Thursday, September 30, 2010

[Druid]The 4.0.1 specs/gear/glyph post

Still working on the haste articles, which is proving a bit difficult in the mathy departments. The short answer is that haste is going to be awesome as long as bleeds aren't more than half our overall dps - but there's some good evidence that this might actually be the case. In which case...mastery is teh bombzors, and haste is teh sux.

Anyway.

Since it's very, very likely that the 4.0.1 patch is going to come out next week, I thought I'd prepare folks for the coming...uh...cata...err...patch day. Yeah, patch day. There's lots of articles about what you should be farming or what you should be doing in the final days. I just want to let you know how to gear and how to spec.





At level 80 you won't be able to get either perseverance or master shapeshifter, so you might as well go for some of the less useful talents in the tree - like Stampede and Fury Swipes. You could consider taking King of the Jungle instead if you have 4pT10, but I wouldn't bother - and definitely don't if you don't have 4pT10. Otherwise it's about as cookie-cutter as you can get; there are almost no talents in the bear arsenal that are nice to have, and most of the skipped talents are either entirely for cats or are super weak for bears.

For glyphs, here's what I'd use:
Glyph of Rip (right now you literally only have two prime glyphs for bears, so might as well)

Glyph of Maul (substitute Glyph of Rebirth for this if you like)

Not horrible, but I hate having that rip glyph.

For gear, you'll lose a LOT of extra armor from your armor rings/trinkets/cloak, so it's a good idea to pick up the higher ilvl gear without extra armor and get some of that stamina back. Double stamina trinkets are likely not a bad idea if the armor nerf on gear continues. Do consider getting non-'tank' rings and amulets as well for dual use; they'll have as much stamina as the tanking jewelry, and should be good options. As an example, the often hated Seal of Many Mouths is probably going to be one of the better pieces you can get. Remember that you'll be losing 10 expertise right off the bat, so replacing that should be up on your list; while things won't parry haste, it will be good to keep threat high. Similarly, Sindragosa's Cruel Claw should rival some of the better tanking necks out there. Gemming should almost certainly be stamina, since it now provides both health and some mitigation via vengeance-fueled Savage Defense

On reforging: haste is the least useful stat for bears given how GCD-focused they are; while it does increase rage income, it won't do so insanely well. As a result, you should consider reforging all the haste you can into mastery or dodge. My gut feeling is that your dodge% will be ridiculously high, so mastery will likely provide a better overall result of reduction of damage per point. If you really need to, get expertise or hit, but try to get pieces that have that instead.

For raiding, Bears are actually in a pretty decent spot. Their threat has been rebalanced significantly and mangle proccing off of lacerate combined with mauls every 3 seconds and very high frontloaded swipe threat and lacerate threat puts bears in a good spot. Savage Defense in a weird, backwards way ends up scaling with bosses...kinda...and certainly is better than it is now. Reforging helps bears quite a bit, and all that extra stamina on agility pieces gives bears tons of options they didn't have before. Skull bash even acts as a mini-charge of sorts, and the cooldowns are significantly better than they were. If you were tanking at a high level before, I think you'll be in great shape to tank now.




Similarly to bears, as a cat you can't get master shapeshifter, so those points have to go into the feral tree. This time, though, there are a lot more options. This build focuses more on utility and survivability, picking up Nurturing Instinct and Furor and skipping Predatory Strikes. In fact, that's the only dps-related talent that was skipped, and it was skipped because at level 80 you will have an absurd crit rate and don't need the boost to ravage.

The rotation hasn't changed at all, though the length of some of the buffs have. At level 80 it should be absurdly easy to maintain a rotation, given how fast you are with haste.

As to the glyphs, there's some interesting ones:

Glyph of Claw (which will be Glyph of FB)

Not the best options for DPS in the major glyph categories, but hopefully these will change soon.

