Wednesday, July 29, 2009

[Druid, 3.2] Emblem of Triumph items for cats

And with the bear one out of the way, time to do the cat emblem of triumph gear. (Really, it's not that bad Jacemora). Fortunately all the prior items are basically covered save a couple trinkets, so that saves some time - there really aren't that many pieces from purely Emblems of Triumph that are obtainable, and the ones you might care about are fairly small.

The scary list - the one that has all conquest, valor, heroism and triumph loot - is coming.

As before, the gearing list is based on FeralByNight's system of weighing. Keep in mind that this is very, very dependent on other stats; do not use these values without understanding why they're good, and what to do if you don't have a ton of armor pen. In general a good rule of thumb is to use armor pen whenever the gear isn't going to be shared, and use agility whenever it is. If you don't have grim toll or mjolnir runestone, armor pen isn't as valuable as agility - but it doesn't really change these rankings at all.

The tier gear question is a lot more interesting this time around. The 2pT9 bonus is pretty good, though it's not as good as the insanely great 2pT8 bonus is. The estimates I've seen value the bonus at around 50-100 DPS, given that it's basically giving you a free shred while keeping higher rake uptimes up. The 4pT9 bonus is not nearly as sexy, by comparison. Ferocious Bites are a very small amount of overall DPS, even with higher gearing. 5% more crits on rips is very good, but it's not as good as the 2pT8 bonus or even the 4pT8 bonus, so realistically you should only break 2pT8 if you're getting a significant upgrade in the item quality.

Now, another interesting facet here is that because the gear is becoming better 'itemized' for rogues, it's becoming much better for cats and much worse for bears per item point. This is why, for example, hardmode Ulduar gear is so great compared to hardmode Coliseum gear for bears. The flip side of this is that cats are just getting more and more awesome.

If you want a summary of what to aim for, here goes:

For the Ulduar-25 raider who's got everything: Get the Ring, the Idol and the Hood. Don't bother getting T9.10 gear at all, as you'll likely get 2pT9.25 soon enough.

For the U-10 or lower U-25 geared cat: get the ring for sure. Get the Duskstalker shoulders. Get your T8.25 head and chest, then get T9.10 legs and gloves. If you're really wanting to push your DPS, go for the Hood of Lethal Intent first.

For heroic-gearing cats only:Get the ring first, then the duskstalker, shoulders, then the idol. Get the T8.25 head and chest as the biggest priority. Get the leggings of wavering shadow, and then get the T9.10 gloves - and maybe consider the T9.10 legs if you're running a lot of heroics. But since you're so starved for Triumph badges, the best upgrade by far is that ring.

  1. Dexterous Brightstone (35 emblems) - This currently ranks as the second-best ring in the Coliseum, second only to a hardmode drop. While the loot lists are clearly incomplete, that doesn't stop this from always being great - it's about 5 points higher than the best ring from Ulduar save the quest rewards from Algalon. Because it's so cheap and because it should remain best in slot or close to it for a while, and that rings are traditionally hard to come by since so many other classes need them, I'm recommending this as the best thing to get early on. Of course, if you have two awesome rings disregard. :)
  2. Hood of Lethal Intent (75 emblems). Remember where I said this was more of a cat item? Well, here it is. This is actually better than the T9.25 piece due to having hit and armor pen instead of crit and expertise, and all the other stats are identical. The reason this isn't quite as high is because of the expense, but make no mistake - this is 11 points higher than the Ulduar equivalent, and it will be an excellent upgrade for a while. It's also only a couple points worse than the hardmode T9.25 gear, again thanks to having armor pen and hit over expertise and crit. Definitely a good choice.
  3. Duskstalker Shoulderpads (45 emblems). These are about a 7-point upgrade from the bset non-hardmode item in Ulduar, and a couple points better than Shoulderpads of the Intruder. They are almost identical to the T9.25 shoulders as well; they trade haste for crit. Personally I find crit to be better, but FbN values haste slightly more. In nay case, they're both very close, and because they're so cheap and require no Trophy of the Crusade, they should be a high priority for most cats.
  4. Idol of Mutilation (25 emblems). Unlike the data for bears, this is a great idol for cats; Rawr has this as about an 100 DPS upgrade over Idol of Worship, assuming that the 200 agility is essentially up all the time. Given that it lasts for 16 seconds and that it triggers off of shreds and mangles, I suspect that's right. This isn't as high a priority because simply other items give a bigger bang for your buck, but if you're hurting on an idol it's hard to go wrong here.

