My math is counting the +0.5% weapon attack....
When I said that, I thought it was 21 luk to go past 91 str, not 11.
Sorry I was looking at LUK as LUK/5 like DEX is, but it's LUK/3.
Also before I continue let me say the only thing I am contradicting you on, is the implication that anyone who chooses > 90 STR over getting at least 11 LUK is making some kind of mistake.
The way I see it:
90 STR = 90 ATK, +45% Wep_ATK
Assuming a weapon with 100 base weapon attack (which is pretty low.)
940 Stat.Pts. = 91 STR = 91 Status_ATK, +45 Wep_ATK
940 Stat.Pts. = 90 STR, 6 LUK = 92 Status_ATK, +45 Wep_ATK <-- better by 1 ATK
951 Stat.Pts. = 92 STR = 92 Status_ATK, +46 Wep_ATK <-- Exact same total atk
950 Stat.Pts. = 90 STR, 11 LUK = 93 Status_ATK, +45 Wep_ATK <-- Exact same total atk
973 Stat.Pts. = 94 STR = 95 Status_ATK, +47 Wep_ATK <-- better by 1 ATK
974 Stat.Pts. = 90 STR, 19 LUK = 96 Status_ATK, +45 Wep_ATK
995 Stat.Pts. = 96 STR = 96 Status_ATK, +48 Wep_ATK <-- better by 1 ATK
996 Stat.Pts. = 90 STR, 25 LUK = 98 Status_ATK, +45 Wep_ATK
(I don't want to split hairs over the fact that status attack is doubled in damage calculation, as we're talking like 2 damage at most which is irrelevant).
The way I figure, LUK is never really "better" than STR unless you want CRIT or Lucky Dodge more than weight capcity, with a pretty low ATK weapon. With a high ATK weapon (say a hurricane fury or a pile driver), STR is clearly better for damage-- always.
LUK only looks like a good investment at 5 total LUK, IMO. And only if your WEP_ATK is under 120 or so.
Edited by Jaye, 27 February 2013 - 02:23 PM.