Let me restate the method I like to use. Granted, it's not an official method, but I think it offers the best of point buy and random generation. Besides, rolling multiple characters and selecting the best one isn't an official method either.
The method is based on the assumption--no, actuality--that everyone else in the D&D world draws from a 3d6 distribution. If this is the case (and it officially is), then 1 in 216 people will have an 18 strength. I think it's safe to assume that even fewer than 1 in 200 NPCs are adventurers.
So, why not marry these two statistics? Let every player have a single 18. The catch is they have to roll 3d6 for the other 5 scores, no rerolls, and no multiple character selection--unless you want your character to be a 1 in 400, 1 in 600, etc.
Alternatively, let them have two 16s. Roll 3d6 for the other 4 scores. You get the picture. The advantage of this method, I've found, is that everybody gets something, but there's still the feel of randomness to the rest of the character.
If you think this is a gross cheat, you'll be shocked to know I've even played it with a two 18s assignment method (1 in 46656 have an 18 Str, 18 Con). That was for a high, high challenge campaign, with half the party slaughtered by level 7. But they sure had fun with it. Of course that's the point: everybody having fun, even when Elliot steps on the pizza.
The method is based on the assumption--no, actuality--that everyone else in the D&D world draws from a 3d6 distribution. If this is the case (and it officially is), then 1 in 216 people will have an 18 strength. I think it's safe to assume that even fewer than 1 in 200 NPCs are adventurers.
So, why not marry these two statistics? Let every player have a single 18. The catch is they have to roll 3d6 for the other 5 scores, no rerolls, and no multiple character selection--unless you want your character to be a 1 in 400, 1 in 600, etc.
Alternatively, let them have two 16s. Roll 3d6 for the other 4 scores. You get the picture. The advantage of this method, I've found, is that everybody gets something, but there's still the feel of randomness to the rest of the character.
If you think this is a gross cheat, you'll be shocked to know I've even played it with a two 18s assignment method (1 in 46656 have an 18 Str, 18 Con). That was for a high, high challenge campaign, with half the party slaughtered by level 7. But they sure had fun with it. Of course that's the point: everybody having fun, even when Elliot steps on the pizza.