Thursday, November 26, 2009

Moonkin Icecrown Itemization

Ever since we saw the moonkin t10 stats, I've been following MMO with interest to see what's been data-mined.  It finally seemed like this is the tier for solid moonkin gear - even the emblem gear includes two SP/haste/crit leather items, a nicely itemized belt and decent gloves.

As more and more items have been found, however, I've been getting more and more uneasy.  The items I was expecting just weren't showing up - there's only one moonkin leather item so far, and it's 10-man.  Everything else leather has spirit, without exception.

Now this post is obviously a bit premature, because the loot lists we have (I'm going off of is of course incomplete.  And there's always the "data-mined info is data-mined" maxim, and there is some truth to that.  But I find it worrysome that there are zero 25-man leather items for moonkins found so far, in the final tier of the expansion.  And by the time we have complete loot lists, it will be (practically) too late to raise these concerns.

This was true in ToC/ToGC as well - we generally ended up with leather boots/belt, and they had spirit.  But that's not a significant issue because the cloth gear was identical or worse - there was no SP/haste/crit cloth belt or boots, for example.  And we had the Anub legs, which are very nice but also are only one slot of gear.  However, Icecrown cloth doesn't suffer from the same drought moonkins are - it looks like there will be nicely itemized pieces available for almost every cloth slot, and both below pieces are so far almost certainly BIS for moonkin:

I guess the current loot just leaves me with questions and concerns, especially if you consider this interview and the following excerpt:
Greg: The nice thing about Icecrown Citadel from an itemization perspective is that we have lots of bosses. That's something of a relief coming from the Crusaders' Coliseum, where we had only five bosses. For Icecrown, we can afford to have multiple options for a particular slot -- say a healing leather bracer or a melee trinket -- and even itemize some of the more unusual pieces.
So my issue becomes - how did we get a moonkin-specific leather item in ToC (5 bosses) but miss out in Icecrown with 12?  Why is it ok to have multiple healing leather bracers but nothing for a moonkin - especially if you consider that many resto druids would be able to use a "moonkin" bracer and put it to good use?

So, what do we do?
It seems to me that the best solution is simple and pretty straight-forward.  There's a lot of leather spirit gear on the Icecrown tables, and a lot of it has crit but no haste.  For resto druids with their new haste cap, this makes it feel rather subpar especially with how poor crit is for Trees right now.

It seems that some of this gear (and it doesn't have to be a lot - 2-3 pieces, tops) could easily be turned into moonkin leather and solve two issues at once.  Resto druids can afford to use some items without spirit - if anything it will make their mana use more interesting - and those items would help them toward their haste cap.  And this would give moonkins leather alternatives, instead of forcing us to go for cloth.

Monday, November 23, 2009

Icecrown Citadel Raid Progression

Blizzard released their plans for ICC progression last week and the implications are pretty interesting - I wanted to give myself some time to think about it though.

Here's the post.  I'll try to cover the important bits:

Gated Release
Icecrown Citadel is going to be broken up into four distinct sections: The Lower Spire, Plagueworks, Crimson Hall, and Frostwing Halls. We plan on releasing these four sections of Icecrown Citadel over time and not all immediately when patch 3.3.0 goes live.
The first section that opens will include the Lord Marrowgar, Lady Deathwhisper, Icecrown Gunship Battle, and Deathbringer Saurfang encounters. Progress beyond that point will be prevented for several weeks.
In a nutshell, this means we'll probably get at least three weeks per wing before being able to progress behind it.

My Reaction: This is amazing, and thank you Blizzard.
If you're in an endgame raiding guild then you probably experienced (at least in the past) that the first 1-2 weeks of a new instance are hell.  When Ulduar came out my old guild raided 6-days for the first two weeks because of the progression rush - very top end guilds require more, with the very top (Ensidia/etc) going 24/7.  The really nice thing about this change is it spreads out the "learning the new bosses" phase across a period of two months or so.  I like raiding and I'm willing to put a lot of time into it, but I never liked the first few weeks of an instance.  With this change, we get to learn and see some new bosses, we get gear - but we also have time to go back and run ToGC, to run 10-mans, and to do heroics while experiencing all the new non-raid content.  It'll make for a nice change of pace.

