I've been trying to decide to what extent, if any, the Bear Form stamina modifier, Heart of the Wild, and the buffed Survival of the Fittest compensate for the problem of our gear's scaling versus a warrior's. Without itemization for +armor on feral gear, the bear form AC modifier becomes simply a compensation for the fact that we don't wear plate. The fact that bears and warriors have relatively similar AC on live (as in previous posts) confirms this.

That leaves us with four tanking stats at our disposal (stamina, agi, dodge rating, and defense), and defense is so laughably bad a stat without the need for crit immunity that I'm tempted not to include it in the list at all. Itemizing for dodge rating hasn't happened in the past, and anyway it's worse than agi.

According to WoWWiki, the item budget formula is
Suppose we have an idealized tanking item for each class that has exactly equal "item budget's worth" of each tank stat.

The druid one might have:
60 agi
90 sta(=60 "points")

The warrior one has stats distributed between 5 of their stats, say
35 def
35 block
35 strength
35 dodge
52 stamina(= 35 "points")

These two totals result in roughly the same item budget for the item, i90, if I've done my napkin-math correctly. That would result in the druid item having 120 item points' worth of stats on it, and the warrior's has 175. That ignores all the % scaling talents that bears have.

Applying SotF and the bear form modifier, 60 agi becomes 63.6 points (60*1.06), and 90 stamina becomes 98.58 points (90*1.06*1.55)*0.67 for the item budget factor. So, then we effectively have 162.2 item points' worth of stats.

Doing the same for warrior's talents, they get 5% more stamina and 10% more strength from Vitality, so they have 38.5 effective points from that 35 strength, and 36.4 from stamina, for a total of a 179.9 points. We still lose.

However, I think it's unrealistic that a warrior item would include so little stamina vs. avoidance as in my example--they need to end up with roughly the same HP pool as the bear, right? So maybe their item has:
20 strength
20 defense
20 dodge
20 block
110 stamina(=73.3 "points")
(I'm rounding a bit sloppily so it isn't exactly the same item budget, but you can see where I'm going.)

That's 77 effective budget points of stamina and 22 of strength, totaling only 159 points for the warrior.

In that more realistic scenario, we come out ahead in terms of the "effective" points gained. These are very coarse examples, but it's not unreasonable to think Blizzard could tweak items such that they have the same "effective" tanking budgets for bears, warriors, and paladins given the existing class mechanics.

How those "effective budget points" translate into time-to-live is complicated (as outlined in Tossk's analysis), but at least we're still getting our money's worth of item budget if the warrior has any HP to speak of. (Of course I'm discounting the diminishing returns on dodge, which will hurt us more than warriors because we have more of our eggs in that basket, but that's another topic.)

The problem right now is, looking at the Naxx feral set, it really looks like cat gear to me--so much of the item budget is spent on DPS-only stats like crit rating and AP. Okay, these buff our threat, but they don't help us out with our survivability. Compare with the warrior tank set, which has four or five of the stats on the item devoted to survivability. (To say nothing of the annoyance of getting AP instead of strength, more less-than-ideal budgeting...)

I guess in the end I really don't understand why Blizzard seems to suddenly hate the notion of itemizing specifically for ferals, even outside taking away our "niche" as the high-armor tank. Blue posts on druids keep including things like "but then you can't share gear with other classes." Well... so what? Introducing all the specifically feral items they did over the course of BC (with nice distributions of AC, agi, strength, stamina, and a little intellect for good measure) worked out great in my opinion. They turned out to be viable for both bear and cat. T6 is close to perfect in that regard, with its awesome set bonuses. Why ruin a good thing? Why try to shove furry animal-shaped pegs into square warrior holes? (Or, for that matter, pointy rogue-shaped ones?)

Regardless of the answer, I see a problem with the question. Stamina can only compensate for gear scaling if you're concerned about 'how long can I live without any heals?' The more important question is 'how much damage do I take?' Outside of soloing, the % of damage you take is more important than how long you live without heals.

