Safeguard 2/2 is 30% damage reduction. So it's 30% damage reduction 20% of the time, which is fairly overpowered. But my point of getting gibbed still stands. I didn't even consider Council P3, or Pit Lord Argoloth in my usage of Safeguard, but they are still very relevant to maximising its usage. I'm sure we'll see Safeguard come into more use in later tiers, but right now, I would still only pick it for a couple of fights (like the examples in my previous post). I would definitely refrain from using it on ODS as both tanks are taking near constant damage unless they're casting and not melee-ing. But used on a non-tank to counter Magmatron's single target flamethrower would be the only safe usage of that in the encounter, methinks. I'd wait a tier or two and see if the encounters allow a more free usage of it.

My apologies for the mistake. Indeed the maximum potential would be 6% overall damage reduction, when considering mana usage on the tank.

As for ODS, all 4 of the bosses stop their melee damage frequently to cast as you've already mentioned. The intervalls aren't too far off 30 seconds so intervening the other that at that point isn't dangerous at all. All in all it probably won't make or break any of the current fights but I wouldn't dare to say that investing the points in victory rush possibilities is a better choice.

I have a question about blood craze, i am really not sure it is a viable survie talent. 3.5 seems extremely low and most of all the proc is totally random. I actually made some test it can proc almost 2 time on the row or not procing for 30 sec. Of course i am aware it depend on how often we get hits. But after read some post above about improve VR i am tempted to put the point there instead of blood craze.

Exactly, I think people are looking at these things in too much of a vacuum. There's a lot of stuff to consider, but one of the most important things is that as a tank, there will always be some form of HoTs and healing effects being placed on you that require you being sub 100% HP to be effective. That simple fact alone means that a tank who (just as a completely random example) takes 90% less damage at all times is far more efficient to heal than one with 90% avoidance, but 10% to get smacked.

While one could argue that avoidance streaks let the healers catch up on raid members, in practice it doesn't really work like that...I likely have a tank Beaconed so any avoidance leads to wasted healing potential even when I'm healing the raid, Druids don't really move their Lifeblooms around so those are being wasted while you sit at 100%, just a couple examples. My point is that irregardless of math telling you that various Mastery levels optimize your total avoidance over time, in the real world you're ideally trying to just smooth out the damage, because that is what makes it the easiest to effectively heal you. I'd say the correct way to play a tank is to make it the easiest for the healers to keep you alive, not to min/max some % because a spreadsheet says it averages out to less damage over time. Mastery accomplishes the former, avoidance the latter.

From anecdotal personal experience, Warrior tanks are the #1 easiest to heal for me. They more or less take SOME damage all the time, but it's very smooth and reduced which allows me to get a LOT out of Beacon of Light's free healing...it rarely goes to waste on Warrior. Conversely, I find Bear tanks to be the hardest to heal. Lots of my Beacon heals go to waste due to avoidance streaks and then I'm forced to use my big, expensive heal to get them back to reasonable health when they do take some hits. Even if over time, the Warrior averaged out to take 1% more damage than the Druid...the style of mitigation increases my healing efficiency by a LOT more than 1%.

As far as reforging goes, the most gains overall to both threat build and survivability that I have seen is when following the
Mastery>Parry>Dodge stat priority (this stat priority list is only in regards to reforging as reforging for more stam is impossible, a decent health pool should be maintained but not at the sacrifice of the avoidance stats. The dual cut gems ie Mastery/Stam etc cuts of gems should be used to achieve this). The best practice for reforging is to focus more towards the parry and mastery stats. So for example if a piece of your tanking set has mastery and parry already on it then no reforging is necessary. Where as if you are wearing a piece with secondary stats Mastery and Dodge or Parry and Dodge then reforge as follows.

Mastery & Dodge on gear = reforge dodge to parry as this will help with threat gen and dodge is the weakest stat for threat gen when combined with its use in avoidance. As well as this will help with Hold the Line procs.

Parry & Dodge on gear = you should consider reforging the dodge to either mastery or to expertise depending on your level of gear and if you are or are not expertise capped at your current gear level.

The options your have for reforging should be balanced with the stats you are prioritizing in gems as well. Anyone that is using straight cuts to certain stats should use reforging to make up for stats they are not gemming for. In the Cata expansion its no longer the clear cookie cutter list of reforging and gemming options to follow. A healthy balance to all of your stats will make you more adaptable to fill the role of Heroic, Raid(main) and Raid(off) tanks.