Hard modes would of course take up more time, but it seems Blizzard thought of that:

Easy Mode First
There are other elements that gate access along the way. Players may not attempt any Heroic versions of 10 player encounters until they have defeated the Lich King in a 10 player raid. Similarly, players must defeat the Lich King in a 25 player raid before they can attempt a Heroic 25 player encounter. So players must master every normal difficulty encounter in Icecrown Citadel before attempting Heroic difficulty.
The Lich King may not be attempted until Professor Putricide, Blood-Queen Lana'thel, and Sindragosa are defeated. Furthermore, the Heroic difficulty of The Lich King encounter may not be attempted in any week unless the three aforementioned encounters have been defeated in Heroic difficulty that week. 
Simply put - no hard modes until you've killed Arthas once on that difficulty.  I feel this is another pretty positive change - it's nice to experience all the content once, and it also helps prevent weird situations where one guild may kill a boss months earlier then anyone else and skew rankings.

The Lich King gating is also good because it provides a more "final" Hard Mode boss, unlike Ulduar (where you could go kill YS-3 before doing any of the keeper HMs).  This gives a little more "linear" feel to a winged instance that otherwise wouldn't have it.

One could argue that the nice pacing would eventually go away once you get to hard modes, but no worries, once again we have:

Limited Attempts
The initial number of attempts provided for defeating Professor Putricide is only five. When Blood-Queen Lana'thel unlocks, the amount of total attempts remaining will increase to 10. Then when Sindragosa and the Lich King unlock, 15 total attempts will be available to defeat all four bosses. After a raid has exhausted their attempts for the week, the Ashen Verdict must withdraw their support and the four most difficult bosses all despawn and become unavailable for the week. The limited attempt system is a feature of both Normal and Heroic difficulty.
Yay, limited attempts!  This is a nice equalizer between the more hardcore guilds and the ones that aren't, so in general I'm in support of them.  This has a couple of interesting twists, though:
  • Only four bosses are limited - Putracide, Blood-Queen, Sindragosa and the Lich King are all limited.  The other bosses aren't.  This is a new look at the tribute system, where the "easier" bosses are left available to progress, but the harder "gate" bosses are limited.
  • Limited attempts on normal modes - this is completely new, and we're also not 100% sure how it works.  I am hoping that normal and HM attempts are separate (IE, you would have 15 HM and 15 normal attempts for the gate bosses) but Blizzard may go with a homogenous system.  Separate attempts would be preferrable because it would reduce the problem of "how many HM attempts can we try and still kill Arthas?", which can lead to raid drama and problems.
Of course, this means we'll have another instance where we sit trials and spend hours hoping the DC-boss loves us, right?  Well...

No Insanity
There will be no explicit rewards for defeating the Lich King with a specific number of attempts remaining as there was with Trial of the Grand Crusader. There will also not be an achievement to complete Icecrown Citadel without being defeated by a boss encounter, or letting a raid member die. (i.e. A Tribute to Insanity).
The only real problem I have with limited attempts was the Insanity achievement - not a problem itself, but the fact that it was tied into progression and gear.  And a lot of other people seem to agree, so Blizzard took them out.

There is one more section to Blizzard's post, and I think it's easily going to be the most controversial:

Progressive, Automatic Nerfs
In the weeks and months after all twelve encounters are unlocked, additional attempts against the final four boss encounters become available. This represents the Ashen Verdict growing more powerful and gaining a stronger foothold in Icecrown Citadel. To further help raids, Varian Wrynn and Garrosh Hellscream will begin to provide assistance by inspiring the armies attacking Icecrown Citadel. This is represented as an additional zone wide spell effect applied to all players that will increase their hit points, damage dealt, and healing done. This effect will also increase in effectiveness over time. Players may opt out of the spell's effect if they so wish.
There's two effects here, and both serve to make it easier to complete content.  Over time we will get additional attempts, and we will get a zone wide buff to health, damage, and healing that gets stronger each week/month (or however they do it).