  5. Runetotem's Shoulderpads of Conquest (30 emblems) - These are a fairly reasonable upgrade from T8.25 shoulders (3 points) and just barely below Shoulderpads of the Intruder. They're also quite cheap. If you desperately want to have 2pT8 + 2pT9 right away, this is a good spot to have. It's up to you whether the tier bonus and look is worth the slight downgrade from Duskstalker's.
  6. Runetotem's Handgrips of conquest (30 emblems) - These are a slight upgrade from the T8.25 gear, (about .75 points), and are loaded with hit, making these possibly somewhat awkward to use. They're not a huge upgrade but are fairly cheap, and could be a decent spot to go 2pT9 if you've got good options in other slots.
  7. Runetotem's Legguards of Conquest (50 emblems) - while these are a raw upgrade over the T8.25 item, they're not that great. Why? Because for Emblems of Conquest you can get the Leggings of Wavering Shadow instead, which are .08 points worse than these. Between the high cost and the alternatives that are available, the only reason to get these is because you really want a 2pT9 bonus.
  8. Runetotem's Headguard of Conquest (50 emblems) - not a good choice at all. They're a small (1 point) upgrade over T8.25, but because of the excellence of Hood of Lethal Intent, they get marked way, way down. Since you can get T8.25 via emblems of conquest, these become even worse. They're also not well-itemized to begin with. A poor choice to upgrade overall.
  9. Runetotem's rainments of conquest (50 emblems) - A decent upgrade over the T8.25 chest, but not an upgrade over a number of other Ulduar items like Winter's Icy Embrace. Still you should consider getting T8.25 with conquest badges anyway and do this later.
  10. Shard of the Crystal Heart (5o emblems) - this one was tough. With the huge amount of hit, chances are you'd go over the hitcap on this. But it's otherwise great as far as trinkets go, and if you've gotten unlucky with trinkets it may be worth it to you. For me sampling this in Rawr, it looks like it's significantly below Mirror of Truth and DM:G, much less something like Mjolnir. And that's with all of the hit being useful; if it's not, then it's even worse. I'd skip it and go for those two instead if at all possible. If you're curious, the Mark of Supremacy is similar.

[Druid, 3.2] Don't get Idol of Mutilation for bears!

I alluded to this in my previous post on Emblem of Triumph gear choices, but I'll expand that a bit.

Idol of the Corruptor currently drops from General Vezax in Ulduar and will be obtainable for 19 emblems of conquest as soon as 3.2 comes out - and everything will drop emblems of conquest, at least. The proc is 153 agility for 12 seconds, and it has close to a 100% uptime since it can be refreshed with another mangle. If there is an internal cooldown it is less than 4.5 seconds.

The Idol of Mutilation costs 25 emblems of triumph, which are only obtainable from the Coliseum raids (10 and 25 man) and daily heroics. The proc rate for lacerate/swipe appears to be quite high and it seems to have close to a 100% uptime as well, and there appears to be no internal cooldown. The proc gives 200 dodge rating.

With the nerfs to agility and dodge rating, here's what each item does before diminishing returns factors in:

200 dodge rating / 45.2 = 4.42% dodge
153 agility * 1.1 (kings) * 1.08 (sotf + motw) = 181 effective agility /47.84 = 3.799% dodge

These are with the new 3.2 nerfed agility/dodge multipliers, in case you were wondering.

So the Mutilation idol gives more dodge %. That's not a big surprise, though with all the stat multipliers it's less than you might think. Even if the multipliers don't take into account, it's about 3.2% dodge.

However, that's not all that agility gives you. In this specific case, you're also gaining 362 armor and 2.2% crit. Is .6% dodge worth 362 armor and extra crit?

According to Rawr, nope. Even removing all threat value, the Idol of the Corruptor wins over the Idol of Terror by about 16%. If you factor in threat it improves even more. This is fairly intuitive, especially given what we know about fights in Ulduar.

Now, there's a wrinkle to this. Early in 3.2, it was possible to swap idols and have both procs up simultaneously. Later on in the PTR they added this confusing bit of verbiage:

Items with Triggered Effects: These items generally have cooldowns on how often they can be triggered. Those cooldowns are now triggered each time the item is equipped (example: A trinket has a 45-second cooldown on an effect triggered by player attacks; when a player equips that item, the effect will be unable to be triggered for the first 45 seconds it is worn).
The problem with this is that as far as I can tell, neither idol has an internal cooldown of any sort. Have they added one since? If they haven't and if this doesn't apply to things in a generic way (ie, giving everything a 5-second delay on equip regardless of ICD), this trick will still work. If you macro all your attacks to something like this:

/equip Idol of the Corruptor
/cast mangle

/equip Idol of Mutilation
/cast lacerate

You should be able to get both procs up all the time. Note that this is not tested on live, so don't take it as gospel; while I suspect that it does work, I can see it going the other way. If so, one of the better things you can do when threat isn't an issue is to use those macros - you'll gain 2-3% more dodge just from doing that.

If anyone else on the PTR has the emblems to test this out, I'd appreciate it; mine were wiped along with all the emblem gear I had bought.

Tuesday, July 28, 2009

[Druid, 3.2] Emblem gear for bears

UPDATED: Changed Idol of Mutilation's ranking to reflect that the Corruptor's ranking is higher and only costs 19 emblems of conquest.

With 3.2 just around the corner, it's time to start doing loot lists. 3.2 is going to introduce a huge amount of gear obtainable via heroics, so similar to what 2.4 was like I'm going to first go through the various emblem purchases and tell you what to prioritize so that you don't waste your hard-earned triumph badges on random crap you don't need. No matter how shiny it is.

First off, tier gear: you can only get the 10-man normal mode tier gear from badges, and for the most part it's not that spectacular. Furthermore the set bonus is basically irrelevant for bears at this time; the 2pT9 bonus is odd and not that useful, and the 4pT9 bonus is essentially as good as the 4pT7 bonus.