I am a tank in a 10 man raid guild. Here is an observation I have made that only applies to 10 mans. This is due to 25 mans running many more interrupters than a 10 man most likely would.

Foolishly, I was reforging Dodge into Parry, Parry into Mastery, etc. Last night I was trying to do interrupts on Omnotron and Halfus to find my hit rating really hurt my raid. I would miss an interrupt and big damage hit the raid group. Upon reforging to max hit, I found that over all I lost only 1% pure evasion (dodge/parry) and 3% Shield block. However, I found that I gained 7% hit. Honestly, reforging right now seems to be more beneficially used on hit than on defensive abilities.

Honestly, if you are running a strict 10 man comp, where you may not have that extra interrupt, consider reforging for hit rather than defensive stats.

Unfortunately, I greatly disagree. Lets take the Omnotron fight for example. Arcanotron has a spell that is vital to interrupt. In my 10 man comp the reliable kicks with low cooldowns we have are from the Feral Tank, Prot tank (me), a DPS DK and a Resto Shaman. Other than that we run a mage and boomkin - both of which are extremely less than ideal due to the long cooldowns on both interrupts.

In this scenario we can only really rely on our 3 kicks, due to the healer may be busy and has no hit. However, this is a two tank fight, and in many cases it is safer to tank apart rather than on top of each other. This leaves 2 kicks for Arcanotron (dk and tank). This means that hit cap is extremely important, because if Arcanotron gets a cast of while standing in a puddle (which is entirely possible as he has a chance at casting the puddle then an immediate arcane bolt) then you have 1 dead raid member in a 10 man fight.

Halfus is another important fight, as healing at this point in the game sucks. If Halfus gets a cast off on his aoe it really sets you back.

Interrupting the ice guy on Ascendency Counsil by the tank helps a ton too.

So really, I must respectfully disagree with you with this. Remember, I only lost 1-2% evasion when reforging to max hit. In reality by doing so I am reducing much more damage in these particular fights than 1-2% evasion would.

After encountering this situation with Archanotron's interrupts (as he was powering down from 50-0) when our Elemental Shaman was absent and couldn't get every one, we tried solo tanking it, and successfully, too. It's definitely not a viable strategy for 25 man or 10 Heroic, but if your interrupts are lacking and you've got an evenly spread melee:ranged/healers then it can be quite easy. My healers even commented that I, the single tank, was easier to heal than two-tanking it with a very geared bear. The shields only last for ~10 seconds, in that time its ridiculously easy to stop cleaves and enables melee to focus-interrupt. Or, you could two-tank them together for the same effect.

Not saying it works for everyone, but this is what my 10 man did to overcome it.

Since parried or dodged attacks are fully mitigated, where as blocked attacks are only reduced by an average of 0,3x(1+0.01x %critical block chances), I think it is incorrect to write mastery > Dodge/parry

I can't see this happen before 30% critical block chances and a dodge or parry chances are high enough so that diminishing returns make ~300 dodge/parry rating = +1% dodge/parry chances, which is not necessary the case.

.
If I compare 2 rotations, at the start of the same fight.

First rotation : Rend => TC => SW => gcd => gcd => TC

We'll pick for threats per second (= total threat / Cooldown) the followings :
Rend : 2 142
TC : 2 415
SW : 1 524

The opportunity cost of the first rend is (per target) :
1.5 * 2 415 + 3* 1 524 = 8 194 threat.

Second rotation : TC => SW => gcd => Rend => TC (rend is delayed) :
the opportunity cost of the first TS is (per target) :
1.5* 1 524 + 6*2 142= 15 138

The difference beetween the two rotations, is ~7 000 threat.
So even if it is a bit better (in terms of threat) not to delay the first rend, the overall difference is about the quarter of the threat of a devaste which is not a "dramatic" difference.

The main reason why mastery is more valuable is because you get a lot more block % per mastery rating than you get of dodge or parry per rating point.

Also, dodge and parry are affected by diminishing returns, while mastery isn't. This means that mastery gets you closer to unhittable and smooths out your damage intake.

For instance, I currently have 1097 dodge rating, which is giving me 6.21% dodge. My parry rating of 1913 is translating into 10.10% parry. Meanwhile, my 2273 mastery rating is giving me 19.02% block and critical block.

That's actually surprisingly close to twice the benefit.

I made a small list of current warrior maintanks. It comes in handy at times.

