Pike from Aspect of the Hare, and possibly Honor from Honor's Code.

You may remember that I was featured on TNB back in April as the blogger of the week. I had a lot of fun and it seemed like the people in the chat room enjoyed themselves as well. I'm sure this time will be a lot of fun as well. I just have to remember to let other people talk.

Anyway, the taping will be this Friday 10/23/2009, at 10 pm CST. Click Here, if you would like to find details on how to participate in the live taping. Don't worry, if your unable to attend the taping. They will post the recorded show sometime next week and I will link it when it is available.

Blue Post: Haste and the GCD

In case you missed it Murmurs, started another really good thread regarding Haste and the Global Cooldown. You can find it here.

Since Blizzard is looking at haste a little more closely in patch 3.3 with the possibility of having DoTs affected by haste, Murmurs has taken this opportunity to summarize the Moonkin issues with haste rating.

I'm sure the issue is well known to most of you. Moonkin's have two primary nukes. After Starlight Wrath, Starfire has a base cast time of 3.0 seconds, Wrath has a base cast time of 1.5 seconds, and as always the minimum GCD is 1.0 second. After you account for Celestial Focus, Improved Moonkin Form, Wrath of Air, and Nature's Grace, a moonkin only needs 400 haste rating to cap the affective Wrath cast time at 1 second.

This ends up causing a problem with how we scale. Moonkin are is the only spec that has a nuke with a 1.5 second base cast time and is a core part of the rotation. This causes an inequity when Blizzard tries to balance us against other classes. In WotLK we balance very well in the Tier 7 content, but seemed to have slipped some as we progress to higher tiers. I can't blame this all on haste, Wrath, and the GCD, but it is definately a contributing factor. Since we don't scale as well with Haste rating as Mages, Shaman, and Destro Locks do, it creates and opportunity for us to fall behind in DPS.

That's the issue as I see it and I think this is what Murmur's was trying to point out.

The Blue Post:

Ghostcrawler stopped by the thread and posted these comments.

So in theoretical-space, how much do think these problems go away if haste doesn't bump into the global cooldown for some of these spells?

I know it's cumbersome to explore why that would be the case, because I don't want to stipulate that haste rating just gets cut in half or whatever. I also think it complicates things when you start talking about all the added benefits of spells becoming instant (such as being able to move and not being interruptable).

I'm trying to get at whether it is the way different spells behave under haste or whether it's the GCD that's the problem.
I'll be honest, I'm not completely sure what Ghostcrawler is trying to get at here, and I'm a little confused by his response. I think he is trying to address the solutions that are being suggested. So, I'm going to take that approach from here on out. The two basic suggestions I've see are to lower the GCD or increase the cast time of Wrath.

Lower the GCD:

Currently the minimum GCD you can have is 1 second. Some people are suggesting that this should be lowered even further or possibly eliminated.

Personally, I don't like this possibility for a couple of reasons. Vontre explained it best on post #146 on the thread.

It's the gcd, except the gcd cap is fine where it's at because 1 second is about as low as you want to go for the human brain to process action -> button press -> server command -> return to client -> evaluate result. Problem is Wrath is too fast for a primary nuke.
Lowering the GCD is not a clean solution. When you consider the latency and the human factor, you realize that you can cast only so fast. While we would technically scale better, the practical aspects of the game would limit the impact of the change.

Increase Wrath's Cast Time:

There are multiple ways to do this. Some have suggested increasing Wrath's Cast time using a Glyph. Others have suggested that some of the buffs that reduce Wrath's cast time be changed or eliminated. Let me look at these possibilities separately.

  • Glyph: The suggestion is that blizzard add a glyph that increases Wrath's cast time by 0.5 seconds and the damage by 33%. This would not change the DPS of the spell at zero haste, but would allow it to scale with haste more appropriately. I like this suggestion but I see two issues with it.

    1. The first issue is how this impacts Eclipse. A lot of people complain about how Wrath procs Lunar Eclipse on only 60% of crits but it's done for a practical reason. It makes Solar and Lunar Eclipse proc at a similar rate. If you increase Wrath's cast time without increasing the proc rate of Eclipse you throw off that balance.


    2. The second issue is what type of Glyph would it be? If it's a major glyph then your losing DPS by not having one of your other glyphs. Making it a minor glyph wouldn't have this problem since most of the minor glyphs we currently use have little impact, but it would increase our DPS indirectly and my not qualify as a minor glyph.


