I fully agree, but what I was referring to were the melee hits which are very weak. In such a case avoidance isn't as important as EH. In a similar fight where resistance gear isn't a viable option the same gearing strategy will stand.

I get this one all the time due to the 30 Stam 15 Resil on shoulders:

Does resilience still reduce raid boss chance to critically hit you?

I have no answer for this because I remember it used to but I remember something about a change.

Would this be a good edition to the warrior FAQ?

I personally prefer to keep Imp Revenge for a few reasons. First, I'm not really a fan of Imp Disciplines unless I'm glyphing Shield Wall/Last Stand and I never do. Also, I'm not worried about threat or doing a tiny bit more DPS with Cruelty (also I prefer Focused Rage). Another reason I take Imp Revenge is because there are still many boss adds in ICC that can be stunned. Fanatics and Adherents on Lady Deathwhisper and many of the adds on Valithria Dreamwalker (esp the Abom) as well as many of the adds on LK can all be stunned. These adds do a lot of damage so a 3 second stun is pretty useful. I also keep it because I'm the 3rd tank for my guild so I have to run with a Prot offspec for much of the week. Fury is awful in BGs w/o a healer so when I'm going solo I go prot, and imp revenge is op in PvP.

Resilience does work on raid bosses and reduces crit ADDITEVELY with defense.

One thing to note for sake of completeness (hardly useful, although sometimes handy when you solo old content etc).

Resilience is counted BEFORE defense. Which means that with resilience on your gear IF you spec blood craze it will proc regardless of your defense (converted crits still proc blood craze if they were subjected to reslience reduction).

This is just plainly wrong. Bosses does so much damage these days that if a healer decides to throw a heal elsewhere the tank might very well die before his next heal can land. In the end it's pretty much irrelevant because Beacon of Light will keep hitting the tank no matter where the paladin aims his heal.

Avoidance isn't worthless, but that's because it can save you in those situations where healers make mistakes or other circumstances (such as raid needing heals) stops some heals from coming in. This is a randomness that you cannot depend on, but 20% chance to avoid a wipe is still better than 0%.

How is it "wrong" to say that between HoTs, shields, and intermittent heals, avoidance is more valuable than 330 more hp or a resillience enchant which you derive little to no value from? The whole problem with a lot of tank "theorycraft" these days is that all anyone is focused on is burst and a flawed EH model. The burst in most fights in WoTLK is almost always a known value on a known timer, you almost always deal with it using cooldowns in WoTLK, and it's completely unsurvivable with 2000 more hp (or less since that's what we're talking about with some of the gem and enchant choices).

I don't know how your healing assignments look, but there are people in our raid group who spam the tank all fight long and people who help out on the tank when needed (either because you're in period where there's an expected frenzy or because the tank dipped low when a healer had to interrupt a cast). There's maintenance healing and there's extra healing. I don't think anyone is saying you can survive long periods of time without heals. What I'm saying is that you will always have some, but if you don't need that swing healer to help out, that frees up other heals for other people. You can't just look at 100% to 0%, there's a lot of situations where you go 100% -> 45% -> 75% -> 20% -> 80%. Mitigating and avoiding damage in there makes it easier for them to get to a point where they can comfortably renew buffs, judgements, etc. because you just got topped off and they know it will take 2 hits to kill you.

When discussing the usefulness of avoidance, it is wrong just to focus on the final hit (that some might consider as the one that kills the tank). You must calculate the series of event leading to the death of a tank. Mostly the tank would have taken several blows over a short timeframe, any of which could have been avoided.

If you only consider the final hit on a tank with 50% avoidance, the chance to survive that hit is higly based on RNG. however if you consider the chance of any of, lets say 4 hit beeing avoided, that would lead to a survival chance of 93.8% (or a 6,2% risk of dying). To halve the risk of dying in that senario, you would have to increase your avoidance to 58% (as opposede to 75% to dubble tha chance of avoiding just the final hit).

Revenge is now VERY viable again as of 3.3.3. Thanks Bliz for making me change my bars twice in 2 months :P

Revenge > Shield Slam > Shockwave > Devastate is now our new single target priority list.


P.S Imp Revenge talent is fantastic

Revenge and Shield Slam are actually very close, so I'm not sure I'd rank Revenge above SS in priority. For one, SS has a higher crit rate via talents, and secondly if shield block is up SS is going to hit much harder than revenge. Even assuming it's not up though, I'd argue the 15 percent higher crit chance definitely outweights Revenge, especially given a 2 piece t10 set bonus.

I'll admit I have actually run a TPS parser to test this though because I didn't think revenge had come quite far enough to outweigh the points I made above. I can try to do it and get back to this thread though.

For Shield Slam to have better threat than Revenge you must be at or very near a T10 gear level with 2pT10. The choice to put Revenge ahead of Shield Slam was made under the assumption that most Warriors using this guide would not have 2pT10. Shield Block does push Shield Slam above Revenge at all reasonable gear levels, though the difference is fairly small at the lower end.

That said, I've revised the threat rotation answer with those caveats.

Could anybody try to explain why the suggestet spec uses anger management 1/1 (1 rage every 3 sec) instead of improved beserker rage 1/2 (effectively 1 rage every 2 seconds if used on cd + possibility to better bypass rage starving spikes) ?