Infraction for Casstor: 2. All opinions should be stated as succinctly as possible.

Kaubel · 07/14/09, 4:20 PM

Post: Finalizing the Prot Warrior design
User: Casstor
Infraction: 2. All opinions should be stated as succinctly as possible.
Points: 1

Administrative Note:

Message to User:

Say it with one post, not three.

Original Post:

Originally Posted by Liar

And that is where avoidance just wins against hard hitting bosses. And I only did Algalon a couple of times but avoidance is definitely a factor in being able to heal a tank; generally, if you don't avoid every now and then you will die even with healers being able to pump out massive HP/s (just turn your boss to a hard hitting boss to see what I mean). Block in comparison? Healers will most likely not even be able to tell the difference between 10 unblocked hits in a row and 10 blocked hits. But they sure as hell can tell if you don't avoid some hits of the 10.

A couple points:
1. If your healers don't notice 1200 damage being blocked, why would they notice if you had 1200 more health (~103 more stamina). I doubt that when they top you off again, there is less than 1200 overhealing.

2. Here's my issue with avoidance -

I won't pretend to know the Algalon fight well, but - in general most of the tank deaths I've seen have been during a period where many attacks hit for their full value back-to-back before healers could react to the damage. In my experience, if healers have more than a few seconds to react to damage, they will be able to top me off. Now, consider a 30-second period in which you take 20 attacks (1.5 swing speed). This is 18 sets of 3 attacks in 6 second periods (the 13 sets are 1st 2nd 3rd, then 2nd 3rd 4th, etc.) The expected value of strings of 3 attacks, all of which hit, during this period is:

E = 18*p^3

where p = probability that the tank is hit. Now, assume your effective miss+dodge+parry chance is 55%. Then, p = .45, so E = 1.64. This means you should expect to be hit 3 consequtive times (i.e. 4.5 seconds of unavoided damage) 1.64 times every 30 seconds. Now, assume this you increase this chance to 60% (i.e. ~250 parry/dodge rating, depending on DR). Then, the number of sets of 3 consequtive attacks you take during the 30 second period decreases to 1.152. But, the point is, you should still expect there to be a 3-attack window where your avoidance stats are not providing any protection. If your healers can heal through this anyway, what did you just spend the 250 parry/dodge rating on?

Note: The number of consecutive attacks you should expect to take without avoiding damage is calculable, given avoidance values, swing timer, and fight length. For long periods, you end up taking much more than 3 attacks in a row without getting a dodge/parry/miss, even with as much as 70% dodge+parry+miss. Also, if your healers can react within every 5-second timeframe of spike damage, the benefit of quantifying expected values for the number of 10-second time intervals where avoidance stats do not contribute only becomes as meaningful as the possibility of your healers going and staying oom.

Originally Posted by Liar

To sum it up: Real avoidance beats BV/BR in a normal tanking set and it's not worth comparing them in any other situation because of the trade offs you will have to make to emulate Paladin block.

Warrior block is different than pally block, ours gives us a cooldown, there's gives them a passive EH bump. We shouldn't seek to emulate what they do - we should seek to do what we do - and in the most effective manner possible. By grabbing up shield block rating, we're only amplifying a rather small percentage of our actual blocked amount. This is why block rating is terrible for us.

For the purpose of having said it -

There are plenty of people here who have far more tanking experience and are in far more competitive guilds than mine. Your experience is not insignificant, and if my numbers aren't measuring up to what experience has shown, I guess I am looking to find where my assumptions deviate from the realities others are experiencing. Are my hit size values too low? Is there a flaw in my theory? Am I applying some bonus multiplicatively where it should be additive? I am not above admitting where I'm wrong about SBV's value as a stat. But if I am wrong, I'd like to get a notion of the correct calculation.