I'd go with #2
tho I haven't tested this following piece of advice I believe it to be true (and you can quote/credit me for this):
damage = x * (str/100 + .95) * (EE/100 + 1) * (1+ (%chance to crit * crit multiplier bonus))

that part is true, what I propose later about the dex might be false as I will point out if theres ever the chance it might
the last term being dependant on dex
optimal results will be obtained with a maximization of the product of all involved variables
in this case EE, str, dex
EE you can't really play around with (unless you have an alternative weapon with str and a little bit less EE)
which leaves str and dex
to maximize their part of the product you would want them to be as close to equal as possible (basic calc, even geometry can show this)
now since I don't know the exact formula for the crit and crit multiplier I can't definitively tell you that there isn't actually a point where you should stop pumping into itbut I can tell you that even though it has diminishing returns these may be compensated for by the fact that it is a multiplicative bonus
example of that (entirely made up numbers) ((even contrived phoney example if you'd like to call it that))
100 points to allocate between dex and str on a character starting at 0 dex 5str
25dex = 33%crit with 2*bonus
50dex = 50%chance to crit with 3*damage bonus
100dex = 66%chance with 4*bonus
case 1: 75str 25dex : (1+.75)(1+ (.33*2)) = 2.905
case 2: 50-50 : (1.5)(1+(.5*3))=3.75
case 3: 100dex : (1)(1+(.66*4)) = 3.64
thats totally made up but you see where I'm getting at lol
This post was edited by jbpellerin on May 6 2009 12:26am