  • Lowering Talent Haste: This suggestion is that Blizzard either nerf or eliminate some of our haste providing abilities. If we have lower haste then we will brush up against the GCD less and eliminate the scaling issue.

    The problem with this suggestion is that it is a straight nerf to our DPS. The affect would be moderated by the increased usefulness of haste rating, but in total the affect would be a nerf. Therefore this doesn't fix the problem unless they buff Moonkin in other ways as well.
The Third Option:

The third option is that they fix moonkin scaling in some other way without touching Haste or Wrath's cast time at all. This has been alluded to by several people, but not out right said. Blizzard could fix our scaling issues by making us scale with another stat at a higher rate then most class do.

For instance, what if Crit Rating not only increased your Crit Chance, but also increased the amount of damage your critical strikes did. Any change like this would take a lot of balancing, but my point is that Haste is not the only stat that could be used to make us scale better. The best option maybe to leave Wrath, the GCD and all of our haste the way it is and tweak moonkin DPS in some other way.

In some ways, I think this is the best option, but it's probably the most difficult to implement.

Friday, October 16, 2009

2T8 vs 4T9: A Different Perspective

Looking around the internet I came upon this discussion on the Moonkin Repository. None of the sentiments expressed in the thread are new, but I've grown frustrated with some of the conclusions that have been drawn over time. It seems to me that some people have an unrealistic expectation of what set bonuses should be. Please let me use the 2T8 vs 4T9 debate as an example.

A History of 2T8

The first version of 2T8 was a 6% increase to your Eclipse buffs when the eclipse buffs were mutually exclusive. Math that I and a couple of other people did showed that it was extremely inferior to 4T7 and suggested that Blizzard buff it to 15% to make it competitive. When Blizzard buffed it as we suggested it became a good set bonus that increased Moonkin DPS by about 3% give or take. In patch 3.2 Eclipse was buffed so that the two Eclipse buffs were no longer mutually exclusive, and you could proc one while the other was on cooldown. Not only was this a significant buff to Eclipse, but it was also a significant buff to 2T8. This increased the value of 2T8 so that the set bonus provided about a 6% DPS increase. This is much larger then most other Set Bonuses in the history of the game and many theorycrafters including myself expected 2T8 to be nerfed. It was not.

2T8 vs 4T9: A different perspective

I have seen several comments that said something to the affect of "4T9 sucks. Blizzard is screwing us." I completely disagree.

4T9 is not a bad set bonus. My calculations value it at about a 2.5% DPS increase. Looking at prior moonkin set bonuses and the set bonuses of other classes, set bonuses tend to be worth somewhere between 2% and 4%. 4T9 is on the lower end of that scale, but a 2.5% DPS increase is not something to brush off. It's much more significant than most people think.

2T8, on the other hand, is a bad set bonus. Try and understand what I'm saying before you flame me. 2T8 is obviously a very strong set bonus, but it is an outlier. The reason we don't move from 2T8 to 4T9 isn't because 4T9 is weak, but because 2T8 is strong.

The reality of the situation is that Blizzard is not going to increase the value of set bonuses with each tier of content. It would be an unsustainable cycle. If they did that, at some point only the set bonuses would matter. The trick is to try and maintain the value of set bonuses from tier to tier to make for an easy transition, without making the other factors of our DPS meaningless.

Therefore, every time you say "4T9 is weak" or "buff 4T9", what blizzard probably hears is "2T8 is too strong" or "Nerf 2T8."

Myth: 2T8 was made to Make Moonkin Competitive

It seems to be a popular opinion that Blizzard purposely used 2T8 to make Moonkin competitive with other classes in terms of DPS. Personally I think the suggestion is ludicrous. Why would Blizzard balance any class using a temporary set bonus? It doesn't make sense. To maintain the buff they would have to continue giving that class over powered set bonuses. This would in turn create expectations of what a set bonus should be worth, and the other classes would get up set that one classes set bonuses were always significantly better then their own.

If Blizzard feels the need to balance a class they are going to do so with more permanent changes like having Spirit provide Spell Power, or changing Eclipse so that the buffs are not mutually exclusive.

TL:DR

Like everyone, I would love it if 4T9 was better. I would love it if all moonkin bonuses and abilities were better. I would love it if Moonkin were OP for once in their lives, but that is not good game design. No one likes to hear it but 4T9 is probably closer to what a set bonuses should be then 2T8. If you keep asking for a change to be made, don't be surprised when that change happens on the opposite side of the equation.

