If I remember correctly, it checks on the fly. Whilst mucking about on beta, both [Idol of the Unseen Moon] procs and Owlkin Frenzy procs increased damage while mobs were hitting me. With my amazing 980ish spellpower in my Balance set at 70 (which, incidentally, is actually my healing gear for the most part) I was getting 110-140 ticks of Thorns dependent on buffs, with full Brambles. If they fix this to work properly (i.e. take the casters SP, not the users SP) then it could be quite a nice extra bit of threat/damage on AoE mobs, especially combined with Retribution Aura.

I don't think it's over budget, just that itemization points seem to be allocated disproportionately :

Enraged Feral Staff:
((1092/14)^1.7095+(45)^1.7095+(56)^1.7095+(117*2/3)^1.7095)^(1/1.7095)*1.8 + 0.75 = 265

Origin of Nightmares
((714/14)^1.7095+(77)^1.7095+(100)^1.7095+(150*2/3)^1.7095)^(1/1.7095)*1.2 + 26 = 308

(not my math).

As of right now on the beta bleeds and dots aren't taking rogues out of stealth anymore, although they still take druids out, so having bleeds up on rogues will not make a difference anymore. I am not sure if this is just a bug or a design change, but it has been there for at least two patches.

There's no way that's intended. They already have the best tools in game to remove dots from themselves (Cloak of Shadow and Hunger for Blood), so there's no reason that they shouldn't be knocked out of stealth from a dot.

Uhh... what? As far as I know, swipe does dmg x bear threat multipler, on each target it hits, regardless of how many that is. Is this wrong?

This is not wrong. Not sure where he got that from since the only threat that is even split is healing threat.

Extend your quadratic function up to the 300-400 ilvl range and tell me how intuitive it will be to assign tier item levels if you want to make each tier 13% of an upgrade. I have no idea why the heck you're choosing to pursue this line of debate.

Yeah, Armor is something a druid gets a -lot- out of in terms of item budget : effectiveness. It's one of the only stats in the game that Blizzard has to keep an extra watchful eye out for in terms of itemization. For example most other offensive/defensive stats 'handle themselves' in terms of effectiveness regardless of how badly you mangle an item's stat allocation. Armor on the other hand has to be strictly controlled. In TBC you could see how they had the wrong and right ideas in various points of the development cycle.

The obvious 'wrong idea' would be early on, when we had the god-like green quest reward set which put us roughly at 29-30k armor, vs T4 which (at that time) had no bonus armor. There was a clear disconnect between two itemization sub-groups and the class designers as to what should have been done for druid armor levels.

The 'right idea' would be how the tier sets (+ sunwell tank leather) had their bonus armor rationed out. You can see that although the ratios of the other stats changed from tier to tier (most notably Str:Agi) bonus armor went up smoothly.

Questionable would have been the progression of armor on the rings obtainable at various points of progression, as well as the pretty-much static values on weapons (Only Hyjal Staff had more armor than the rest, and was otherwise horribly itemized as a tank staff anyway)

In terms of how WotLK is looking, they seem to be getting more of the 'right idea', for example looking at rings - dungeon blues while levelling up are in the 200-300 range, 80 blues are around 300-400, epics going higher. Ditto cloaks, necklaces, and even trinket slots. The Enraged Feral Staff seems to be an anomaly.

This is pretty much as anecdotal as anecdotal gets, but at the moment ferals make the best Hateful Strike tanks for Patchwerk for pretty obvious reasons. Napkin math off the top of my head says that druids start losing out on DTPS at higher gear levels, but would continue to maintain a massive HP advantage as is, assuming the current itemization trends scale upward. Not to mention Ancestral Fort/Inspiration/LoH soft capping will hit druids sooner than lower armor classes. Much less distinct a difference than in TBC but there you have it.

Has anyone in beta noticed whether the "chaining" effect on swipe is still there? In other words, on live I notice that when picking the three targets for swipe, often a caster or other mob that is outside of the normal melee range will be hit as well - I'm a little curious how infinite target swipe will interact with the ability to CC in close quarters.

Someone posted a page or so back that the initial swipe target has a 5 yard range, but the extra targets have an 8 yard range, and that this is the current mechanic on live.

I plot log(fap) against level and fit a least-squares line; I plot sqrt(fap) against level and fit a least squares line. I assume the error in the FAP is 1 due to rounding. Then chi^2 for the log fit is about 87 while chi^2 for the sqrt fit is about 11000. Neither fit works, but the log fit is much closer (for 6 degrees of freedom, P(chi^2 > 87) ~ 10^-16 while P(chi^2 > 11000) ~ 0). If I push the error in the FAP up to 5 (assume the itemization team took some liberties), then chi^2 for the log fit is about 3.5 while chi^2 for the sqrt fit is about 437. In this case the log fit is plausible while the sqrt fit still isn't (P(chi^2 > 3.5) ~ 75% while P(chi^2 > 437) ~ 10^-91).

In summary, the assumption that log(fap) ~ level (i.e. fap ~ e^level) yields a reasonable fit while the assumption that sqrt(fap) ~ level (i.e. fap ~ level^2) does not.

I thought Swipe didn't get affected by the bear threat multiplier, so it was basically threat caused by damage. Is that not the case?