My initial reaction was that this is ridiculous, insane and I can't believe Blizzard did it.  More attempts I can see - going from 15 to 20 attempts won't help a lot, but will serve to let guilds working on the content still feel like they have something to do.  The problem is with the second effect.

The problem is that gear already does this - it increases your damage, healing and health over time.  That's practically the definition of upgrades.  So what is the point of adding a "free gear" aura?  Raids that are doing ICC would already get this effect, so what is the benefit of this change?  It will let more people see the content, yes - and this is a good thing.  But from the viewpoint of a progression raider, it just seems like a horrible idea.

However, after thinking about it I started noticing the last line more:
Players may opt out of the spell's effect if they so wish. 
The wording here is kind of interesting - "opt out".  That sounds like something you might do for a more difficult run, and Blizzard has been known to reward those.  Maybe Blizzard intends for the "Glory of the Icecrown Raider" achievements to only be completable if you opt out of the buff?  Or possible additional rewards/titles by doing so?  If so, that would make a lot of sense - more casual guilds could see the content easier, but the hardcore raiders would be able to opt out of the effect for progression or additional rewards.

I think the system looks nice.  There is a good balance between gated/non-gated, and between limited attempts and unlimited.  They also managed to make Arthas the "ultimate" boss, both in normal and hard mode.

The only change that I'm not really happy with is is the raid-wide buff, but even that could be cool if done correctly.

Monday, November 16, 2009

11/16/2009 - General Update

I'll be posting something more substantial soon but I had a couple of updates I wanted to cover that are pretty minor:

1) I've started logging every raid I do to World of Logs, and I've added a link to that over to the right, under my armory and guild site.  There isn't much up there yet, but tonight's 25 ToGC should go up, and hopefully more content this coming Tuesday/Wednesday.

I use Loggerhead and run WoL's live-raid tool, so generally everything I raid will be up there.  This includes 10-mans (which I almost always tank) and anything my DK raids.

2) I've started rewriting my simulation spreadsheet.  The previous version had a lot of limitations - it's handling of DoTs and cooldowns was pretty clunky and it was hard to get good estimates for the value of RNG stats like crit or hit.  The new version should be a bit more robust and hopefully give more accurate values.

I did sit down and ask myself if I was reinventing the wheel (given tools like Wrathcalcs or SimulationCraft) but decided that it was different enough to be worthwhile - a dps simulation generated in a spreadsheet offers some advantages that Wrathcalcs and SimulationCraft don't (and of course, some weaknesses as well).  And there is always the value of having multiple tools look at the same information, just to see the differences in output.

3) The topic of my blog came up in guild chat the other day and my raid leader asked if I had talked about his gearscore yet - he has the best on the server and he gets tells two or three times a day about it.  Sadly this is actually true - he gets tells like, "Hey I heard you have the best gearscore on the server, that's awesome" (although usually with decidedly less capitalization, grammar and spelling).  So this is for you Romo.

Edit: Followup
Owlkin Frenzy - 0 procs Beast, 1 proc Jaraxxus, 6 procs Faction Champs, 1 proc Twins.  So, useless.

Bad News about tonights raid: I went full retard on Twins and died twice.
Good News about tonights raid: We still managed to one shot it and got our first Insanity (and I got my neck).

Friday, November 13, 2009

How Caster Haste Works

This is a topic I've talked about before (although not on this blog), but I still see a lot of confusion about it.  So if you're don't know exactly how haste works, this will hopefully clear that up for you.

Fair Warning: This post will contain formulas, although I'll do my best to keep it simple.  If you are allergic or if you already know the topic, feel free to skip it.

A Definition
Think of haste as increasing the number of casts you can cast in a certain time.  In other words: If you could cast X spells before in a fight, then 20% haste should let you cast 1.2*X spells in the same time.