Wednesday, October 14, 2009

Patch 3.3: Relics

This will be a short post, but after looking at the latest info MMO-Champion there are a couple of things I want to comment on.

T10 Balance Druid Relic - The periodic damage from your Insect Swarm and Moonfire spells grants 44 critical strike rating for 15 sec. Stacks up to 5 times.

At first glance I was pretty unimpressed due to it's similarity to the T9 idol. Lets take a quick look at it for comparison.

[Idol of Lunar Fury] - Each time your Moonfire spell deals periodic damage, you have a chance to gain 200 critical strike rating for 12 sec.

The two idols buff your crit chance by a similar amount and both are based upon a DoT you put up. So, my first reaction was "Don't I already have this idol?" My second reaction a little more honest. While these two idols are very similar it is important to note the differences.

The first difference is obvious. The T9 idol grants 200 Crit Rating. The T10 idol grants 220 Crit Rating when fully stacked. If you assume a 100% uptime of both idols then this is a 10% buff to the idol slot. That's not that bad of a buff going from one tier to the next.

The second difference is something a lot of people will ignore.The T10 idol will have a higher buff uptime. The T9 idol buff has a 70%-90% proc chance and a 12 second duration. It's so high that most players and theorycrafters assume that it has 100% uptime. That's not a bad assumption, but in real life the idol's uptime will be less.

Since the buff is proc based there is always the possibility that bad luck will strike and it doesn't proc for an extended period of time. Also, there are times when casting DoTs is not in your best interest. At these times your giving the buff an opportunity to slip off.

The T10 idol on other hand is not proc based and has a 15 second duration. If you cast both DoTs the buff should be fully stacked within 8 seconds. Since Moonfire has at least a 15 second duration, as long as your cast a DoT every 27 seconds the stack will be maintained.

TL:DR - The T10 idol is not sexy and it isn't something your going to see a bunch of moonkin going nuts over. However, the buff is a little bit bigger then the T9 idol and it will be easier to maintain the buff. Therefore, this is a clear upgrade, despite my first reaction.

Rebirth: The cooldown on this spell has been lowered from 20 minutes down to 10 minutes. Cannot be used in Arenas.

Great chance. If you've read my blog for along time you may remember that I thought that the Rebirth cooldown should have been reduced to 15minutes instead of 20 back in the early days of BC. Reducing it to 10 minutes is even better then my suggestion.

The big problem with the 20 minute cooldown is that if you use it in one attempt you don't have it for the next 2 or 3. Hopefully, with this change it will now be up for almost every attempt. This does make me wonder about soul stone though. They should reduce it's cooldown also.

UI Change: Any party member may mark raid targets.

Oh god, please don't let this go through. I can see the advantages for this change. The leader of the party is not always the best person to mark targets. This is especially true in 5mans when you can't make someone an assistant. I see the issue, but I also see the headache this could cause.

First, Addons like DBM and Big Wiggs like to use the icons to mark people that get buffs or debuffs. Most of the time that is fine and it doesn't matter that 25 different mods are trying to mark those people. However, there have been times in my raids were those marks were distracting and we want to turn them off. If you've ever tried to get 25 people to PvP flag in an instance, I think you'll understand how hard it will be to get 25 people to turn off a part of DBM so that the targets aren't marked.

Second, you always have to worry about people getting silly and changing the marks as a joke or on accident. There are some responsibilities that should be left up to the leader of a group. I think this is one of them.

Tuesday, October 13, 2009

Moonkin Tools: Rawr

Rawr

What: Rawr is a formulation based executable application that models several classes and rolls. The project was originally started by Astrylian as a Feral Druid tool, but many developers have been added to expand the tool to other classes. Dopefish is the current Moonkin developer on the project.

Where: You can download the tool from Rawr's website here. They also maintain a thread on the Elitist Jerks forums to document changes and handle questions. You can find it here.

How to Use: I'm not going to go in depth here for two reasons. First, Rawr is a little more complicated then can be easily explained here quickly. Second, they already have very good documentation on their web. The two main things you will want to look at are the Rawr Video and the Rawr Help section. If you have additional questions I suggest you go to the Discussions site. You can search the topics. If you don't find your answer, you can post the question yourself.

Pros:
  • Documentation: A tool is useless if potential users can't figure out how to use it. I do think that the current documentation can be expanded and improved, but what they have is great. At the very least it is much better then the documentation for WrathCalcs and Simcraft. I posted several links in the How to Use section. What I like most is the discussions site. It's nice to be able to ask a question if you don't understand something or search the questions that other people have already asked.