In terms of gear, you won't have any armor pen - you'll be swimming in crit and haste. Your first goal should be to regem to hit the hit/expertise caps if you aren't already (including the missing 10 expertise from Primal Precision) then go pretty much all in on agility or haste. As I state below consider going for agility, since you can reforge to haste. Ignore crit rating; at level 80, in T10 gear that wasn't BiS, I had something like 82% crit rate. Because of that, the Arpen trinkets lose value compared to Herk or WFS, since you'll have such high crit rates anyway. DBW is still kinda awesome because of the trigger, however.

For reforging, you can't reforge to agility, so consider reforging to either haste or mastery. My money's on haste, but this is one of those things I'll have to test and retest. Numbers are (still) not remotely set in stone for ferals, so this is likely to change some. My gut feeling is that if the 4.0.1 patch comes out on Tuesday and it's anything like what we currently have, cats are going to be doing ridiculously low damage compared to what they do now - likely around half the damage. I'm not sure they're a reasonable option for any hardcore raiding at this point. Hopefully they fix things a bit more.

Tuesday, September 21, 2010

[General]Cataclysm ratings

Just an aside before I start digging into haste, hit, and expertise - some of the complaints about the previous articles were that the ratings at 82 aren't the same as they are at 85. Which is true, but it doesn't matter as I was comparing them to each other.

If you're curious, the ratings at level 85 have been figured out. You can see the results thanks to Whitetooth over at EJ. They're reprinted here for ease of use at level 85.

Dodge: 176.71
Parry: 176.71
Block: 88.36
Hit: 120.109
Spell hit: 102.446
Crit: 179.28
Haste: 128.05
Expertise: 30.02
Mastery: 179.28

A druid starts with 7.4755 base crit %.
At level 85, it takes 324.85 agility to get 1% crit.

The agility to dodge conversion is unknown at this time.

The important point to note here is that everything stated before was correct. Agility is still 1.8x as expensive as crit rating to get crit. Mastery and crit rating are equivalent, and haste is much cheaper. Hit and expertise are slightly cheaper than haste per 1%.

Friday, September 17, 2010

[Cat] Cataclysm cats part 3: the case for crit

Following up on the previous articles:


Now it's time to dig into crit. And not just crit rating - but agility.

From the first article we know that crit rating and mastery rating are basically equal (at least right now) in terms of what they give, and the same amount of crit rating to give 1% crit is the same amount of mastery rating that gives 2.5% bleed damage. That's nice; it makes it very easy to equate them without going nuts.

In addition to that, agility gives crit% as well. It takes 1.8 points of agility to give the same amount of crit as 1 point of crit rating, if that helps.

How does crit help you?

Well, let's see. Crit gives you 200% damage on all critical strikes. This includes bleeds, melee, and any special attacks. The 220% boost on crits is either not implemented or purposely taken out at this point; my guess is the latter. Add the metagem into the mix, and your crits do 206% damage.

For every 1% crit you should in theory get 1% more damage, roughly. That's not exactly true. For starters, there is a cap on crit% for white attacks (76%). There's a soft cap for things like ferocious bite and ravage as well, though that doesn't often come into play. And mostly, it's a matter of degree; if you sit at 0% crit, 1% crit will increase your damage by 1%. If you're at 40% crit, the crit boost only increases your overall damage by .7% (since you can't make your crits...uh...crit again). That sounds awfully close to the mastery bonus in terms of damage potential - so are they equivalent?

Well, without knowing what the crit rates at 85 are going to be, it's hard to say. Chances are that a 40% crit rate is pretty high, actually. At level 80, assuming ICC gear you get absurd amounts of crit thanks to every bit of arpen converting to crit (almost) - but that crit decreases very fast. Instead, we'll look at median damage values for crit rates:

0%: 1% crit = 1% dps gain
20%: 1% crit = .83% dps gain
30%: 1% crit = .76% dps gain
40%: 1% crit= .71% dps gain
50%: 1% crit = .66% dps gain

Note that this is the case for bleed damage too - as your bleed damage multiplier increases, the overall damage increase each point does is less relative. The overall damage increase from a raw standpoint should the the same though, and note that in order for bleed damage to keep relative pace with crit%, bleed damage has to be around 40% of your overall damage.

Anyway, that's part of crit's success, but the neat part is when you talk about the extra combo points generated. Here's where we get more meaty bits.