But let's get to the filthy lucre, shall we? Here's the rankings I've used to evaluate all of this. And the first item up for bids is:

  1. Duskstalker Shoulderpads (45 emblems). Remember, the Ulduar hardmode shoulders from Iron Council are better than this (and everything else in Tier 9), so if you have those - don't bother. And it turns out - don't bother with your tier in this slot either, as these are better than the equivalent T9.25 items. Both are pretty poorly itemized for bears, but these are slightly better - and unlike the T9.25 gear these don't require a Trophy of the Crusade to get them. They're also about 12 points higher than the best non-hard mode Ulduar item, and they're pretty cheap.
  2. Clutch of Fortification (35 emblems). The only reason that these are this low is simply because there isn't such a huge jump between this and the next items in Ulduar. But right now - this is the best tanking ring for bears in the game. I suspect that one of the heroic-mode items from the Valkyr will eclipse it, eventually, but even then it'll still be very competitive. It's about 9 points higher than Leviathan's Coil and 5 more than Fate's Clutch, if that helps. Personally, because my rings kinda suck? This'll be my second purchase after the Idol, assuming Vezax continues to not cooperate.
  3. Glyph of Indominability (5o emblems). This isn't in wowhead yet and might not be around any more (I'll check tonight) - but wow, if it is...amazing trinket. The only reason this is so low is that the cost is very high and it is situational. But wow, what a situation. This is essentially a defender's code that is double the effectiveness of its normal use. With it and defender's code you can have 3000 armor extra. That's pretty nice. Again, I suspect there are other better trinkets out there (like Juggernaut's Vitality), but it's great. If you have good trinkets already, I would wait on this. If you don't, this is an excellent place to start.
  4. Hood of Lethal Intent (75 emblems). This is more of a kitty item than a bear item. It's better than the ilvl 232 tier helm, but slightly worse than the equivalent ilvl Tier piece. It is about 12 points higher than the best stuff from Ulduar, so consider buying it and using it until you get your T9.25 item and then swap this with your cat gear.
  5. Idol of Mutilation (25 emblems). Oh, finally we get a tanking idol that's an upgrade from Idol of Terror. And the proc is great, even with the recent nerf to dodge rating. Testing has it up almost 100% of the time, making this about a 3% gain in avoidance almost permanently. This should be your first priority if you don't have Idol of the Corruptor; nothing gives quite so much a gain in value as this does, badge wise. Except as Gescom points out - Idol of the Corruptor is available in 3.2 as a conquest drop.

    Now there may be some odd things you can do if you have both; you can swap idols and potentially get both procs up at the same time in bear, which is pretty awesome for avoidance if not threat. This isn't quite clear given this patch note:

    Items with Triggered Effects: These items generally have cooldowns on how often they can be triggered. Those cooldowns are now triggered each time the item is equipped (example: A trinket has a 45-second cooldown on an effect triggered by player attacks; when a player equips that item, the effect will be unable to be triggered for the first 45 seconds it is worn).

    It did not appear that the Mutilator or the Corruptor idol had a cooldown internally at all, but it's possible they do. If so, there's nothing useful about the Mutilator idol relative to the Corruptor Idol for bears. I'll make a bigger post showing why, but keep in mind how well druids scale with agility and how many ways agility gets boosted compared to dodge, and you end up giving up very little. The difference in avoidance between the two is about .5% avoidance pre-diminishing returns, and for that .5% you get higher crit rating and about 300 armor. The corruptor wins out unless you must have as much avoidance as humanly possible. On a side note, realize that this is the best cat idol if you have to mangle (and may be the best cat idol period depending on the cooldown rate for shreds), so it'll be pretty useful for cats anyway.
  6. Dexterous Brightstone (35 emblems) - this is here for completeness; don't bother with it unless you really want something for cat and have horrible rings for your bear.
  7. Runetotem's Handgrips of conquest (30 emblems) - these are actually worse than the T8.25 gear, but if you didn't get lucky from Mimiron these may be something important to pick up. But it's a pretty small upgrade from even Naxx. Still, these are the best gloves available purely from badges, so if you need an upgrade, here ya go.
  8. Runetotem's Headguard of Conquest (50 emblems) - these are the best of the lot in terms of an upgrade from T8.25 (about 7 points), but because of the non-tier item that's available, these get something of a downgrade overall. If you can, save your emblems and get the Hood of Lethal Intent instead.
  9. Runetotem's Shoulderpads of Conquest (30 emblems) - an actual slight upgrade from the T8.25 item, but because of the Duskstalker shoulderpads above and because they're only a small upgrade, I put them a bit lower than the handguards. Definitely pick up the Duskstalker's first, if only because they're 7 points higher than these.
  10. Runetotem's Legguards of Conquest (50 emblems) - another T9.10 item worse than the T8.25 items. If you need an upgrade in the leg spot, skip these and use your conquest badges on Leggings of Wavering Shadow instead - those aren't the best things ever, but they're only slightly worse than buying the T9.10, and you won't be wasting your triumph emblems.
  11. Runetotem's rainments of conquest (50 emblems) - these are just barely an upgrade over T8.25. Unless you've seriously been unlucky with drops, don't get this. In fact, it's much better to simply buy the T8.25 via conquest badges and save your emblems of triumph, so don't buy this at all.

Friday, July 24, 2009

[Druid] Nerfing stamina the right way part 3 - unask the question

A continuation of the big long talk from yesterday.

First, I'd like to say that I really appreciate folks taking the time to look at their fellow tanks in their raiding guilds and do their own legwork on this. It was good to get largely reasonable confirmation that the numbers were correct for other people and that the relative numbers compared to other tanks were at least on par. It gave me some insight as to what is fair and reasonable. For instance - is it okay if druids are below warriors sometimes, depending on gear choices and profession choices? Must they be a bit higher?

But in the end, I think that this sort of stamina nerf would be fair and reasonable across the board, relative to tanks.

The question that wasn't asked but should always be is this: is it fun? Well, clearly being nerfed is never, ever fun. On the flip side, being sat because someone else happens to be playing an overpowered class isn't much fun either. So balance is somewhat important for a lot of people in raiding guilds fighting to get a spot in that guild.