Kungen - Tarilya - Grafarion - Toffel - Xav - Sco
* list is in no specific order

Some of the maintanks preferred gemming for mastery while others still go all out on stamina.
Gemming for mastery seems to decrease the damage you get while gemming for stamina increases your effective health.
But since the mana pool of healers has become an issue mastery will trump stamina if you have enough health to survive all of the boss attacks.

I'm interested in the post Mokkhyr made (#16), discussing whats currently the most feasible meta of Eternal Shadowspirit Diamond and Austere Shadowspirit Diamond.

Apparently the block value meta only buffs the block value to 31,05% and not 31,5% (0,3*1,05). Currently we're sitting at somewhere around 36000 armor and somewhere close or semi-close to 60% block depending on gemming.

I'm interested in why most people (not all) seem to pick the 2% armor meta, when the block meta seems to be superior already now at current gear levels.

In my case, the Austere was easier to come by. It's still early in Cata so it can be difficult to find a JC with a specific pattern. There wasn't a single Eternal on the AH any time that I looked.
It seems to me that at this point in Cata, a lot of people are gemming and enchanting with whatever is most conveniently available, knowing that much of their gear will be replaced in the short-term, especially if they still have a blue helmet.

I agree with you that the Eternal seems much better, but I haven't done the math for myself.

To all of the mastery "nay-sayers," do the math before you start complaining that it can't compete with dodge and parry. Diminishing returns are a big deal and significantly devalue dodge and parry rating even if you don't make a big effort to stack them. I'm stacking mastery at all costs and only getting parry and dodge where mastery isn't available, and I still am getting worse returns from dodge and parry because of DR.

I made a spreadsheet to model average damage mitigation and the variance in mitigated damage. I should note that I'm wearing mostly heroic level gear and I'm reforging and gemming mastery. With this gear, dropping 100 points in dodge/parry rating an increasing mastery by 100 has a negligible effect on my average mitigation (.01% or something) but the standard deviation changes by a much more significant amount.

The conclusion is that for a typical heroic geared tank, mastery is roughly equivalent to dodge or parry for providing average mitigation, but it does so with far less variation, making the job easier on a healer. It's obvious that mastery is the best rating stat for mitigation.

If you are reforging and gemming for dodge/parry because it "feels" right, you are in the wrong place. We don't decide things by feeling and intuition here. Trust those of us who have done the math, or do it yourself.

I see no discussion on where the soft cap for mastery is. Because of critical block, I don't believe there will be a hard cap. There is a theoretical point where it might be more efficient to start reforging your stats back into dodge.

I made a macro to calculate my current %hittable and the amount of mastery needed to become unhittable.

With SB up, I am 7.71% hittable, ~926 mastery from unhittable. This should be more than attainable.
With SB down, I am 32.71% hittable, ~3927 mastery from unhittable. Not sure if this will be attainable, but we have a few teirs of gear coming so who knows?

When I become unhittable with SB up, does it become more efficient to reforge back into avoidance (dodge specifically)? Does SB's cooldown make it better not to rely on it and OK to waste the 'overblock'? Or does the critical block aspect of mastery still make it a clear choice (does it outweigh the dodge diminished returns)?

I need one of those statistics math monkeys to crunch the theoretical damage numbers.

The macro: (for those who wish to play with it themselves)

Shyrian, there is no 'overblock' from SB.

In fact, from looking roughly at the possible caps, Shield Block's caveat for when it brings you past unhittable leads me to believe there won't be a legitimately negative cap for stacking mastery until the excess crit block from unhittable shield block brings you past 100% crit block.

I doubt this will be reasonably attainable as that kind of mitigation would be almost as bad as the invulnerable block sets they did away with with the block overhaul.

I see. Theoretically then with hold the line and shield block rolling up the full 25%, 65% crit block would be the true mastery soft cap (~6363 mastery).

Might be lower than you think. Take a look at your shield block BUFF next time you pop it.

Accounting for Shield Block - your Mastery cap is dependant upon your total Avoidance. As your Avoidance goes up, your Mastery cap goes down.

6363 Mastery Rating would get you Mastery capped if you could somehow lower your Dodge, Parry and chance to be missed all to zero.

My understanding is that Critical Block is a 2nd roll that happens when you successfully block.

What is the reason for not tanking with a fast weapon? I thought a faster weapon would simply mean more Heroic strikes/ Cleaves. I didn't realize there was any downside.

This used to be the case when heroic strike was a spammable, on-next-hit attack.

Now it has a cooldown.