The cycle in cataclysm is mostly unchanged from wotlk; you keep savage roar up, you keep rip up, and you keep rake up while shredding when you have spare energy. Rip always at 5 CP. SR when you can (and as I'll point out, it's actually a bit more efficient to SR early). The change comes from the endless carnage talent. Rake now will have a duration of 15 seconds (up from 9) and SR will add an additional 6 seconds to its duration regardless of combo points. Rip's duration is unchanged, and because the glyph of rip is being changed to a straight DPS boost instead of a duration boost, the maximum duration rip will have is 22 seconds.

This means in practice that you must generate 5 combo points every 22 seconds to keep rip up - and do three shreds in that time as well. In addition, you should be keeping Savage Roar up. For ease, I'm going to assume a 5CP savage roar, which means 40 seconds total. Realistically this means that you need to be generating 15 CP every 40 seconds or so just for the basic rotation.

That doesn't sound so bad until you look at the energy that takes. In 40 seconds you'll have 400 energy with zero haste, along with another 60 from at least one Tiger's Fury. You need energy for two rips (30 energy each, so 60 total)and a SR (25 energy) in that time, and we'll assume you're a pro at keeping rake up too - so you need energy for 2.6 rakes (90 energy). That leaves 285 energy for shreds, and shreds are 40 energy each currently.

In other words, 7 shreds. You have 2.6 rakes, 7 shreds, and you're needing to generate 15 CP in that time. The effective crit % required to keep that up is 56% if all you did was that. Now, you'll hopefully get at least one OoC proc - probably 2, so let's assume that. Even so, in that case and assuming no extra CP get wasted (going from 4 to 5) you need an effective crit rate of 30%.

So forget about ferocious biting; you won't have the CP generation. In fact, if you're below 40% crit rate, your rotation is going to be pretty horrible. Rake is the most efficient combo point generator you have, but you won't be raking that often compared to now (every 9 seconds or so vs Cataclysm's 15 seconds or so). What that implies to me is that crit is going to be very, very important early on.

Which makes sense; some of the basic advice from most ferals in the early and midtier of WotLK was to get crit at 50% at least before going for arpen, and the same advice applies here as well. Except we don't have anything like arpen in terms of boom or bust stat, so it's much more forgiving to go for.

One of the things I expect is some criticism for this post as I assumed 0% haste - which isn't realistic, not by a longshot. The point is to illustrate how important crit rate is for ferals (especially compared to mastery), not to devalue haste. I'll get into this in the next post, but as a teaser - the above assumed 0 haste. If you assume around 20% haste (which with stats and buffs is about right) you get 100 energy every 8 seconds. That means that in 40 seconds time instead of 460 energy, you get 560 energy, meaning that even in the worst case scenario (no OOC procs) you get 2.5 more shreds than the 0 hasted situation - which requires an effective rotation crit % of only 25%! if that sounds awesome, it's because it totally is. That it increases your OoC procs by 20% and increases your melee damage by 20% doesn't suck either. But like I said - we'll get into that next time.

Monday, September 13, 2010

[Cat] Cataclysm cats part 2: the case for mastery

This is the second article in the series, following up from the basic one. Today we're going to talk about mastery - and what mastery does for a cat.

Upon gaining mastery, Feral Combat specs get 8 points of mastery, which increases your bleed damage by 20%. Not bad - this would be (if on live) about a 3% DPS upgrade. Each point of mastery - which is similar to points of expertise - increases all bleed damage you do by 2.5%.

At level 82 it takes 79.15 points of mastery rating to get one point of mastery skill. This is the same value that it takes to get 1% crit from rating as well, so mastery and crit are likely equallly valued as far as gems and itemization goes. Which makes crit rating a very good comparison, overall, in terms of value.

For crit rating, since all dots can crit and all attacks can crit (up to the glancing blow cap) we have almost a pure 1% DPS upgrade per 1% crit. It's actually better than that due to how the meta works (gives 3% more crit damage), whether or not the 10% more crit damage boost is baked into the spec or not, and the addition of extra combo points on crits. But let's just assume that 1% crit only increases overall damage by 1%.