But how many of those out there are there, really? How many people would be benched if (for example) paladins, not druids, had the 15% health lead?

It's not an easy thing to answer. With DPS, it's very simple: either you bring amazingly useful buffs for the rest of the DPS or the raid, or you put out great DPS. This is mitigated by things like 'do you stand in stuff' and 'do you loot whore', but for the most part, we know where DPS stand. If you buffed one class's DPS by 20% over others, tons of people would reroll to that class, and tons of people would welcome them with open arms into raiding.

With healing and tanking I think it comes much more down to the player, not the class. Healing has occasionally been measured by healing meters (which is stupid), but most of the time you bring the healer with the mechanical capabilities you need (Chain Heal, CoH, MC, whatever) or the better role to play (awesome tank heal, awesome movement, etc). It rarely comes down to raw healing output. It's not nearly as quantifiable.

And even less quantifiable is tanking. Tanking has a lot to do with odds and minimizing odds. It has much more to do with being in the right place at the right time, using your abilities the right way, and allowing others to do the right thing. Yes, you can look at health and armor and avoidance and make guesses as to how well a tank could do compared to another tank - and that does happen from time to time. But the bigger limiter isn't class; it's skill.

Now, with DKs a couple things happened. For starters, they were really stupidly overpowered. It wasn't a small degree, it was a huge one. Second, their abilities matched up really well to the encounters. But most importantly, every guild had a TON of DKs waiting in the wings. If you needed another tank, plenty of people would chime in and be happy to go to town. With Naxx, alts had gear, main specs had secondary pieces, etc. It was easy to find a DK with enough stuff ready to tank at least at a decent level, gear wise - and with their overpowered abilities and stats they could find success.

But is the same thing true for ferals? Not really. Having to start from level 1 on a reroll of the flavor of the month is a big difference than getting a ridiculously geared 55 that rolls through early content. Are there a lot of guilds out there with feral tanks waiting to burst on the scene? I don't think that's really true. I can't reasonably study spec/class populations in raids and who gets to do what; I'll leave that to the smarter folks like Flyv. But anecdotally, I hear a lot about not many ferals are seen tanking, or many have chosen to do DPS more than not. It's hard to find a feral tank on my server, and harder to find a good one. Anecdotally, I don't think the population of good feral tanks is particularly high.

So if there's no real danger of people rolling a new feral, and not a lot of guilds have ferals, is it okay if they're not balanced?

I used to think no - it's never okay that they're not balanced. It's not good for the game no matter what. But I talked it over with my wife, and she changed my mind.

The thing is, people do want to have fun playing their character. Some people have fun playing the underdog type - the one that is underpowered but they want to show that they're so awesome they can do well with it anyway. (this, I think, applies to a LOT of protection paladins out there, or did - I know it did for me when I was playing it). Others want to be creative and figure out new, better ways to do old things.

Others want to feel needed - that something they bring to the table isn't done by others. That makes them feel important .

And others simply want to feel like they're super powerful - not only relative to the content they're fighting, but relative to other players. They want to nuke everything and laugh. They want the big numbers, the big crits, the giant stamina. Heck, some don't even care if they're ridiculously awesome tanks; high stamina beats overall survivability, right? It certainly can when trying to find a PUG.

Everyone feels these to some degree or other, but the point is that all of these things exist for people. And I think for a lot of people, if you took away the health and armor advantages that druids have, they'd really cease to be fun. At that point, they wouldn't be overpowered relative to other tanks. They'd be the same - except they'd have worse cooldowns, a more boring rotation, no armor graphics, weird buggy behavior, odd itemization and no real unique abilities.

If you get rid of that, what's left? And it's not a matter of 'why would you bring one' - you'd bring one because they're a good tank, and good tanks are in short supply. No, it's a matter of 'why would you play one?' I think Blizzard did a really exemplary job of making the death knight not only viable but really fun to play; the resource system and the way the trees work along with the graphical awesomeness of the DK made for an excellent hook.

But I don't know whether bears really have those sorts of things.

I'm not saying that everyone would abandon ship if something like a nerf in health combined with a fixing of cooldowns happened. I know I wouldn't; I'd once again figure out how things worked and move on the best I could. Many others I know are the same; they're happy because they can MT anything and do it well, and they're no longer constrained by stupid mechanics requiring shields. Others are just happy they get to tank at all. Yet others really love playing druids because druids are awesome in their own right.

But I do think for many people it would make them change. It would remove the parts of being a druid tank that are fun for them - the big numbers, the feeling like a mighty meat shield. With that gone, they'd move to something that gave them that feeling in some other way.

So right now, I'm not really supporting any big nerf like this would be. Maybe next expansion bears can be revisited heavily, and when that happens things like making bears care more about druid abilities, rotations, interesting shifting mechanics, resource management, itemization and graphics can be fixed. And if bears need a stamina nerf now, perhaps a small one will work to do the trick and balance them reasonably.

But I don't think there's a need for it, not yet at least. Because sometimes, fair and balanced isn't what people want, and it's not what's actually fun.

Thursday, July 23, 2009

[Druid]Nerfing stamina the right way 2 - more examples

In the last post, I mentioned a way to make druid stamina scale at any level of gear in a reasonable way compared to other tanks that would require very few nerfs going forward. Basically, it uses the same idea that bear armor had and only uses the bear multiplier for stamina found on leather and leather alone. Everything else just gets the normal feral multipliers, same as cat form does currently.

And by far the biggest concern in the comments (both here and at the WoW forums) was that this was way too big a hit on the starting druid and the midrange druid, and that they'd be hurt the most. (the second was 'would this be fun and is it even necessary, which I'll address tomorrow).