This isn't exact - spells can run into the GCD (especially for moonkins), some spells have cooldowns or static durations (so you really can't cast more of them, but instead get more of other spells) and there are also queuing issues with instant spells.  But as a general idea, you need to think of haste working this way.

Keep in mind that while haste does reduce cast time, it isn't a straight reduction - 20% haste doesn't mean taking 20% off of a 3 second cast, for example. 

Calculating Cast Time
This is the current formula for cast time:
Final Cast Time = (Base Cast Time) / ((1 + haste1)*(1 + haste2)*...*(1 + hasteN))
I posted that first because I'm going to refer back to it, but let's start with a simpler example to get the basic idea.  Don't worry if you don't understand the formula at this point.

Let's say you only have one source of haste: Nature's Grace, which is 20% haste.  And you want to calculate the cast time of your Starfire when Nature's Grace is procced.  Then the formula is:
Final Cast Time = (Base Cast Time) / ((1 + haste1)*(1 + haste2)*...*(1 + hasteN))
Final Cast Time = (3) / (1 + .2)
Final Cast Time = 2.5
Note that I use .2 instead of 20% - they mean the same thing, but .2 makes more sense if you are plugging this into a calculator or other tool.  With only one source of haste it's pretty simple - you just take the cast time and divide.  Now take the above example and we'll add in Celestial Focus (3% haste) and Improved Moonkin Form (3% haste).
Final Cast Time = (Base Cast Time) / ((1 + haste1)*(1 + haste2)*...*(1 + hasteN))
Final Cast Time = (3) / ((1 + .2)*(1 + .03)*(1 + .03))
Final Cast Time = (3) / (1.27308)
Final Cast Time = ~2.36
What happens is that every source of haste you have is multiplied together.  Instead of being additive like crit (where 5% from talents + 5% from gear = 10%), we actually get a little extra - 20%, 3% and 3% ends up being the same as 27.308% haste.

So now, looking back at the original base formula:
Final Cast Time = (Base Cast Time) / ((1 + haste1)*(1 + haste2)*...*(1 + hasteN))
  • Base Cast Time - this is the base cast time of the spell before any haste effects, but after talents like Starlight Wrath.  The base cast time on Starfire is 3 seconds - the base cast time on Wrath is 1.5 seconds.
  • haste1/haste2/hasteN - these are placeholders for a variable number of haste effects - you could have two, three, four, or even more depending on raid buffs and talents.  In general every distinct source of haste should be separate and multiplied - each individual talent or buff, and then your haste rating as it's own category (note that this includes haste from stuff like Potion of Speed or haste food as well as haste rating on gear).
I'm going to finish this up with an example using my own numbers and then with in-game proof (just so you don't have to take my word for all this).  For this example I'm going to use Hurricane - Blizzard's spell book only gives you 2-digits after the decimal point, so a longer cast time gives better precision.

My Haste Buffs:
3% from Celestial Focus
3% from Improved Moonkin Form
20% from Nature's Grace (procced with Starfall)
17.17% haste from gear

Final Cast Time = (Base Cast Time) / ((1 + haste1)*(1 + haste2)*...*(1 + hasteN))
Final Cast Time = (10) / ((1 + .03)*(1 + .03)*(1 + .2)*(1 + .1717))
Final Cast Time = (10) / (1.491667836)
Final Cast Time = 6.7
And now in-game:
6.7 seconds, on the dot.

Thursday, November 12, 2009

Tier 10 Moonkin Stats


I know everyone else is already talking about this, but I think it deserves a post.  MMO data-mined the t10 druid sets, and the balance is amazing:

There also should be a Heroic set around ilvl 277, but that isn't available yet.  It should be pretty obvious though that the stats on this are incredible:
  • No Spirit
  • Three Spell power/haste/crit items (helm/chest/shoulder)
  • Two Spell power/hit/crit items (gloves/legs)
  • Only red/yellow sockets and only spell power bonuses, and every item has at least one red socket.  So you can gem for every bonus and be gaining dps.
I'm going to go out on a giant moonkin-shaped limb and state that either gloves or legs will most likely be our off-set.  Hit is going to be incredibly easy to get at t10 (moreso then t9) and using two tier pieces with hit will probably put you way over cap, especially with the rep ring adding hit.