What would it take to have mastery do the same thing?

Well, simple math says that as long as bleeds are doing about 40% of your total damage, each mastery skill point will increase your DPS by 1%.

That's not how much bleeds currently do, on live. And for various reasons, that's certainly not how much damage they do in the beta, either. But that can be fixed.

In live, assuming you have a fairly good rotation and a reasonably static fight, bleeds account for about 30% of your total damage assuming 4pT10 and a high-armor pen build. (I'm basing that on my performances on Saurfang, which is a fairly good gauge of a basic single-target fight). On beta, there are a couple problems that cause this to be even worse:
  • Savage Roar was changed to only improve autoattacks - which nerfed all bleed damage and all other yellow damage by 30%. This hasn't been baked back into the yellow damage, and as a result bleeds do less.
  • While autoattack, mangle and shred are based now on weapon DPS, rip and rake are based only on AP. With the removal of Feral AP, this results in about a 30% reduction in those damage sources as well.
  • The glyphs had not been particularly well-optimized for bleed damage until recently. Both rip and shred increase rip duration, but with dot clipping gone from cataclysm duration of bleeds is not nearly as important. The new glyph of rip increases rip damage by 15%, which will help this, but shred is still suboptimal (especially since it...sigh...doesn't help with the horribly newly named talent Feast of Flesh)
Now, all of these things can be improved. And I expect some of them to be improved a great deal; rip and rake can't only scale on AP forever, for instance, and glyphs will get a serious overhaul. At the same time, other abilities will get buffed too. And while it would be an interesting way to go, it's hard for me to imagine that two bleeds will do 40% of overall damage for ferals. It's possible - but I think it's unlikely. It's more likely to me that they'll buff mastery so that it's a bit more inline with the other skill points.

In any case - my recommendation for the beta right now is to either ignore mastery or largely avoid it. I'd like it to improve; it would be very boring if the new skill we can get is basically meh. But we'll see.

Tuesday, September 7, 2010

[Cat] Cataclysm cats part 1: the basics

Cataclysm changes a lot of things in terms of stat values - far more than they were changed from TBC to WotLK. While this doesn't have a huge effect on bears (the tank changes were largely around block, which bears don't have), the cat changes are pretty massive.

Specifically:
Agility provides 2AP
AP is gone completely
DPS of a weapon is used to calculate base ability damage
Arpen is gone completely
All bleeds can crit
Mastery exists and improves bleed damage only (the opposite of arpen, basically)
Haste improves energy regeneration as well as attack speed (ETA)

With all that, I thought I'd go through a series of articles that talks about the various ratings and what they mean. Note that I am not going to evaluate them based on the current values of damage, as they are no where near current. Instead, I'll evaluate them based on what they would do given a certain plan. That plan is based entirely on our current ratings of damage, with the savage roar change factored in.

Currently my damage spread (I'm close to BiS, so take that into consideration) is something like this:

Melee: 38%
Shred: 25%
Rake: 15%
Rip: 15%
Ferocious Bite: 6%

Savage roar is being changed from a 33% boost to all damage to a 53% boost to melee only. The new breakdown just from that would be:

Melee: 56%
Shred: 18%
Rake: 11%
Rip: 11%
Ferocious Bite: 4%

I'll be using these numbers to showcase the relative strengths and weaknesses of ratings. Note that this is in no way a perfect scenario of what Cataclysm will be like in terms of damage spread. Here's some quick thoughts on that, however:

  • I do believe melee damage will increase and yellow damage will decrease due to how Savage Roar now works.
  • Ferocious Bite will likely increase in use due to the lengthening of duration of rake and savage roar and the effects of Nom Nom Nom.
  • Because dots cannot be clipped, the uptime of dots should be very high - close to 100% with a near optimal rotation. This should increase bleed damage a small amount.
  • Because there is no armor penetration, non-bleed damage will decrease some.
  • Combo points are going to be harder to come by instantly; 4 or 5 yellow attacks are going to be necessary to get 5 CP easily due to less crit%, at least to start.
  • Hit and Expertise may be required depending on whether refunds on yellow attacks still work. I suspect the refund will be built in, but it may not be.