Before I give the answer to this, I want you to think about this for a bit. For starters, I gave an example of a druid not in BiS (myself) and compared them to other tanks and how they were - and trust me, I'm no where near best in slot geared. That gave an idea of what the scaling would be like if you didn't go full stam on everything - it was almost 2.5k difference compared to the best in slot full stam gearing druid.

Now, think about what the actual change does. It emphasizes stamina on leather first and foremost - that's unchanged. It then nerfs stamina from gems, enchants, and buffs.

Lower level gear, by comparison, is not affected nearly as much. There aren't nearly as many gem slots as higher tier gear (point of fact, this is one of the biggest ways that items are an upgrade for bears right now - if they have another gem slot). If you're starting out, you're not going to have the best enchants in the universe. In a heroic, the only buff you're going to count on is mark of the wild.

In short, the worse-geared druid will be affected by it, but not nearly as much.

That's the theory. Let's look and see what it's like.

First up, a bear geared in some of the better gear that they can get that is not a BoP from heroic or above, is not crafted with Ulduar items, is not from a raid or is otherwise easy to get. Note that some of this is expensive - the trollwoven set is expensive, after all. But it's worth it, and it's reasonable otherwise to assume a tank would try and get some of these. The enchants are cheap and favor stamina to some degree simply because the agility enchants are expensive. No profession bonuses are used, and rare gems are used only in a couple places and aren't perfectly ideal. This also means that the total stamina multiplier doesn't include kings, so the total stamina is .1486x normal, and the non-bear multiplier is only 1.189.

Again, gear lists at the end.

slot leather nonleather gem enchant other
head 69
0 0
neck
49 0

shoulder 79
0 30
back
75 0

chest 169
72 8
bracers 61
24 6
weapon
117 0

hands 71
0 18
belt 72
0

legs 67
0 45
feet 61
0 18
ring1
34 12

ring2
51


trinket1
45


trinket2
0


fort




motw



51
base



121
food



40
total 649 371 108 125 212

Now, here's the best in slot values (just the totals)

slot leather nonleather gem enchant other
total 781 828 569 222 426


That's right. Compared to the best in slot stamina, the preheroic only loses about 132 stamina directly from leather. Part of that is using the Polar chestpiece - which is perfectly normal for early bears to do. Part of it is simple: there simply isn't that high of scaling between stamina pieces per itemization level.

But look at the gems! 461 stam just from gems. And the nonleather pieces are huge too - 457 stamina difference. And that's the part that's getting nerfed the most.

The final numbers for the preheroic geared bear with few enchants are thus: they start at 23.2k health. With the nerf, they lose 2420 health, bringing them to 20.6k health. That's a bit low, I think. One easy way to fix this would be to include the base stamina (121) as part of the bear multiplier. They'd then go to only losing 206 stamina total, or 2060 health and bringing that preheroic-geared bear to just about 21k.

Now a reasonable comeback here is that, much like the heavy clefthoof set from TBC, bears could be forced to depend on the Polar set simply because it gives way too much stamina. And that's fair. It may be reasonable to simply apply the stamina bonus on all gear but not enchants, gems and buffs. That would help quite a bit in the low-end case, still nerf those best in slot druids with massive enchant and gems, and help scaling otherwise. Just food for thought.

Okay, what about a more modestly geared druid than yours truly? I'll try this with one of the forum naysayers - Azshannya of Kirin Tor. They're modestly geared, only having a couple of pieces from Ulduar 25, a couple from U10, and the rest from Naxx and various other places. They've gone another route, choosing to maximize armor whenever possible. They start out at 44172 total health after buffs (again, 25-man). What do they get knocked down to?

slot leather nonleather gem enchant other
head 90
56 37
neck
88 0

shoulder 85
18 0
back
84 0

chest 100
48 0
bracers 57
0 40
weapon
150 0

hands 75
0 0
belt 75
24

legs 99
24 55
feet 57
0 22
ring1
94 0

ring2
84


trinket1
111


trinket2
0


fort



214
motw



51
base



121
food



40
total 638 611 170 154 426


Again, comparing to best in slot:
slot leather nonleather gem enchant other
total 781 828 569 222 426

The leather total doesn't increase that much. The gem gulf is still ridiculously huge though. Might be interesting to consider just nerfing gems, no?

However, the total comes to something a bit alarming - this druid would lose about the same as what I would - 4450 health. That's quite a bit, bringing them down to just above 40k health. Is that reasonable?

It might be. The thing is, this druid didn't gear for stamina only. They don't have a second stamina trinket, their professions aren't stamina-centric (at least one of them), they've enchanted for armor and avoidance over stamina in many places, etc. Whereas that comparable warrior or paladin is putting stamina in every single slot they possibly can. It might be the case that this is too much, too. Is it fair that warriors have to gear for nothing but stamina, but bears have the luxury of branching out? I don't know. It feels like this might be a bit too extreme to start with. Then again, one of the goals was to nerf bears so that they wouldn't have to be nerfed again.

On some of the commenters (Nandy and others) that mentioned that they're only 3k ahead of other tanks in their guild - I'm really curious now. My suspicion is that there are either buffs that are missing or that the gear isn't perfectly full - on stamina centric, or professions aren't. I could be wrong, in which case this would put that bear behind the warrior. But again - might not be so bad.

Anyway, tomorrow I'll talk about the fun factor of it a bit more - where the summary is "My wife is more often than not right".

