[Relambrien]

Originally Posted by devonin

I'm not sure how my example of a healer getting a lag spike suggests that they are failing to play optimally. To me, optimal play in the case of a KT fight, is that anybody who has healing spells that are quick try and get a heal on the frost tombed player. The loss of DPS from the 2 seconds you aren't DPSing is lower than the loss of DPS caused by a DPS player DYING in the middle of the fight (Or a healer dying, or the off-tank(s) dying)

A healer getting a lag spike isn't really something that should happen often enough to worry about it. An occasional blip in their connection due to the ISP having issues isn't something that can be helped, and if it happens, then "Oh well, nothing to be done about that. Might as well try again." Hopefully, someone will notice their healer is having trouble before a person is in danger of dying, and another person can take over. However, people shouldn't be -looking- to take over. They should be -prepared- to, but shouldn't expect it. In the case that they have to, then yes, I agree that would be "playing optimally."

1/ There are many people who won't do this, they won't sacrifice their DPS to make sure decursing gets done

These people are idiots and probably fall into the "I want the biggest number I can because it makes me feel good" group. Either that or they have inordinate amounts of faith in their raid members.

2/ You -view- sacrificing your DPS to make sure the group gets decursed as part of being the best raid member you can be which puts you in a seperate group from the people I'm talking about.

Point taken.

To me the ideal way to "strive for perfection" would be to either figure out the numbers, or use their figured out numbers, and then use that knowledge to create your own spec. If you're going to just use their spec and rotation anyway because you "trust them to be correct" then what's the point in also reading about the underlying numbers? You're going to just use their conclusions too anyway.

Reading about the numbers allows me to develop my own ideas when I don't have the ability to create the optimal spec, i.e., virtually always. For instance, according to EJ the current ideal Mage spec is 19/52/0 with heroic Ulduar gear enchanted a certain way. Obviously, as I haven't even completed heroic Naxx yet, I have absolutely no way to reach that level. Instead, I use what EJ has written about hit rating vs. spell power vs. haste vs. crit to determine what is an upgrade for me and what isn't, and what spec to use when I have certain stats.

The reason I'm Arcane still even though Fire is the "superior" spec is because I don't have the hit rating for it. I only have 9.49% hit off of my gear, and thus I need the 6% from talents afforded to the Arcane spec (in addition to raid buffs from Draenei etc.) to reach the hit cap. And because I'm using Arcane, which I've read to be a less crit-dependent spec than Fire, I try to stack haste more than I try to stack crit, with spell power trumping them both. This is what I'm talking about when I say I try to understand the numbers as well as I can, even though I leave optimal spec-crafting to the experts.

My spec is not the ideal spec or rotation as specified by those websites, and yet I like to think I'm a pretty optimal tank at least in terms of the fact that if everybody else is doing their jobs passably well too (Like say, the biggest burst DPS in the raid isn't opening up on the guy I clearly haven't established full aggro on yet [Like guy number 4 or 5 in a big pull]) I do a damn good job building and holding aggro, stacking enough defense and avoidance to be surviviable for the healers to keep up with a minimum of effort, and yet here I am, passing on "better" pieces of gear, putting points in "bad" talents, and yet being diagnosed as "geared to be 2nd or 3rd tank" and having to either pull aggro off the main tank, or hold back to keep myself third on the list.

"Good," or even "excellent" are not the same as "optimal." I personally enjoy running with you because I know you're an excellent tank, and I gladly accept that you prefer your own developed ideas over reading cookie-cutter builds off of EJ or something. But when I say "optimal" I mean "the absolute best as determined by the game's system of calculation with respect to a person's job." With DPS it's easy, just pick whatever gear and spec allows you to put out the most damage. It's more difficult with tanks, as there has to be a balance between threat generation, maintaining aggro for long periods of time, and survival. Because of this, there can be different flavors of tank that still result in success (mitigation vs. avoidance, for example), but there's only one flavor that is considered "optimal." So while you may have great success in your particular flavor, your success potential is less than that of someone using the optimal flavor.

This is likely why you are relegated to off-tanking. When people who know the optimal spec see you in a non-optimal spec, they automatically assume you do not have the ability to perform as well. When in reality, you may perform better in your non-optimal spec than a person in the optimal spec. In this case, the difference between "optimal" and "non-optimal" is that the hypothetical "limit" to a non-optimal spec is less than that of the optimal spec. People may misinterpret this to believe that the non-optimal spec is always inferior to the optimal one.

But like I said, with dps, it's a lot simpler. There are no "flavors" in a ranged DPS class like mine, as we have only one job: put out as much damage as possible. Granted, crit vs. haste is something of a flavor, but because any change will either strictly improve or worsen our ability to do our job, it's different from the tank flavors in that sacrificing one facet of the job may improve another facet.

Because of this, barring significant gear differences, the optimal spec will generally outperform any non-optimal specs fairly significantly. But in my case, as mentioned above, the optimal spec will -not- do as well for me as my non-optimal Arcane spec, due to the fact that I don't have the gear to support it.