Now which piece is off-set will depend on what other loot is available (and there's also a small potential for us to actually use 5/5, depending on what else is available), but at first glance this is very promising.

One more last thought - get used to having high haste values again.
  • 3x Tier pieces have haste, and you will probably end up using all of them.
  • The rep ring has haste.
  • The Emblem leather belt has haste (and is SP/haste/crit so you'll probably use it unless something better itemized is in 25 HM).
  • Wrist/Boot pieces that are BIS will probably be SP/haste/crit if the pattern continues.
So, to sum it up:
You will likely have haste on every single piece of gear, barring trinkets and 1-2 other pieces that have hit instead.  So be prepared to break 600 haste again, possibly at high velocity.

Tuesday, November 10, 2009

Power Auras And Innervate

About one month ago I read this post on the WoW druid forums.  Since then it's become especially useful on fights like Anub, where mana-management is extremely critical to performance.

Power Auras is an extremely flexible addon available from Curse.  It allows you to create a visual "aura" on your screen based on almost anything: buffs, debuffs, buffs/debuffs on your target, mana/health percentages, abilities off cooldown, abilities ON cooldown, and more.  What's interesting about that WoW forum post is it utilizes multiple auras that reference eachother, which creates a neat "AND" logic.  The first aura tells you when Innervate is off cooldown.  The second aura activates whenever your mana drops below 65% (I find this to be the bext amount where I can hit Innervate and not waste anything gained).

By linking these two together you get an indicator of when Innervate is available and when your mana is low enough to justify hitting it.

First you create the Innervate aura:
  • To access Power Auras type /powa.  Then hit "New" to create a new aura.  After selecting it, hit "Edit" to start.
  • Set the Activation By to "Action Usable".  This tells Power Auras to display the aura when the named ability is available for use.
  • In the bar under Action Usable, type in Innervate.  Make sure this is exact, although you can tell it to ignore capitalization by checking "Ignore Uppercase" below.
  • Only if in combat - this is optional.  Checking this will make the aura only activate if you are in combat, which will prevent you from seeing it every time you change specs or otherwise drop in mana.  I don't use it because it rarely matters for something like this.
  • Set up the display - you will want to set up the aura to display in whatever manner you want.  I make the innervate aura nearly invisible by making it small and low opacity, because I let the mana aura dominate.  The post where I got this idea said you could turn one of the auras off (so the display wouldn't matter) but I had issues with getting it to work when I did this, so I leave both of them active.
Second, create the mana aura:
  • Activation By should be "Mana".
  • The bar below this will turn into a slide.  Set it to whatever percentage of your mana is around 10,000 down - I usually use 65%.  This ensures that after you hit Innervate you won't waste any mana from other gains, while still letting you use it early.
  • Set up the display.  This is the aura that I want to be visible, so I make it go bang - it's in the middle of my screen, it's pretty large, and it's blue (because my brain associates that with mana/innervate).  I use a middle opacity so it doesn't totally interfere with my vision, but it still shows up immediately.
Finally, you need to create the linkage.  Type in /powa and hover over each of the auras you just created.  If this is your first time with Power Auras then those numbers should be 1 for Innervate, and 2 for the Mana.  Personally, the numbers are 3 and 4 for me.  Whatever they are, write them down or remember them.

