You gain much less dodge from agi as today, so if you gem your whole equip with agi gems you gain approx 13x10agi=130agi=~3,6% dodge. Gemming stamina gives you 13x15=195 stamina=around 3k life. For me this is a clear decision towards gemming stamina. With level 80 gems this is 7,2% dodge vs. 6k life, but in the beginning our avoidance will suck with around 30-35% so 6k life will be clearly better. Perhaps if the gear gets better and we are at an bt/hyjal level gemming agi becomes better, because overall avoidance is better. But in the start at lvl 80 gemming stamina as tank will be the route to go.

13 X 10 dodge rating= 130 dodge rating= 6.87% dodge, suddenly the gap between your choice between avoidance and hp just got closer. I'm not arguing for agility at all at level 70 since it's no longer a winner, but at 80 it passes dodge rating point for point for avoidance, and that's all I am stating.

To me stamina is near-worthless past a certain point for survival (I use 21k for Brutallus), and mitigation and avoidance is much more important to me. In addition, although we lose dodge from agility, we gain 5% base dodge (I believe, correct me if I'm wrong), 6% from NR (talent), and the removal of Sunwell Radiance is worth 25% avoidance. So strictly speaking in sunwell, unless we lose more than 36% dodge in regular tanking gear in sunwell, stacking avoidance will still net better results than stacking stamina. Now I cannot say for sure for my M'uru avoidance tanking gear, but I definitely won't lose 36% dodge in my standard tanking gear come 3.0.

Dodge rating passes agility at every level in terms of avoidance per point. The relation always remains the same, regardless of level.

Apparently M'uru would be different anyway, since the adds don't suffer(/enjoy) the aura, so there the hit to your avoidance will be much larger.

I do agree with you about the avoidance vs stamina though. At some point, stacking stamina just doesn't bring anything more. You want to survive a worst case scenario and once that's done, the next level to hit would be one more boss attack, which tends to be a wee bit more than you could gain from socketing additional stamina. It remains to be seen what exactly that perfect stamina level is, but with the reported health lead bears have, we may not even need to focus on stamina all that much.

Good point about Sunwell Radiance. Regardless, in my heavily avoidance-stacked, 15sta's still beat 10agi's.

One other thing to note... Defense CAN surpass Dodge/Agi in total avoidance gain, due to how the DR works. DR for Dodge and DR for Miss are calculated separately. If you already have a very high dodge, such that you're losing alot to DR, adding a def gem CAN provide more avoidance due to miss being almost unaffected by DR.

Again, these are pre-alpha calculations, not guaranteed to be correct in any way yet.

It looks like on Beta, SR and other +%AP buff don't stack with each other.

So, it is still a myth to say what blizzard will address this problem.

1. Make it intended, it doesn't stack.
2. Make it stack, but nerf SR accordingly, make the overall effect doesn't change after stack.
3. Make it stack, and it is 140% * 110%
4. Make it stack, and it is 140% + 110%

There is nothing wrong with it at all, maybe you misunderstood me.
I was merely stating that since the relation always the same regardless of level so you don't have to do the math again for level 80, as the result is the same to level 70.

If 1 agi > 1 dodgerating after kings at level 80, the same will apply at level 70.

Let me do the math about end-game damage reduction.

Formulas and numbers are from here
http://elitistjerks.com/f31/t29453-c...gs_level_80_a/

+25 defense skills = 1% miss/dodge/parry
+250 defense skills = 10% miss/dodge/parry = 30% pure damage avoidance before diminish returns for war/pal/dk.

You need 1229.62450975 defense ratings at level 80, because 4.918498039 defense rating = 1 defense skill.

For same item value points of the defense rating, druids get 1229.62450975 agility. After blessing of the kings, it is 1352.586960725 agi. Which equals 32.4620870548% dodges for druid, since 41.6666667 agi = 1% dodge at level 80.

Let's assume that miss rate shares same DR formula as dodge, although it might be a little better than dodge, if a boss has overpower ability or tank need to turn its back to the boss for some reasons.

War/Pal/DK has
x'(miss/dodge) = cx/(kc + x) = 88.129021*10 /( 88.129021*0.9560 + 10) = 9.350425913177
x'(parry) = cx/(kc + x) = 47.003525*10 /( 47.003525*0.9560 + 10) = 8.5561497238594

Total Avoidance after DR x' = 9.350425913177 * 2 + 8.5561497238594 = 27.25700155

Druid has
x'(dodge) = cx/(kc + x) = 116.890707*32.4620870548 /( 116.890707*0.9720 + 32.4620870548) = 9.350425913177 = 25.9756304

Almost 1.28% avoidance short from 30%ish avoidance.a

Edit for my math if not wrong, druid need 5kish agi, and from that point, defense rating is better than agi in term of avoidance for druids.

The numbers are a bit skewed since you're comparing defense and agility separately when it shares the same diminishing returns on dodge.

The break-even point is actually closer to 3.5k agility. You'll have about 81% dodge pre-DR and around 49% after DR (from agility only). The DRs start affecting dodge so much that the miss gain from defense increases avoidance more. However, agility will still be more valuable far past that point for the dps value.

Aye, I just noticed it happen when I activated avoidance trinkets.

First, the math here is not to find break-even point for defense and agility. I didn't write math for break-even part yet. I was comparing if what happened if warriors stack defense rating and druids stack agi. My math showed that druid might be lack of avoidance at very end-game, although we need real numbers of T9 to have an exact idea if druids are underpower or not.