Now, here are the specific effects of each stat as you gain them:

1% crit:
  • At level 82, it takes 79.15 crit rating to give 1% crit
  • All of your direct damage melee attacks crit 1% more often (for 220% damage)
  • All of your bleeds crit 1% more often (for 200% damage)
  • Shred, mangle and rake provide 1% more double combo points
1% haste:
  • At level 82, it takes 56.5 haste rating to give 1% haste.
  • You attack with melee attacks 1% more often
  • Your Omen of Clarity has a greater chance to proc
  • You regenerate energy at a rate 1% greater
  • Your bleeds tick 1% faster (removed per GC recent comments)
1 point of mastery:
  • At level 82, it takes 79.15 points of mastery rating to get one mastery point.
  • All of your bleeds do 2.5% more damage
1 point of agility:
  • At level 82, it takes 143 points of agility roughly to get 1% crit.
  • Increases your attack power by 2x, increasing all damage by a small amount.
1% worth of hit:
  • At level 82, it takes 53.2 hit rating to give 1% hit.
  • Increases your chance of all abilities hitting by 1%, up to a cap.
  • Does nothing for bleed damage
  • Increases your chance for spells (FFF, hibernate, soothe) to hit by 1%
1 point of expertise:
  • At level 82, it takes 14.05 expertise rating to give 1 expertise skill. (The assumption is that this is the same value for hit as 4 expertise skill would be)
  • Increases your chance of all abilities not being dodge (or parried) by .25%, up to a cap
  • Does nothing for bleed damage
  • Does nothing for spells (growl, FFF, soothe, hibernate)

Of these, does something stand out to you? It certainly does to me. My theory going into this is that haste is going to be a very, very desirable stat for ferals - possibly the most desirable stat by a long margin. In the next few articles, I'll explore that.

Of course, this can be adjusted for ferals or whoever. The problem is that haste isn't that valuable for a lot of other classes due to the energy regeneration. My suspicion is that they're going to nerf that part hard once they see ferals and rogues going nuts with haste, but it'll be a while.

Wednesday, September 1, 2010

[Druid (and really, what else is there] - setting Cataclysm expectations

It is almost September. The Cataclysm approaches.

I don't know any more about when it'll launch than anyone else. Except that it will launch sometime this year, and Activision likes late November dates. So I thought it was a bit educational to look back on what was done by this time 2 years ago, when WotLK was about 2 months away.

  • mostly finalized talents and a talent tree (berserk had the last stand effect)
  • removal of crittability
  • removal of crushing blows

What we didn't have:
  • A finalized talent tree (survival instincts was still not there, as an example, and FFF was)
  • Fully balanced skills (rake wasn't even considered for rotation, barkskin wasn't usable in forms)
  • Any idea of what itemization was good or not
  • Glyphs
  • All the stuff that came in WotLK proper (Feral AP removal, bonus armor removal, nerf of stam, FF's buff, mangle's duration buff, primal gore

  • A completed talent tree (essentially)
  • More tweaks to skills (bash interrupting, growl taunting at range)
  • Additions of things like an OoC res, barkskin in forms, pots in forms
  • More of a clear idea of what is going to drop
  • Removal of +armor from leather (but not from items)
What we didn't have:
  • All skills finalized
  • Full lists of all gear
  • All numbers balanced

  • Talent tree still complete
  • addition of skills like Savage Roar (seriously, it came THAT LATE)
  • Swipe hitting infinite targets
  • Mostly finalized glyphs (the SI glyph wasn't in there yet)
  • Things like on hit procs working with cat/bear form
  • Mostly finalized professions (including the bonus for each being largely normalized)
What we didn't have:
  • Many of the early fixes in WotLK (like bonus armor not being multiplied, Feral AP being removed)
  • Some balance issues (like swipe doing a lot more threat)
  • knowledge about the armor change or the FAP change (this came early November)
Now, let's assume for a moment that the release schedule is basically exactly the same as WotLK. There are reasons to expect otherwise - such as the huge amount of stuff that simply doesn't work in the beta right now - but let's go with that. What does that mean?