That said... Toskk did a pretty good evaluation of how much damage ferals take in level 80 blue gear.

WoW Forums -> The State of Feral Tanks (8885) by Toskk

The important things he found:

In good Wotlk blues against 100% physical damage...
We have slightly more armor, probably not as large of a difference as before.
We have about 7% less avoidance.
We have about 900 more health.
The damage taken by a druid is about 10 more dps from the same mob.
The druid lives about .6s econds longer without heals.
(And proving what we already knew)
We scale better with armor.
We scale better with stamina.
We scale worse with defense.

This is a good evaluation because it covered both time to live and have much damage we take compared to other tanks in the same level of gear. I was honestly surprised to find out we're so close at level 80.

The problem is that this evaluation didn't consider how we scale with improvements in gear. A similar analysis using all of the new purples would be quite useful in letting us know if we scale with gear as well as the other tanks.

Personally, I'm still concerned because we care only about armor, agi and dodge for damage reduction. 2 of those stats are each mainly (only?) available on half our gear? So, our damage reduction is based on basically 2 stats: armor + (whatever stat gives dodge on that half of our gear). They can balance it with those (at least for reasonable range of gear). But 2 stats with high multipliers is just bad design for balancing and itemization compared to several stats with smaller multipliers. Some avoidance or mitigation from ap (and therefore str) would make things easier to balance, even if it was small, and even if our other stats had to be slightly less valuable for total damage reduction.

Stamina can't be totally ignored. It allows us to survive spike damage. It reduced overheals. But it can't make up for avoidance and mitigation.

I see what you're saying, Tappin, and I agree. What I was trying to illustrate, though, is that warriors have to spend more of their item budget on the necessary stamina than we do (to get to the "minimum required" HP pool), so they actually have less of their item budget left over to dump into their 7 different mitigation/avoidance stats. Our bear form modifier isn't a HP buff from that perspective; it's an item budget coupon. We are still handicapped by only having dodge to work with, but it's not as extreme a scaling problem as it might at first seem in terms of the gear itself. The problem is that Blizzard isn't giving us purely tanking items.

Really, our extra AC was a substitute for all that missing block and parry (and the chance to be missed from defense we didn't stack).

I am intrigued by the thought of AP giving us some form of mitigation-equivalent. Since stacking more and more dodge runs us into diminishing returns, maybe it could somehow add chance to be missed? That's one avoidance stat we do have at our disposal, even if Blizzard doesn't want us to be armor capped anymore, but we can't gear for it effectively at the moment because defense rating is so weak and crappy item-budget wise.

Another notion is to make defense give ferals some bigger benefit to miss and dodge, to compensate for the block, parry, and crit immunity we don't get from it, and to work that in somehow with the DPS stats. Or a "feral xyz" stat that was a DPS stat in cat form, but the good part of defense in bear... but that goes the feral AP route, which they seem to dislike.

Even from an item budget, you kind of have to look at the bigger picture.

Look at the naxx gear. Stats that are entirely worthless for survivability:
Warrior: Hit
Druid: AP, Crit, Haste, Ignore Armor

Then look at shared items like rings, neck, back:
Warrior: None
Druids: Str, and the block and parry item budget related to defense

We waste a lot of item budget already. Our huge multipliers may or may not make up for it (in blues? pretty close. In early purples? In late purples?) We won't know unless we look at an entire set and all of the survivability stats/mechanics/talents.

This idea also makes sense to me. But, it only covers our shared gear and enchants, while ap (str) covers all of our slots, making it easier to balance? Before we got the 2 new talents, I also suggested giving us NI in bear form, which would be better as we (and our healers) geared up and give us some tanking flavor.

However they do it, as long as our % of damage taken is comparable (not necessarily equal) in early blues into deep purples, I'll be fine with their solution. I have no idea if that's true, but it seems ok in early blues so far.

It seems to me that all the dps stats we have available share a common property; they all increase our rage generation. If there was some way of converting rage into avoidance or mitigation (e.g. 40 rage - guarantee the next 3 attacks on you deal 20% less damage) then that would solve the problem.

