Mages

The Akil'zon fight has two major sources of damage: Static Discharge and the soaring eagles. The bit where you get thrown up in the air is annoying but causes no damage if you have your Light Feathers handy. And if you're getting hit by the storm, you're doing it wrong.

Static Discharge does 3500 damage on its initial hit. However, with three points in Playing With Fire, that's 3605 damage for me. The soaring eagles hit for 1600-1700 with my armor.

I look at the Akil'zon fight as having three phases:

1. Before the first storm, I DPS Akil'zon.
2. After the first storm, I play flak battery and attack the eagles.
3. When Akil'zon gets to 10% or so, I switch back to him.

For phase 1, Molten Armor would seem to be optimal. This is one of the two damage periods I get so I should make the most of it. Further, Mage Armor only increases the resistance by 18. Seems hardly worth it. [Read more about @title]

I expect this'll become the first post of a short series on Zul'Aman. But it might not.

Up until Akil'zon, there's nothing particularly "interesting" for a mage. Pre-Nalorakk trash is pretty normal and the pre-Akil'zon gauntlet isn't that odd either. The main concern about the gauntlet is running out of mana.

Nalorakk is cake. Because of the threat games needed for the tanks to swap between phases, you shouldn't need to worry about aggro after the first transition. Ever. It becomes a fight along the lines of Ragnaros, Patchwerk, Murmur, Lurker Below, etc. You just sit there and shoot.

Akil'zon is more interesting because of the execution elements. If you're lucky, you get assigned to DPS Akil'zon, in which case it's a standard boss fight aside from dealing with the storm. If you're unlucky, like me, you get assigned to anti-eagle duty once they spawn. For me, this tends to be a lot of tab-targeting and praying that Fireball actually works. [Read more about @title]

In the quest of determining how to mathematically model spell haste, I've come to realize a handful of things:

1. Spell haste only provides a DPS buff when considering a duration-based model. In my standard mana-based model (damage based on exhausting the mana pool), spell haste breaks even at best.

2. There's a handful of things where the mathematical modeling gets a little weird. For example, the casting time for a spell can never drop below 1.5s. I'm not sure how to handle bounds like that in a mathematical formula other than saying something like t = blah, blah, >= 1.5.

3. You know, using a spreadsheet is probably just a lot simpler. Vontre says he's updated his spreadsheet.

Spell haste seems to be the new interesting stat for gear. Just about all of the new 2.3 heroic and ZA mage gear has spell haste on it (and so do some of the healing pieces). I'm just not sure if it's worth it.

I would like to work out the mechanics of spell haste so that I can come up with a damage comparison formula. However, this'll take some work since I will need to determine a formula for average MPS. Fun.

zurmagus' DPS simulator looks like an interesting project. I'm just not entirely sure how to make use of it in place of Vontre's spreadsheet.

Another issue I see with the current implementation is that zurmagus has written it in C++ so it's command-line and therefore not really usable. Maybe I'll try to convert it into a web-based project in my copious amounts of spare time.

This site has a ranking of trinkets for mages based on build: http://wiki.shadowpriest.com/index.php?title=SimulationCraft/Trinkets/Mage. Kinda makes me feel better about having spent the 1000 gold or so it cost me to assemble a Blessings Deck for Phaedra's Darmoon Card: Crusade trinket.

(Link found in post 2903 of the TBC Mage Theorycrafting thread on Elitist Jerks.)

Edit: Of course, then there's further discussion about these results later on in the thread. I suggest reading it.