Second, I gave coarse math for break even point. From above math, you can see agility give roughly same benefit for druids as defense rating to warriors. that's for same item value points, 1% miss/dodge/parry from defense rating ~= 3% dodges from agi, pre-DR. For druids only, defense rating gives druid 1% miss and 1% dodge when agi gives 3% dodge. Because defense and agi share DR on dodge part, we can see, when we stack enough dodge, at one moment, additional agi gives x=3%, x'=1.5% as dodge, for same item points, defense would give druid 1% miss(not DR), 0.5% dodge(same 50% penalty on DR on dodge). That's the break even point. When will you have 50% penalty on dodge?

x' = cx/(ck + x), when x= 120.1636468, x' = 60.0818234, you got half gain from pre-DR value.

to have 120.1636468% dodge, you need 5006.8186 agi. it might be lower if you have kings+SotF, which is 4294.01254 agi.

Either way, I think it is big enough for us to consider defense rating until maybe T12.

IMO it's never a good idea to evaluate stats outside the context of the entire system. I simply took my spreadsheet with full epic tank gear and tried inflating agility to 5000. I then evaluated the updated stats for Mitigation and it showed defense much more valuable. If I reduce dodge rating and defense down to zero and increase agility to around 3300, defense and agility are roughly equal in Mitigation value. However, this is also slightly skewed as we won't have only agility on our gear anyway.

When considering non-DR avoidance, more weight is placed on increasing miss than when just viewing the increase from gear on the two stats alone. The comparison really needs to be taken in relation to total avoidance and damage reduction. I haven't done the full math, just some trial and error stats. Regardless, I don't see us hitting this much agility in WotLK either so it's probably not worth the time getting exact values.

Edit: To clarify, my point is that you cannot judge the value of defense based on a comparison to agility alone.

I don`t have access atm to an calculator, but just from my mind the value of +hit and +exp affects white dmg the most because we get energy back in case of a failed yellow attack. If i remember correctly there was also a blue statement about shred (in general they where speaking about attacks that require you to attack from behind) not being able to be dodged in wotlk. So the value of hit and exp for dmg can not be that great. For threat in bear these two are for sure the best values you can get.

The question i have is if its not better to gem all gems with armor pen in the dmg gear, because it still gets better the more you have.

EDIT: fixed an error with FB

Regarding the discussion with Tang about the best cycle, I've tested his ideas and... well we both have partial right.
I'll explain myself:
basically I used a 4SR/2 x finisher cycle, but Tang said that it's better to free up savage roar doing it when expire and use always RIP, well we were both wrong. At least at the gear level I'm testing (full t7.10, full heroic gear, full epic gems, full raid buffed, enchanted, food, scrolls and glyphs) it seams that the best results are with a free SR cycle and free Finisher cycle doing the finishers at 5 CP and doing SR at 4 (but it's really marginal)

Basically this is the cycle (prioritized):

1) keep SR always up (do it if you have 4 or 5 cps or simply do it, it's basically the same)
2) keep mangle always up
3) keep rake always up
4) use shred at energy = shred+rake-10-lag
5) if you have 5 cps use your finisher (RIP or FB) depending on RIP ticks left, buff-debuff, spec and energy.

2SR/5RIP cycle:
DPS: 4853

Free-SR/5RIP cycle:
DPS: 4838

4SR/2 x 5cp finisher cycle:
DPS: 5075

Free-SR/choce finisher cycle:
DPS: 4991

Half-Free-SR (wait 4 cp) /choce finisher cycle:
DPS: 4995


The cycle is pretty complex:
basically we need to watch and refresh 3 debuffs (SR, Mangle, Rake), watch for other debuff (like trauma and bleeds), watch at RIP ticking and energy to choce between rip and FB, watch at our energy to makimize the usage of Tiger's Fury.

His mistake was that he took the value where the entire amount of dodge was diminished to half. At that point additional dodge is diminished to far below half. So the amount of dodge for which additional dodge is diminished to half is considerably lower.

But we can't really compare dodge and defense (=dodge + miss) until we know the cap value for miss. My guess is that miss cap will be lower than the dodge cap.

A warrior who wants to stack avoidance will also use items with defense, dodge and parry on the same item instead of only defense (the itemization formula favors that).

Also there is the OoC free energy procs. This is really too much.

I realize that this is a very non-EJ attitude, but here goes.

Considering the massive amount of buffs/debuffs/durations/cooldowns and energy levels that we need to keep track of when DPSing as cat: what is the trade off by just using FB instead of Rip?
It has been calculated some pages back and as I read it the difference wasn't very big.

I am wondering if the DPS lost from trying (and failing) to manage the optimal rotation is larger than settling with a less complicated rotation: Keep SR up and then FB?

Personally, I'd love to get rid of Rip from my rotation and just rely on Rake for the bleed. The great thing would be that an infinitely complicated rotation really seperates the wheat from the chaff.

It is safe to assume that Deep Wounds will always be up and as such, its not necessary to concern yourself that Rake must be up at all times for FB.

Using only FB is 5026 dps vs. 5108 with my setup (1.6% difference), rake should always be up not only for the bleep but because is our best DPE move.

I really like having a very complicated cycle, I don't want to be a "one button" class, but what I really hope is that Blizzard will not balance Feral druid DPS on that "maximum efficiency" cycle because I think there will be a very little percent of Feral druid comunity able to keep track of that cycle. I really hope that they balance us on a "less perfect" cycle in a way that the average player have "on par" dps and if you are pro you have an "extra".

It's not completely a matter of just player skill either. The more complex a cycle, the easier it is to get penalized by mobility fights of any kind.

With all the smaller guilds that will be focusing on 10 man content and the certainty of a proliferation of death knights, it is not safe to assume this.

Hmm, read back a few pages... regarding gemming, I'm probably going to go bonkers with hit gems and put them in every yellow socket. Growl glyph is nice, but there's already too many to choose from. Cycles are going to be complex enough that missing is going to be quite a significant hit. Way more than it used to be.