Now go to your Innervate aura and hit Edit.  Down under where you typed "Innervate", and to the right of the "Exact Name" checkbox, there should be a second little box.  In this enter the number for the Mana aura.  Your final Innervate aura should look something like this:

Do the reverse in your Mana aura - in the same little box, enter the number for your Innervate aura.  So if Innervate is 1 and Mana is 2, then you put "2" in the Innervate box, and "1" in the Mana box.  This tells them to reference each other and only activate if the other one is also true.  Your final Mana aura should look something like this:

The end result of these two auras is that they will only show up when you are both below 65% mana, and have Innervate off of cooldown.  My display shows up like so:

One more thing I do that's related this is create another aura that displays prominently when I'm below 10% mana.  This is the range where I can't afford to Hurricane, so it's significant - it lets me know when I really, really need to Innervate, or when I should consider hitting GotW.  You can simply copy your Mana aura and remove the reference to your Innervate aura (or leave it in, if you feel like), and then change the display to make it show up better - I changed the color to bright green and put it at 100% opacity.  The end result is something like this:


I find that this combination really helps me with mana management and lets me utilize Innervate as much as possible.  And while on most fights this isn't key, it's especially useful on stuff like Anub'arak where AOE - and therefore mana management -  is critical to both the fight and maximizing your dps.

Wednesday, November 4, 2009

Movement And The Hybrid Dps Tax

About one week ago Vaneras posted a sticky about hybrid vs pure dps.  The key part is about the hybrid dps tax:
In our design, the pure dps classes (hunter, mage, warlock and rogue) should do slightly higher dps than hybrid damage-dealers all things being equal. All things are rarely equal. Player skill, gear, raid comp, latency, random luck and most importantly the specifics of the encounter will often favor one class, spec or player over another.
This raises the question of what mechanics tend to influence dps.  There are two basic types: 

Direct.  These are buffs like Storm Power on Hodir or Shadow Crash on General Vezax.  They have an immediate effect and will obviously favor certain dpsers over others (high-crit classes for Hodir, casters for General).  Direct mechanics are generally a positive increase in dps, although there are exceptions. 

Indirect.  These are primarily mechanics that force movement (although something like XT's Heart that favor a 30-second window of dps count too).  These are present in practically every fight, and they generally decrease dps.
  • Ulduar: Nearly every fight has something that forces movement, with the possible exception of Flame Leviathan and XT (if you get lucky and never get a Light Bomb or Gravity Well).
  • Northrend Beasts: Snobolds and Fire on the Ground for Gormok.  Poison and Burning Bile for the Worms.  Moving to dps Icehowl when he's stunned.
  • Lord Jaraxxus: Legion Flame, and movement to keep Nether Portals or Infernal Volcanos in range.
  • Faction Champions: Yes.
  • Twin Val'kyrs: Light and Dark Orbs, and switching Essences for a vortex/shield.
Direct Buffs are the exception - the only one in ToC is from Lord Jaraxxus, and only one class (mages) can benefit from it.  Most mechanics that effect dps are of the indirect type, and practically every fight has them - because a fight where you stand still and never move gets pretty boring.

The problem with this is that fights with movement tend to favor pure dps, and most fights are movement fights.  Not for player skill reasons, but because all of the pure ranged dps classes have an instant/high movement ability:
Hunters: Disengage
Mages: Blink
Warlocks: Demonic Teleport

These abilities aren't perfect or flawless but they are much better then what moonkins, elemental shamans, and shadow priests have - which is nothing.  On any fight with a decent movement requirement (which is practically every fight) they give the pure dpsers an even bigger advantage then they might otherwise have.  I never expect to beat a good mage on Hodir, because I know that mage can blink to pick up Storm Power, to avoid crashes, to get fire - the list goes on.  I know that a warlock on Twins should beat me because he can teleport back and forth between essences as necessary, and he has an easier time avoiding Orb Surges.  A hunter on Mimiron should beat me because he can jump over fire, and I can't.  And the same holds true for most current fights - there are always mechanics that favor instant movement abilities, and therefore favor pure dpsers.

To a large extent I don't have a problem with this.  The issue is that hybrids come under a double tax, once for being hybrids and then a second time for dpsing a fight with movement.  If I can be within ~5% of a mage on a stand-and-nuke fight then I'm ok with that - the problem is when I drop to 10% or 20% below a mage because he has Blink, and I don't.

I hope that Blizzard fixes this in Cataclysm by giving all the hybrid casters some way to move around instantly, but I'm not super confident - it seems to be an issue that floats under the radar most of the time.