  • User Friendly: Rawr is the most user friendly tool I've seen out there. Since it has a GUI interface, even people that are not that computer savvy should be able to use it effectively. On top of that, it allows you to download your profile from the Armory to get started quickly.

  • Functionality: Rawr has some neat little functions that you won't find on other tools. The Optimizer is a particularly cool feature. If you select the gear pieces you have available to you, clicking the Optimizer will tell you what is the best set of gear possible. You can also swap pieces of gear in and out to see how they will impact your DPS. You can also change up your gems and enchants to see what affect that will have.

  • Alt-Friendly: If your like me and have a couple of alts one of the nice things about Rawr is that you can use it to evaluate several classes and specs. The tool doesn't cover all specs, and I don't know the quality of the other modules, but Rawr is a good place to start if you want to evaluate your Alt without downloading another spreadsheet.
Cons:
  • Black Box: Unless you understand C# Rawr is a just a Black Box that you put numbers in and have numbers spit back out to you. Since I am not a programmer I don't understand the guts of the program, and I don't like the fact that I don't know how it got from point A to point B.

    The reason I don't like it is because every tool makes assumptions and has limitations. If you don't understand those assumptions and limitations it is impossible to account for them in your choices. I know this won't be a big deal for some users, but I like to know whats going on behind the curtain.

  • Calculation Questions: I am not saying they are wrong, and according to Dopefish, with the exception of trinkets, the calcs in Rawr are the same as the calcs in WrathCalcs. However, I have experienced unusual results when using Rawr. For example, this was a couple of upgrades ago, but when I first put my profile into Rawr it recommended a MF/SF rotation with no IS.

    Obviously this goes against the conventional wisdom and raises some questions about the tool. However, Dopefish says that Rawr.Moonkin has been reviewed and fact-checked by third party reviewers. At this time they are unable to find what is causing the differences between the tools.


  • Out of Date UI: The big question I had when I first looked at Rawr was if it was updated for the 3.2 version of Eclipse? The current UI doesn't give a good indication that it is, and the current UI shows DPS results by Rotation. However, Dopefish says the current version of Eclipse is fully integrated into the tool.
TL:DR

Rawr is a very nice tool for evaluating upgrades and the different gear options. The documentation provided and the User Friendliness of the tool sets it apart from the other options.

That said, I do have some small concerns about the calculations. I think they are probably directionally correct and can accurately predict most gear upgrades. However, some of the results provided are confusing.

Thursday, October 8, 2009

The Scotty Principle

I am not a big Star Trek fan but there is one episode of The Next Generation that has always stuck out for me. It is the episode called Relics where the crew finds an old Federation ship wrecked and abandoned. When they examine the ship they find that the teleporters are still active and holding an aged Scotty from the original Star Trek in stasis.

To be honest, I had to look most of that up on Wikipedia to remember what happened, but there is one scene in the episode that has always made me laugh. As usual there is sort of problem that needs solving and Geordi La Forge has an idea of how to fix it, and tells the captain it will take 2 or 3 hours. When Geordi and Scotty get back to Engineering, Scotty asks Geordi, "How long will it really take you?" Geordi says, "2 or 3 hours."

Scotty's response is shock and he says "You told him how long it would actually take you? How do you expect to be considered a Miracle worker if you tell them how long it will really take."

The moral of this story is that, if you lower expectations, it's much easier to please your audience.

This may seem like a random story, but I wonder if Blizzard is trying to use this process to manage their player's reactions. Lets face it, the only way they could make a majority of the player base happy in one shot would be to make bonuses and buffs that are so clearly awesome that everyone loves them immediately. That may sound nice to players but it is a slippery slope. Players will always expect the new bonus to be better than the last one. If you give into that desire then you run into issues with stat inflation.

So, how do you keep the players happy, prevent inflation, and still provide interesting buffs that provide a decent buff? You lower expectations.

With a set bonus you give them something you know is worse than what you intend. That way when they complain you buff it to what you really intended. The players still may not be completely happy with it but it's better than what you originally showed them, and you give off the illusion of taking player concern into account.

I'm not saying Blizzard does this, but it makes me wonder when I see a set bonus so clearly bad like the original version of 4T10 and then a quick buff within 24 hours.

The New Set Bonuses:

2 Piece Tier 10: When you gain Clearcasting from your Omen of Clarity talent, you deal 15% additional Nature and Arcane damage for 6 seconds. (Changed from 10%)

There were a lot of different valuations thrown around yesterday regarding this set bonus. Most of them valued it as about a 2%-2.5% dps increase. I think those estimates are a little optimistic, but let's assume they are in the ball park. A 50% increase to the buff would make the estimated DPS increase 3%-3.75%.

