Short Version: Depends on crit chance.
Long Version: Basically at best you're looking at 77% uptime on Arcane Lightning, and that's with taking Arcane Wave and Shield as your slots (I think not taking shield as a hammer spec is suicidal, but if you can make it work without it you could improve the uptime by using ArBlast instead). Though again this assumes you're using shield immediately except when it will overlap, rather than as a defensive tool of any kind. Normally utilizing shield will drop it to about 60% uptime depending on length of the fight. Only taking wave would reduce it to 50%, and only taking shield to 20%. This means on average your crit dmg bonus from arcane lightning would be 7.7%, 6%, 5%, and 2% respectively.
The way damage calculation works, the closer you are to 100% crit chance the more your crit dmg modifier becomes an actual dmg modifier. I.e. at 100% crit chance, 10% extra crit dmg and 10% extra dmg are the same. Similarly though that's what makes direct dmg modifiers so strong, they are independant of crit chance and thus always apply to your base dmg (assuming they don't require some other condition to be met, like vital striking). Thus a dmg modifier is always an increase to your crit dmg, but not necessarily vice versa due to crit chance being a factor.
Assuming 7.7% from AL, 10% from the air points, and 4% extra chance I have it coming out ahead at about a 70% crit chance (not bad) in my very rough calculations. The deal here is they're very rough calculations and operate in a vacuum (i.e. I assume perfection). That's simply never the case, and even operating under those conditions the spec is not pulling ahead much, to the point that at 100% crit chance the direct dmg actually catches back up (it's only 1.2% behind) due to being unable to realize the benefit of the extra precision (it would logically peak around 96% crit chance then).
So again, I'd be willing to say that not having to manage AL at all, freeing up your slot skills, giving you a more versatile trait slot, and more toughness (which is kind of a double edged sword) is better. Though if you could provide the reddit link, I'd be interested to see their math as well.
The mantra basically works like this, you never are penalized for taking zerker stats (unless you hit crit cap but thats not feasible 99% of the time). Meanwhile I see bleed caps hit fairly often depending on group makeup. Honestly you're not casting churning earth for the bleeds anyways, you're doing it for the massive direct dmg and blast finisher (partially what makes Evasive Arcana so awesome is instant Churning Earth). The problem is most of your other attacks don't deal condition dmg, and the ones that do (Churning Earth being an exception here) deal very weak dmg. So it's not that taking condition dmg would be bad so much as it is that you're losing out on taking another far more useful dmg stat to pick it up (in this case crit dmg).
Also bear in mind, condition dmg is counter to precision, as conditions can not crit. Conditions still on a mob when it dies are in effect, wasted dmg as well since more direct dmg itemization would have ended the fight even faster. Honestly there's only two classes in the game at this point who should really bother with condition dmg (Engineer & Necro) and in both cases their main rotation is riddled with condition applications. Trust me man, I have an 80 engineer, Ele condition dmg can't hold a candle to what I can with her grenades.
Slotting survival stats tends to work along the same rational. Survival stats would have only been worthwhile if you hovered in the grey area that would cause you to go down if you didn't have it, yet barely stay up because you did. If you never got hit, they're a waste, and if you went down while still having them, they're also a waste. As for % dmg modifiers, yes they're worth it. Ele has the most in the game and that's partially why our base coefficients are so weak, we're balanced around them. Essentially it means that you need to take them to deal good dmg, but that if you're focused on a support role and not taking them, you won't still be dealing as much dmg as the other classes who are specced for dps.
Thanks so much for doing the math! I hated not being 100% sure on how that played out.
If you stagger Arcane wave and Blast, which happens in the normal rotation after the very first use, you easily maintain 100% uptime on Arcane Lightning.
I don't see any reason someone woul drop either of those moves in PVE, so I don't think about how things would play out without them-- but you raise a good point that I will update my guide to include-- if you aren't running 2x arcane utiltiies, don't go 20 air. Good catch and I might be messing some people up who need to run with more defense but still stick with 20 air.
Also the fact that they're so close means I will probably emphasize that you can choose based on what you prefer.
One last thing, on Churning Earth. Most of the groups I run in do not hit bleed cap because I do not run with condition damagers (or allow shortbow rangers, etc in my parties). So the stacks of Bleed from Churning Earth actually do a TON of damage, particularly when at 25 stacks of Might. I realized I removed the cond damage math from my OP and I can't find it anywhere, but the bleeds + burning added up to almost 1k DPS by themselves.
You are perfectly right on everything, just wanted to point out that it is actually very strong-- the sad thing is that adding condition damage simply doesn't increase your damage much beyond its high base strength.
Edited by Hamartia, 07 August 2013 - 04:15 PM.