I know I'm late to the party and zombie-posting, but the breaking the outcomes down to per alignment kinda buries the lead.that makes sense. how many people are going to make a CN weapon ?
The table actually outputs 50% good alignments (C/L/N), 35% neutral alignments (C/L/tN) and 15% evil (C/L/N) alignments. This means that, at a minimum 40% a weapon will be one that's going to sit inside the parties general alignment paradigm, which makes the table a bit more interesting than absolutes.
As I'm converting this table for my NPC generator, the thing that bothers me most is the extraordinarily high likelihood of an intelligent sword being in play. I mean, 1 out of every 4 swords??
However, when taken in combination with the aforementioned %'s on alignments, it really means that 1 out of every 10 (25% *40% = 10%) will be compatible with a given party. Based upon that, the likelihood of a serious problem sword (say, top 3%) is actually (3% * 40% * 1% * 1% -->> Int 16, 17; align 40%; Special Purpose: 1%, Max Languages: 1%) which comes out to 0.00012% or 1 in 833,333 swords. One of these would be nearly impossible to keep control of, frankly. A +1 sword in this category would be an Ego 36 which would require a character with both 18's in Int and Cha if at 1st level, so only likely for a Paladin. This would be a weapon ya just take straight to Mordor without taking it out of the bag of holding (sheathed of course).
Even a toning down of these odds to an ego of 24 (+1 sword with 3 abilities, 16int, 5,6 lang) is still 1 in 2,500. Using average stat rolls (~4 on a d6) means that even a first level character could control an Ego 24 sword (Int + Cha + Lvl = 12 + 12 + 1).
Most of these weapons would be little more than a nuisance or things simply left behind (because no one can pick them up without dying).
My only issue with the whole Alignment damage thing is that there's no way to determine the alignment without actually picking it up. This means, back to the original odds, there's a 10% chance your low level character is going to die just looting, which seems... um... rough.