I still think these estimations are a little high, but they are definitely in the ball park of what I think a set bonus should provide. I still don't like the fact that it is another random proc that we must rely on for DPS, but that doesn't look like it is going to change. In short, I am happy with this buff at the moment.

4 Piece Tier 10: Your critical strikes from Starfire and Wrath cause the target languish for an additional 5% of your spell's damage over 4 seconds.

This is much better than the original set bonus, but there are some questions about how it will impact our DPS.

How will the extra damage be applied and how often? We know that is applied over 4 seconds, but is it every second? Every 2 seconds? This is important because we cast spells faster than every 4 seconds. We can cast Wrath almost every second, and it is quite possible to have two crits in a row. Does the second crit over write the first? What if the first crit was a Starfire crit with more damage than the second crit which was from Wrath?

These are very important questions when trying to determine the value of this buff. Especially when it comes to Lunar Eclipse where some people will have close to 85% - 90% crit chance. Will this set bonus be useless in that situation?

I have done some napkin math to try and put a value on it. Assuming that the dots don't overwrite each other, a Chaotic Skyflare Diamond is equipped, and the caster has a very high crit chance around 45%-55%, this buff would increase the damage of Starfire and Wrath between 3% and 3.5%. Assuming that these two nukes make up about 80% of our DPS , this set bonus would increase Moonkin DPS by 2.5%-3%.

There are some other considerations that I've excluded to make the math easier, but I think this is a good estimate of the value, and probably a little optimistic. If my napkin valuation is correct then it is a little lower than I would like. I thing the set bonus should be worth between 3.5%-4%, but that is just my opinion. One thing, that can't be argued is that it is a whole lot better than the original 4T10, and that may have been Blizzard's goal all along.

(Edited by Flamedor)

Patch 3.3: Tier 10 Set Bonuses

/sigh... Here we go again.

If you don't know what I'm talking about, MMO-Champion has pulled apart the latest PTR build and found the Tier 10 set bonuses. You can find them here.

One of the themes for Moonkin in WotLK has been to fight against sub par Set Bonuses. The tier 7 bonuses were satisfactory for the most part, and everyone loves 2T9. However, it has been a bumpy road other than that. The 2 piece T8 use to buff Eclipse by 6% instead of the 15% it is now. The 4T8 set bonus was bugged and we had to fight to get it fixed. The 4T9 set bonus bordered on insultingly bad as just a 3% increase to crit chance for Wrath and Starfire, but was upgraded to a 4% damage buff for both spells. Given this history, I guess it is fitting that we have issues with the Tier 10 set bonus. The good news is that Blizzard has reviewed our bonuses in the past and made changes. Hopefully, Tier 10 won't be different.

One last thing. It is important to remember that data mined information is data mined. While much of it is usually correct, there are plenty of examples where the data mined info included out of date info or info that is just plain wrong. It is quite possible that Blizzard has something else planned but is not currently included in the patch build.

4 Piece Tier 10: Reduces the cooldown on your Eclipse talent by 6000.

I'm talking about this one first because it is the most troubling. Obviously, we can't take the number 6000 seriously because you can't reduce Eclipse by 6000. That said, it is still an obviously bad set bonus. The set bonus could be that Eclipse has no cooldown and it would still be obviously bad.

To highlight the issue I have to ask only one question. Since 3.2, when has the Eclipse cooldown ever been a limiting issue?

Sure, with the use of WiseEclipse the Eclipse cooldown can be slighly limiting, and it would probably be benifical to cancel Solar Eclipse to reproc Lunar Eclipse during Bloodlust/Heroism. However, those are either rare or very minor situations.

The reason this is a bad set bonus is obvious once you consider some facts.

  1. The Eclipse buff lasts 15 seconds. The Eclipse cooldown is 30 seconds. Even if you chain proc the Eclipse buffs, the two buffs combined would eat up the 30 second cooldown. This isn't entirely true if you use an addon like WiseEclipse, but I have a hard time believing that Blizzard would build a set bonus around an addon that most users think is a little shady.

  2. Math that I and many others have done shows that the two Eclipse buffs are very similar in value a majority of the time. Therefore there is no real value in using one over the other in most situation.

If there is no reason for you to force one Eclipse to proc over the other, and there is no way you would proc Eclipse more quickly naturally, then it is obviously not a buff.

