Unless you take into account that arpen can apparently go over the cap.

http://elitistjerks.com/f31/t29453-c...8/#post1210542

Hadn't read that post yet, thanks

I would think that DMC: Greatness would still be the best trinket, even with ulduar loot...

The ArPen trinkets are both better by a good deal due to the average amount given is much higher and that ArPen is better than Agi.

My list was of item values at the time. There were more items discovered recently, and certain items will be re budgeted or fixed. Blizzard has done this with several pieces of loot so far in WOTLK including the heroic COS quest gloves.
The newest feral weapon that I've seen is Dark Edge of Depravity (not on wowhead yet). It is better than [Furious Gladiator's Greatstaff] even if it is not re budgeted.

Does this take into account how those procs might stack? I'm fairly certain if both go off simultaneously you can over cap arp which would severely decrease the value of them combined.

(ignore the Warlock bit on the left; I rerolled feral druid, 'Fite').

Just to add to the confusion on this topic -- I regemmed everything to armor pen and ended up with 415 arpen rating (33.70% reduction). This reduced my ap by about 600 from full agi gemming. I put on as much arpen as I could get, breaking T7 bonus in the process on the theory, promoted here, that maintaining rip is less important than weaving in FB.

Gear is moderately close to Naxx25 BiS, plus some pieces from U10.

I then went and blew up the boss target dummy for a few hours, and my dps gradually settled in at around 3850, motw only. Rip had a few breaks owing to some clearcasting droughts, and I weaved in a decent number of bites when possible. Priority was SR, Rip5, Rake, Bite5, Shred.

I usually do a little more dps -- over 4000 -- with pure agi gemming.

And indeed, despite people with very similar gear reporting that Rawr is giving them arpen > agi, even with full arpen gear Rawr is telling me that agi is better -- which would contradict the notion offered here that arpen has increasing returns, if Rawr is right, which is always pretty questionable.

Given that disparity, it seems likely that it's not so cut and dried that arpen is better. It may be better for certain gear combinations, but for mine, which is probably pretty reasonable for 75% or more of the druids reading this thread, it's questionable. Couple that with most of the fights in Ulduar involving movement requirements, and at least for me it looks probable that optimizing rip/rake (and swipe!) over FB and shred seems to have better play in real world scenarios.

This may be an example of where target dummy testing won't give you an accurate picture. In a raid setting you'll have increased combo point generation (through increased crit) and increased energy regen (through haste and hit, increasing your OOC procs), not to mention more clearcasting procs once you get 2T8. These changes will have little effect on your bleed dps since our rotation revolves around maximizing Rake and Rip uptimes, but will greatly increase the number of Shreds and FBs you're able to pull off. I'd suggest comparing combatlog parses or Recount from raids before your gear changes and after to get a better idea of what's really happened to your dps.

Wait, how does that work, did GC's post show that you couldn't pass more than ~80% arp on a boss mob; as in, you cant even drop 100% armor altogether or did i misinterpret the post?

The equation GC gave us was nothing to do with caps. It really just determines the effectiveness of ArPen based on the amount of armor the target has (and your level). They might have intended for there to be a cap when your paper doll says 100% but there isn't.

The biggest factors in a raid is CP rate and bleed uptime. Crit and haste (OoC) directly influences cycles and can mean the difference between being able to FB or not. You can also sacrifice bleed uptime for FBs. When you are on a training dummy your crit will be lower and your ability usage will be more stable. This will most likely lead to higher bleed uptime. The increasing value of ArP in relation to Agi is directly dependent on increasing the frequency of Shreds and FBs. Additionally, some abilities scale better with AP which means proportion of damage contribution from different abilities changes in raid.

Yes you can draw some conclusion from dummy testing, but it requires huge sample sizes and data extrapolation.