That returns to one of our core issues. Lack of an activated defensive ability. We've gained last stand in the form of berserk, but compared to the other 3 tanking classes who all have a 50% damage reduction ability on a short cooldown (5 minutes or less) along with other activated abilities, (shield block, holy shield, etc) we are rather lacking in this regard.

I see a lot of people leaving out improved leader of the pack in WLK and I think that's a terrible mistake. In sunwell it provides 2-4% of the entire raid's healing for two talent points. Now it effects all groups and not just the druid group, so that number should increase.

As a healer there's nothing I can spend points on that would increase the entire raid's healing by 4% for just two talent points.

I think it should be in every tanking build because it not only provides healing for you, it also heals a lot of raid members which means you get more attention in raid healing.

I've looked at WWS for about 6 2-4 hour sunwell/bt raids now. I've found LotP doing about 1.5% - 3% of total healing per feral. But looking at each person, I also found this to be consistently about 70% overheal. So, I think it's safe to say the for effective heals, LotP makes up about 1-2% of the raid's effective healing. This might be slightly worse because it doesn't consider overheals is causes others to make: you take some damage, someone HotS or starts to heal you, you proc unexpectedly, they overheal some.

Obviously if you have raid wide effect, it will hit more (maybe 10-15 since it doesn't help casters?). That makes it closer to 2-6% of effective heals if no other ferals with iLotp are on the raid.

Still, I agree with you about getting it for a tanking build. I'm not sure about a heavy dps build. If they ever make it a Hot or a damage shield, I'd even go so far as to say every feral (who isn't guaranteed to have another feral on the raid) should pick it up. Right now I'm having trouble fitting it into a 'usually tank/offtank but some bosses I'll dps' build, but I do want it in there.

Seems like a pretty good build. As much as I'd like to pick up master shapeshifter, I agree with your choice of not picking it up.

The choice to leave out savage fury is interesting. It's only an issue if you're having threat problems, though. But if you are, 20% more to mangle and maul is a pretty big chunk of threat.

Also, I'm not sure about primal tenacity for pve tanking. If you have instincts, I believe that covers ae stuns/fears and that the 2 don't stack (could be wrong, just something I read). If it's single target fear, mobs stick to you now, and you probably have fear ward, tremor totem and/or trinket ready if it's important. For stuns... I dunno, I don't think I've lost aggro or died that often because I got stunned by a single target stun.

I'm have a tough time chosing between that and those 2 points in iLotP or infected wounds or dropping a couple of kitty dps-crucial talents like KotJ or shredding attacks for my 'mostly tank' build.

Maybe you'll get expertise capped and that will free up 2 more points? There is a ton of expertise on our naxx set already.

Infected wounds: we need to see if the effect on mmo is a typo. I can't imagine they want -50% attack speed. We also need to see if it starts sticking to more bosses. If it's 20% on 2 hits and sticks to bosses, I'll be picking it up. There will probably be times where others forget to thunderclap or 5/10 mans where you're the only attack debuffer?

King of the jungle: seems pretty important for dps'ing (about 16% more energy moves available?). The benefit for tanking seems marginal.

Rend and tear: savage fury would be a larger bump in bear threat. For cat, 10% to shred is about 3% more damage overall? Right now, FB as part of a sustained dps rotation doesn't look very good, especially if it takes 10 points to make it good. But, that could change.

WWS by default shows effective healing and excludes overheal, so the 1.5-3% is correct for effective healing (unless there's something I've missed but I don't think so).

70% overheal still isn't that bad either. A typical healer will be at about 50%.

iLotP is of course useful and a good talent but whether it is more beneficial/useful than, say, infected wounds is going to depend on your raid makeup.

my bear build will be WorldofWarcraft.com -> Info -> Classes -> Druid -> Talent Calculator

with the last two points being prioritized to infected wounds and then to iLotP if another raid member will be doing the boss attack speed debuffing. But really if you're the 'main tank' odds are probably fairly high you'll be responsible for the boss debuff in which case IW > iLotP (imo).