I haven't done any math on it but the best way to modifiy this buff in my opinion is to have it extend the Eclipse buff rather than reduce the cooldown. I'm pulling numbers off the top of my head, but if it extended Eclipse's length to 20 or 25 seconds instead of 15, then it would increase the uptime of the buff and it would work better with other buffs like Heroism and on use trinkets.

2 Piece Tier 10: When you gain Clearcasting from your Omen of Clarity talent, you deal 10% additional Nature and Arcane damage for 6 sec.

To be honest, I'm not 100% sure how Omen of Clarity works. Historically, it has had a speed based proc chance that may or may not be impacted by haste. The proc rate is reportedly based around 2 procs per minute. WoWWiki has a report that says it has about a 6% proc rate. To find out the real impact I will have to look at some logs but let's assume for a second that it procs 2 times per minute. Let's also assume that the procs don't over lap.

If these two assumptions are true then the buff would be up for 12 seconds per minute or 20% of the time. Therefore, if fully utilized then this set bonus would buff your Arcane and Nature Damage by 2% on average (10%*20%). In actuality it will be less than that, because part of the buff will not be effective since damage is calculated at the end of the cast .

Another big question is how this buff would be applied? Is it additive or multiplicative?

Historically, buffs that affect specific spells, like Moonfury, are additive. Global buffs, like Earth and Moon, are multiplicative. It's hard to tell which one this buff will be since it specifies schools of damage, but it covers everything we could cast.

It's important to ask, because it will have an impact on the value of the buff. A multiplicative buff of 10% is a 10% buff to damage. An additive buff of 10% will be worth less due to the way it combines with other buffs. Improved Moonfire is a great example of where a 10% buff can get completely lost and overshadowed by a larger buff.

I can't make a firm statement on this set bonus yet, because I don't understand all of the parts completely. However, I will say this. It does not give me a warm fuzzy feeling. I don't like the fact that it is based upon a talent that we don't necessarily need since there are some moonkin dropping OoC. I don't like the fact that it is another random proc for a spec that is already swimming in random procs. I don't like that the buff is only 6 seconds and could easily get swallowed up by movement or a delay in fight mechanics. It could be great, but it doesn't feel right.

Tier 10 Art:

I don't usually comment on the art of the game, but take a look at this picture.


I haven't decided if I like it, but it is definitely interesting. It's kind of like a demonic tree. This image was taken from MMO-Champion.

(Edited by Flamedor)

Friday, October 2, 2009

Patch 3.3 & Firefighter

Patch 3.3:

As you probably know Patch 3.3 has hit the PTR, and we have a new set of patch notes and undocumented changes to look over. It's not surprising to me that the patch notes are fairly brief. Reportedly Patch 3.3 is the final chapter in Wrath of the Lich King and I would hope that they have the classes pretty much where they want them by now. That said there are a couple of things to discuss.

Icecrown Citadel - Obviously the new raid instance and 5man dungeon are the primary features of the new patch. Not much is known about the raid, but Blizzard has released some information regarding the 5man. You can find it here.

It sounds pretty cool from the limited information provided, but I do have one question. I really want to know how the fight against Arthas will play out at the end of Halls of Reflection. Obviously we can't kill him because that will be done in the raid, but I hope they have a better explanation than, "HoR was just a set back," like they did with Kael.

That said, I don't have a problem with them recycling bosses in this way as long as it fits the story, and provides a play opportunity that wasn't widely available before. To illustrate the difference, lets look at two examples.

Kael'thas is an example of good recycling. The Tempest Keep version of Kael is a very fun and difficult fight that few players saw let alone complete. Making a 5 man version of the fight made sense from a lore perspective because you made a strong boss weaker and therefore vulnerable to a 5man group. Plus, it allowed the majority of players that couldn't see him in the raid to get a taste of one of the best TBC fights. This one made sense from a lot of angles.

Anub'arak is an example of bad recycling. They took a heroic boss that a majority of raiders had already seen, scaled him up a bit and made him a raid boss with little explanation. This move didn't expose players to content they hadn't seen before because Anub'arak was already widely available, and from a lore perspective it feels like Blizzard thought, "We need someone to end the raid with. How about Anub?"

Hopefully the two versions of Arthas will be more like Kael and less like Anub.

Pet Avoidance - Now reduces the damage your pets take from area-of-effect damage by 90%, but no longer applies to area-of-effect damage caused by other players. - Sounds like a PvE buff and PvP nerf. It makes sense to me. Raid Boss AoE hits a lot harder than player AoE. You want the pets to survive in raid, but to be killable by players. Its not that big of a deal but nice for us.