Gear lists:


Preheroic Azshannya
Head Eviscerator's Facemask Weakness Spectralizers
Neck Torta's Oversized Choker Mark of the Unyielding
Shoulders Trollwoven Spaulders Valorous Dreamwalker Shoulderpads
Chest Polar Vest Tunic of Indulgence
Waist Trollwoven Girdle Belt of the Tortured
Legs Mind-Expanding Leggings Proto-hide Leggings
Feet Jormscale Footpads Dawnwalkers
Wrist Drake-Champion's Bracers Thrusting Bands
Hands Handwraps of Preserved History Dislocating Handguards
Finger1 Ring of Earthen Might The Leviathan's Coil
Finger2 Iceforged Battle Ring Gatekeeper
Trinket1 Goblin Rocket Launcher Essence of Gossamer
Trinket2 Scarab of Displacement Defender's Code
Back Durable Nerubhide Cape Cloak of the Shadowed Sun
MainHand Stave of Shrouded Mysteries Origin of Nightmares
Ranged Idol of Terror Idol of Terror

Wednesday, July 22, 2009

[Druid] Nerfing stamina for bears - the right way

I'm back to hiding behind spreadsheets!

On the official forums, there's a lot of talk about nerfing druids health once again now that DKs have been brought down to the level of mortals (or at least, undead mortals...err...).

And for the record: I agree with it. I've agreed with it for a while Given non-pvp gear/polar gear and comparing best in slot pieces with all tanks across the board, druids have about a 7k health lead over warriors, paladins, and any non-blood DK right now. That's before epic gems and before the DK nerfs as well, so if anything this gulf will only increase in the future. That's about 15% more health than other tanks. Now, part of that is that druid itemization is shockingly 'better' than a lot of plate tanking itemization; there simply aren't that many hardmode T8.5 items out there that are exceptional, and there are quite a few hardmode leather items that are. But it's a continued saga from WotLK release, and it'll continue.

The fact is that bears aren't quite on a level playing field. They haven't been for some time - basically, since crushing blows got removed. With that removal, the idea of bears being the high armor, high HP tank just doesn't make sense. It made sense originally in that bears were supposed to be this buff, hard-to-kill thing that took tons of damage too; that was the concept. Along the way, though, they got about the same avoidance as other tanks, higher health, and higher physical mitigation. They also got pretty decent magical mitigation too. And they ended up taking exactly the same hits without crushing blows.

But without crushing blows, why does a bear need higher health and armor than another tank? This might make sense if avoidance numbers were far apart from each other, but it turns out that this really isn't the case; avoidance is largely close to each other as tanks. So they're taking about the same incoming damage physically. And with Protector of the Pack, they're taking about the same magic damage.

And barring cooldown use, it doesn't make sense for one tank to simply be better in all ways than other tanks.

Now, you can't take cooldowns out of the equation entirely. That's not reasonable, especially with the design of so many fights being 'blow cooldowns in scary place and survive'. But balancing around lack of cooldowns with average mitigation doesn't make sense either; it means that either you're the best tank because cooldowns don't matter, or you're the worst because you do. There's no middle ground.

Part of the real problem here is how well bears scale with only a couple stats. I've talked about this before and predicted the actual gain bears would get, and have been largely accurate. The biggest deal here is, of course, how well bears scale with stamina. Specifically, for every point of stamina on gear, enchants, gems or buffs bears gain the following multipliers:

1.25 for bear form
1.1 for heart of the wild
1.1 for Blessing of Kings
1.06 for survival of the fittest
1.02 for improved mark of the wild

For a grand total of 1.63 stamina for every stamina point. Now, part of this is to make up for the deficit in stamina on leather vs. plate tanking gear; leather has about 25% less stamina than plate, so that makes sense. And most other tanks get close to the same kinds of multipliers other than bear form for their stamina; 1.1*1.06*1.02 is a 18% multiplier for base stamina, which is in line with other tanks to a large extent.

But what doesn't make sense is the bear multiplier applying to everything else. For example: a warrior gains about 1.16x stamina per point of stamina after all buffs. A druid will get 1.63 stamina for every point. So improved fort gives 214 stamina normally; for a warrior that will be 249 stamina. For a bear, that's 350 stamina - a difference of 101 stamina!

Just from fort.

Similarly, every solid sky sapphire gives a warrior about 26 stamina. It gives a bear 39 stamina.

Every leg enchant gives a warrior 64 stamina. It gives a bear 90.

Heart of iron gives a bear 265 stamina; it gives a warrior only 188.

These sorts of things are why bears get nerfed every patch. With that high of a multiplier, bears simply win more every time they get an upgrade relative to other tanks. Warriors have some compensation in that they have higher base stamina and they get things like a gun slot and typically more gems, but that's a static value; by comparison, bears simply have a huge multiplier to wield scarily away. And at some point that isn't going to be balanced.

Which requires a nerf - until the next time.

So let's try a thought experiment (with math!) What would happen if the bear stamina multiplier applied only to the stamina on leather, and only the base stamina? In other words, all stamina found on non-leather (cloaks, trinkets, rings), all gems, all enchants, all buffs - they're all going to only be given the non-bear multiplier, which is 1.1 (hotw) * 1.1(kings) *1.06(sotf) * 1.02 (motw) = 1.31, about.

What would that mean?

Let's take my current character in their balanced set as an example. This doesn't take into account sockets or anything like that, but it's otherwise correct. It doesn't stack stam exclusively, has a few shared slots with cat that use agility, has a mix of agility and stamina enchants when it made sense, and only uses one PvP piece (the bracers). I also have JC and LW for my buffs. A full list of the gear can be found at the end.