First off, the talents are largely finished. They might juggle some (I expect berserk to be changed significantly) but we're not going to see game changers like Titan's Grip or Chakra. For the most part, the tree is pretty well done in terms of what's there. We didn't see a lot of changes to the feral tree between August and October 2008 - the changes were to SI, FFF and Berserk with a couple of modifications to how many points each talent cost. Not a lot else was changed.

Secondly, the skills aren't remotely finished. Between August and October we had Savage Roar added, rake buffed, barkskin buffed, FFF and SI changed (and FFF made trainable), growl given a range, swipe given unlimited targets and a threat boost, bash made into an interrupt - and that's just the stuff I can find easily. There is likely to be a lot of churn in the next couple of months on skills, and even moreso on skill numbers.

Thirdly, the overall themes of the expansion are pretty well set for ferals. Don't expect big sweeping changes or big design overhauls like you saw for paladins recently. The expectation is that cooldowns will be normalized, things like incoming damage rates will matter (because mana of healers will matter), avoidance will be important, CC will be important and the longevity and time to live of tanks will increase while the survivability of DPS will hugely increase. Talent trees will still continue to focus on giving interesting talents, not ones that boost DPS or tanking specifically.

Fourth, expect a LOT of itemization coming later. Itemization happens quickly and easily for the most part. Don't worry about it any time soon.

So what kind of predictions can we make about this?

Well, here are mine:

Berserk is going to get changed...again. With Mangle being completely without a cooldown and with cleave/AoE abilities largely being nerfed or deprioritized, berserk is really without a purpose. As the marquee 31-point talent, it will need to be changed - or bears will spend that point elsewhere. Expect it to be redesigned.

Expect a couple more skills down the pipe. My gut feeling is that they're going to add at least one bear skill (or allow a non-bear usable skill to be used in bear form like thorns) and possibly two. I expect there is going to be at least another way for a bear to give themselves some more rage as well.

Talents will get juggled around a bit. The tree is weirdly unbalanced right now for both cat and bear, requiring way too many or too few talents (tier 2 for bears is almost all things you want; tier 4 is almost all things you don't). That will be fixed or at least improved on.

Numbers aren't remotely final. Don't bother number crunching unless you want to figure out (for example) what the effect of savage roar going only to melee damage is (it makes haste really awesome and hit/expertise even more awesome), or how bad fury swipes is right now (it makes baby Phaelia cry).

Rotations aren't set in stone either. Rake got added late, FFF got added as a useful damage/TPS ability late, SR even got added late!

Balance -especially tank balance - hasn't been touched. And bears are going to be nerfed at least in health and likely in a couple other places. I wouldn't expect Savage Roar to be looked at for a while, even if you know it's going to be pretty lame.

What this means for you, as a member of the community:

  • Don't spend a lot of time working on huge revisions to talents or adding tons of abilities. That's just not going to happen. Juggle talent trees, but don't expect big changes.
  • Do think about what the bear and cat are missing and what is fun and isn't fun, and how it can be improved. Barkskin got changed to be usable in forms, and SI got split from berserk because of this kind of input.
  • Don't stress too greatly that ferals haven't gotten a lot of love; ferals historically have been late bloomers.
  • Don't expect huge revisions in general. As far as I can tell developers felt ferals were in a pretty good place at the end of WotLK and weren't going to rock the boat too hard. If you didn't like them before, this is a good opportunity to switch classes. If you did like them before, chances are you'll get some nice quality of life fixes.
With that all said, I'm not planning on doing a lot of number crunching for a while - at least until level 85 folks are able to do some threat and damage assessments. Until then it's pretty much a waste of time. When we get more solid numbers I'll start looking at the various rotations and the value of stats for different rotations as well as showing how bad certain things are (like fury swipes). But doing anything before that is a bit of a pointless exercise.

Also, thanks to folks who asked me to start blogging a bit more over on wow.com. And thanks for everyone that checked that article out; Skunkworks is one hell of a special guild, and I'm very proud to be a part of it.