Nature's Grace - now procs from non-periodic spell criticals. (Old - All spell criticals) - I'm a little confused at what this means. When I first read it I thought it was a tool tip change so Blizzard won't get a lot of complaints about Moonfire DoT crits not procing Nature's Grace. Allison Roberts over at WoW.com thinks it will also prevent Hurricane and Starfall from procing Nature's Grace. (source MMO-Champion)

We will have to test it on the PTR, but I still think my original interpretation was correct. I just don't see why Hurricane and Starfall procing Nature's Grace would be a big deal. However, if Allison is correct its not a big deal. Nature's Grace will still be up 75-90% of the time.

EDIT: Erdluf, and the EJ forums make some good points I had forgotten. Hurricane doesn't proc Nature's Grace currently. The damage from Starfall has two pieces. The splash damage is AoE damage and does not currently proc NG. The damage from the stars is direct damage and should still be able to proc NG. Of course it will need to be tested to be 100% sure, but this is very likely just a tooltip change.

Genesis - now increases the damage and healing done by your periodic spell damage and healing effects. (Old - All periodic damage) - This really is a just tooltip change to clarify that this talent won't buff Feral DPS.

FireFighter:


Mimiron hard mode is done.

I don't normally report on individual boss kills anymore, but this is one of the special ones that is likely to stand out for me. I love the fight but LoE's attempts on him were very frustrating. First off, we lost a lot of our core raiders right as we started doing him. This forced us to take a bit of a step back and not put as much time on him as I would like. It even lead to a misunderstanding that caused me to cuss out one of my guild leaders (sorry Sev). After we finally started to put some serious work into it our attempts always seemed to be one step forward and 2 steps back. Countless times we would have 5-10% enrage wipes and then not be able to get to phase 4 for 3 or 4 attempts in a row. It was very frustrating to constantly see such progress followed by such failure.

It really all goes back to my post about the Illusion of Success. Though we were getting to phase 4 every 4 or 5 attempts we were still having big issues with phase 3 and killing it fast enough. It took all of those failures to figure out how best to coordinate when and were to drop the head. In the end we killed him with only 30 minutes left in our raid, and it gave me that warm tingly feeling that only a few bosses can give.

Great job guys.

(Edited by Flamedor)

Thursday, October 1, 2009

My Goals: 10/01/2009

Lets take a quick look at where we are. Wrath of the Lich King was released a little over 11 months ago. The first two big content patches have been released. Cataclysm, the next expansion, was announced a little over a month ago. The third, and reportedly final, content patch is likely to hit the PTR very soon, and Blizzard has said several times they would like to release expansions more quickly then they have in the past.

Patch 3.1 was released 5 months after release. Patch 3.2 was released 4 months after patch 3.1. At this pace we could see the expansion in early to mid summer 2010 or late summer to early fall if you give it a little padding. Of course, these are just slightly educated guesses, and could be very far from the truth. However, I think it is reasonable to assume that WotLK is at least in its midlife and could be well on its way to retirement. With that in mind I would like to lay out some goals for the time we have left in WotLK.

Raiding:
These are pretty easy to guess.

Goal: Kill everything on the on the hardest mode possible before a 3.0 style of nerf. I really like that I killed everything in TBC before WotLK was released, but a good portion of the kills were not progression kills, and we didn't get anything down in Sunwell before the big nerf. I would like to improve upon that in WotLK.
Probability: Very Good. Times have been a little rough lately, but the guild was very stable from when I joined last year in September to May of this year. If we can repeat that stability again then we should kill everything. If not, it will be difficult.

Goal: Get all e-peen enhancers. This goes right along with "kill everything," but I want to emphasise it. Basically, I want all of the 310% drakes and all of the titles that mean anything. I would love to get the server first titles, but I know that's not going to happen. Everything else though is up for grabs.
Probability: Good. My biggest worry here is the Ulduar Meta. We are very close to killing Mimiron hard mode. Vezax and Crazy Cat Lady should go down fairly easily once we focus on them. The big obstacles are Yogg+1 and time. We have at least 30 days to get this all done, and I think we can do it in that time if we focus on it, but you never know. Luckily we probably have a little more time then that. My new guess is that they will remove toe drakes with patch 3.3.

Alts:
My Warlock and my Priest are already level 80. I am doing some raiding with the Warlock, but I don't have any big hopes for either of them. I wouldn't mind completing all of the phasing for them but that is unlikely to happen.