Here's the breakdown of how stam looks on that gear:
slot leather nonleather gem enchant other
head 99
73 37
neck
94


shoulder 67
24 30
back
84


chest 119
82 10
bracers 73

90
weapon
121


hands 93



belt 73
60

legs 114
48 55
feet 90



ring1
84


ring2
89


trinket1
162


trinket2




fort



214
motw



51
base



121
total
728 634 287 222 386


That's right. 728 stam from the leather only. 1529 stam from everything else.

With this setup we have roughly 3690 stamina currently, after all buffs save food. If we applied the bear multiplier to stamina only on leather, what would be the result?

3191 stamina.

In practical terms, this is a reduction of almost exactly 5k health in this situation. This bear would go from 48.7k to 43.7k, give or take a bit. Now, 43.7k sounds awfully low compared to some other tanks, I will admit - but that's just my general bear prejudice talking. As far as tanks in a similar gear set, it's almost precisely on the money; this is a tank without any hardmode gear, not fully T8-upgraded in all slots and only wearing one stamina trinket.

So that's...actually pretty close to reasonable. It's a big nerf, mind you, but the value is pretty good and it wouldn't require a lot of balancing relative to warriors (which I'm considering the benchmark right now). Let's take the next test case - the bear best in slot. This will be without any PvP gear or polar gear, and while it will gear for stamina heavily, it will not do so at the cost of juicy bonuses or socket bonuses. It will use two stamina trinkets though. Again, full list of gear at the bottom.

At the start, this bear would have 57.7k health. A similarly geared warrior has about 50.5k health.

slot leather nonleather gem enchant other
head 122
73 37
neck
96 47

shoulder 96
48 30
back
96 24

chest 127
72 10
bracers 61
24 90
weapon
129 36

hands 93
47

belt 65
72

legs 127
72 55
feet 90
30

ring1
96 24

ring2
94


trinket1
162


trinket2
155


fort



214
motw



51
base



121
food



40
total 781 828 569 222 426


In this example, the bear has 2826 stamina from all sources, of which 2045 is from non-leather and 781 is from leather. Currently that bear has 4621 stamina.

With bear only applying to leather? 3952, or a loss of 668 stamina, taking their health from 57.7k to 51.0k.

Which is again almost exactly the same as a similarly equipped best in slot warrior with a similar gearing strategy (it's about 1k higher). Which is just about perfect.

What's good about this strategy is that it's future-proof. Bears will scale a bit faster than warriors, but that's acceptable to blizzard (as shown by paladins in TBC) so long as it's not by more than 10% or so.

What's not good is that without that health lead, druids become much less interesting on certain fights because of their lack of potent cooldowns. So if this did get changed, I would hope that the cooldown situation would also be addressed.

Now you might say that you could get to this result in other ways - and you could, at least to start with. You can remove HotW's bonus, which would reduce health by about 3.5k to 4k. You could nerf SotF and MotW as well. The problem with all of these things is that it still leaves bears with a huge, huge stamina multiplier as a base. Sure, it's a nerf now, but it would mean that bears would still get 30 more stamina than a warrior for every 100 stamina they both get. Eventually, that will break. This way, it won't break until sometime well past the end of the expansion; it can't, since the numbers are essentially the same for all tanks.

Furthermore, it means a lot of degenerate gear choices go out the wayside. Gem slots are not the sole arbiter of whether gear is awesome or not. DPS jewelry becomes somewhat more usable.

It doesn't fix the PvP or polar set gear, and that is a legitimate problem. However, PvP gear only gives about a 2k health lead over non-PvP gear, if you use all of it. That's still within a 10% margin of error, and still better than what we have now.

It also doesn't do anything to fix armor. I think that the 4-5% difference in avoidance between the druids and the rest of the tanks can deal with the armor, honestly. 4% more mitigation vs 5% avoidance seems like a wash.

Now, I hear some druids really like having huge health pools. Okay. They also like having huge armor. Hmm. A bear can't have the same baseline mitigation as other tanks but have higher armor and higher stamina while maintaining similar avoidance levels. Either their avoidance would have to be nerfed, or their base mitigation would have to go away.

If you really want to have health and armor aplenty, I think the only way to fix that would be to remove protector of the pack. In that situation, druid health and health scaling would be unchanged (which would break druids later on in icecrown), but their incoming damage would be 12% higher than currently. They could even get a buff to armor to compensate for this. I don't think this is a good long-term design idea, but I can see the value in it; a lot of bears don't care about whether or not they're easier or harder to heal so long as they have tons and tons of stamina and armor. That would be one way to go about it.

Anyway, there's my idea for fixing the stamina issue so that it's no longer so insane for bears. I do expect some criticism here, but try and make it somewhat constructive if you can. Also, check out the gear choices and see if they make sense or not. Enchants were kept the same across both sets for ease of use.