Goal: Level Grayvik the Mage to 80.I know I can buy the herbs of the AH with out much problem, but I don't like not having gatherers incase I need to farm a little bit. My mage is entirely a gatherer with Herbilism and Skinning as his two professions. This may not seem like a big deal, but I like the peace of mind and it will support some of my other goals.
Probability: Almost guaranteed. he's level 72 at the moment. Its really just a matter of time.

Goal: Level Graybel the Shaman to Northrend.I would love to get him to 80, but that is not my primary concern. He is a skinner/leatherworker, and I want to get him to a point where he can max out his professions.
Probability: 50/50. He is currently level 23 and has a long way to go. In the past it has taken me about 8 or 9 months to get to max level. Grayvik is taking a little longer, but that is understandable given the expansion and the time I took to level my 3 level 80s. I think I can do it. I just depends on when the expansion hits and how well my other goals go.

Goal: Level Graydeth the Death Knight to Northrend.You haven't heard about this guy before. I started him to see the Death Knight starting area, I haven't played him much. However, I think he would be a good choice to level up a Blacksmith. A this point I would level him up purely for the professions.
Probability: Doubtful. Again, time is the big issue. This goal is pretty far down the list and I doubt I will get to it by the time the next expansion is released if I'm honest with myself.

Professions:
On my three 80's I have Alchemy, Inscriptions, Tailoring, Enchanting, Jewelcrafting, and Mining maxed out or within a few points of being maxed out. Skinning and Herbilism aren't far behind on my mage.

Goal: Max out Fishing.I hate fishing. I don't like it, but there are some advantages to being able to fish. I would like to have it maxed out to take advantage of those benefits just in case I need them.
Probability: 50/50. Thank god for the changes they made to Fishing during WotLK. At the end of TBC I was about 50 points away from Max skill, and I couldn't bring myself to level it. Now, I knock out a little at a time doing fishing dailys or just to kill some time in Dalaran. I think I will get there but who knows.

Goal: Max out Leatherworking on Graybel.I would be nice to have a leather worker to back up Graylo since he can wear leather. Plus, there are other opportunities I would like to be able to take advantage of.
Probability: 50/50. If I get him to Northrend, he will max out leather working. I've used the skins farmed by my mage to level his leather working and his skill is well beyond his current level. The only thing currently holding back is skill is his own level.

Goal: Max out Blacksmithing on Graydeth.I want this purely to make gold.
Probability: Doubtful. I think if I level him it won't be hard to max out his professions, but I doubt I have time to get him there.

Goblinish Goals:
As you can probably tell, a lot of my goals have been leading up to this one.

Goal: Have 100k gold when the expansion is released.If you want to make a lot of gold in WoW you have to find the imperfections in the market and beat them to death. Since everything in WoW is available to every player, these imperfections are usually the result of a lack of knowledge. Consumers don't realize that they could buy the herbs for that glyph for just a couple of gold and buy the glyph for 20g instead. Producers don't realize they can make 1000% profit on an item so they don't enter the market. So the people who see the opportunities make a killing. However, even if you see these opportunities you need the capital to take advantage of them. That is why I want 100k gold.
Probability: Iffy. You may have read the report about the MMO-Champion guy making 125k gold in 7 weeks. That makes my goal sound easy, but lets face the facts for a second.

  • First of all he makes his living off of WoW. He is probably very knowledgeable about the different tricks, and can have a computer open next to him at all times to take advantage of the opportunities that present themselves. I don't know how he did it for sure, but I would be surprised if he wasn't camping the AH to some extent. Most people just don't have that kind of time.
  • Second, not all servers are the same. Some are very undeveloped, with few competitors. In such a situation your able to sell your goods at a fairly high price for a lot of profit. Other servers will have a tighter market. Take mine for example. According the the MMO-Champion guides you should sell your glyphs for a minimum of 4 or 5 gold, buy eternal earth and eternal shadow for around 10g or less, and buy green gems for around a gold apiece. On my server, 4 or 5 gold is often the market price for Glyphs, good luck trying to sell them for more. The market price for eternal earth and shadow are around 15g and rarely goes lower then 10. Green Quality Gems go for 2g or more.

Over time, imperfections will be worked out of the market as more people discover the opportunity, thus the market will be come less profitable. My point is that my server is pretty far along that curve. I can easily see several sellers that I am competing with and that getting to 100k is going to take a while if it happens at all.