My gear Best in Slot gear Enchants
Head Garona's Guise Conqueror's Nightsong Headguard Stalwart Prot
Neck Boundless Ambition Bronze Pendant of the Vanir
Shoulders Valorous Nightsong Shoulderpads Shoulderpads of the Intruder Great Gladiator
Chest Conqueror's Nightsong Raiments Embrace of the Gladiator Power Stats
Waist Death-warmed Belt Soul-Devouring Cinch Eternal Belt Buckle
Legs Valorous Nightsong Legguards Legguards of Cunning Deception Frosthide
Feet Runed Ironhide Boots Runed Ironhide Boots Superior Agi
Wrist Furious Gladiator's Armwraps of Triumph Solar Bindings Fur Stam lining
Hands Conqueror's Nightsong Handgrips Conqueror's Nightsong Handgrips Major Agi
Finger1 Gatekeeper Fate's Clutch
Finger2 Signet of Winter The Leviathan's Coil
Trinket1 Heart of Iron Heart of Iron
Trinket2 Darkmoon Card: Greatness Royal Seal of King Llane
Back Cloak of the Shadowed Sun Titanskin Cloak Major Agi
MainHand Twisted Visage Dark Edge of Depravity Mongoose
Ranged Idol of Terror Idol of the Corruptor

Monday, July 20, 2009

[General] 10-mans observations


I've been doing a lot more 10-mans as my schedule's been a bit more free this summer. Up until last week I'd never done all of Ulduar on Felhoof. I had two experiences that were about as wildly varied as could be.

The first was me getting in a bit late, in a group that needed a second tank. I thought it would be okay, and it looked like a solid group - but a number of the people in it decided to leave early, go AFK, or otherwise only stick around for gear on the boss they wanted and then leave. They've since been kicked from the guild, and good riddance. But these all conflated to a very difficult time of it - and it didn't help that the other group that was going at the same time was knocking out hard mode achievements left and right. We finished up without doing Yogg in two whole days, and only managed one hardmode achievement (XT).

It was honestly one of the most demoralizing experiences I had ever had playing. Not just wiping over and over to things we've done before simply because people were leaving and we were scrambling to find people, or people simply couldn't perform at the level we recruited them at - but knowing that it was doable by other people that I raid with...sigh.

The next week I got to go in on the 'hard mode' group's run because they needed a second well-geared tank for Algalon tries as well as finishing up their Glory of the Raider achievement; they had Vezax and Yogg hard mode to go. And sure enough, it was a breeze. One-shot virtually every hard mode in there that we tried (we didn't do Firefighter). We one-shot Freya+3 and did it so fast we got Conspeedatory. The only trouble we had was doing a side achievement on Thorim, and once we stopped trying for that we breezed through things too. We got to Vezax and two-shot him on hard, then got to Yogg and one-shot him on everyone's first try at that hard mode. We went back and did Ignis and Razorscale's achievements, and at the end of the night 7 people had their Rusted Protodrake.

Not me - I started the night only having Hodir and Kologarn done, and finished with only Mimiron to go. Pretty sweet.

I've talked about what makes runs so much better or worse, and in this I think it really came down to leadership and experience. I did the running of the group the first time, and I didn't and don't have a lot of experience on Ulduar. It was frustrating, and I didn't know the strats as well as I should have - and that was compounded by people flailing about and leaving left and right, babies crying, etc. The second group had done most of this many times before and had had practice at it, and the raid leader has been leading these sorts of things for months - and strats were ready, questions were answered quickly, and things went well. I've talked about the importance of leadership before, and it's definitely true now.

At least right now, I shouldn't be doing leading. Too many distractions, not enough knowledge of the fights from top to bottom. Ah well.

Anyway, onto the actual analysis. 10-man is very, very well done. FL+4 and XT are both easy enough hard modes that they're not really gating bosses. The actual fights of the hard modes are pretty nice with the exception of Hodir. Hodir is still a boring hard mode, largely. Thorim adds interesting tension, Freya is quite a different fight.

But Vezax and Yogg are masterful.

Vezax has a tension that isn't in almost any fight I've ever done. Maximizing performance when you can't really do anything is a bizarre mechanic, but one that works. Every move made requires a weighing - if I do this, will the raid wipe? Then you have the Animus spawn, and it goes insane - damage is everywhere, you only have a certain amount of time, and you've still got to watch out for most of the mechanics from before.

Yogg with fewer keepers is astoundingly different. I'm shocked at how important the sanity mechanic is when you can't refresh it. Hodir and Mimiron aren't really missed, but having no sanity is like night and day. It changes how you can DPS, it changes what you're going to do (sending in 4 DPS in portals so that you get fewer brain phases means no healer), and it drastically changes the tension in phase 3 for a DPSer. Instead of worrying that you won't make the enrage, you're worrying constantly whether you're going to take that last bit of damage that'll send you over the edge.

As a cat, what I ended up doing was being absurdly cautious and maximizing bleeds. I'd try and rake, shred shred rip and then rake - and then turn around immediately and await the next lunatic gaze. I didn't start this way, but my sanity was dropping precipitously (to a mighty 4!) and I couldn't afford anything else. It worked, too; I topped the meters with bleeds on yogg.

But the tension was palpable. It was a crazy fight because of that. Just brilliant. Kudos to Blizzard for adding such an interesting mechanic to an otherwise boring fight.

Which brings me to...Algalon. Algalon isn't at all hard. In our first hour of attempts we got him to mid-40s, and that was mostly just learning what to call out and how to move, and what the behaviors are. The only thing that's stopping us right now, and we're having a hard time with, is when to kite constellations into holes and which ones to leave up - but that's mostly because we didn't consider that so important. Once we iron that out a bit more, (likely making sure that no one is near the hole we'll take and kiting constellations into specific ones outside of that) I think it'll be a clean kill. It's disappointing; the actual zone is epic, beautiful, and the lore is great. The fight itself compared to Yogg-1 or Vezax or Freya or Mimiron is much less special.

It's a bummer. It really should be a super hard fight, but the only thing that stopped us was double zones, zones closing early, and melee being stupid about not moving out of Cosmic Smash.

If y'all have any tips, I'd love to hear 'em.

Anyway, will have the cat epic gem comparison later this week.