I think the real question is how do we tell which is which?
That probably could depend upon whether the observer is male or female himself, no?

I think the real question is how do we tell which is which?
YMy first thought is all the minimums are way too high...but then it occurred to me that you're probably starting from a point buy or array setup that (ridiculously, IMO) doesn't go lower than 8.What min/max requirement would like to see for races?
For my self (min/max)
Dragon Born: ST: 12/22, Dex: -/18, Con: 8/20, int: -/20, Wis: 10/20, Cha: 11/21
Dwarf: ST: 12/21 or 22*, Dex: -/18 Con: 12/22, Int: -/20, Wis: 11/21*, Cha: -/18 or 19
Elves: ST: -/18, Dex: 12/22, Con: 9/19 int: 11/21*, wsd: 11/21* Cha: 11/21*
Half-Elves: No min max or as elves for minimums.
Half-Orcs: ST: 12/22, Dex: -/20, Con: 12/21, int: -/19, wis: -/20, cha: -/ 16
Halfling ST: -/16, Dex: 12/22, Cn: 10/20 or 22*, int: -/20, wis: -/20, Cha:11/20 or 22*
Gnomes: ST: -/16, Dex: 11/20 or 21*, Con: 11/20 or 21*, Int: 12/22, wis: -/20, cha -/20
Tieflings: ST: -/20, Dex: -/20, Con: -/20, Int: 11/21, wis: -/18, cha: 12/22.
That is just a sketch. What do you say?
I think the real question is how do we tell which is which?
In my experience, character creation in AD&D starts with rolling dice and ends with picking a name. With female humans capped at 18/50, I could easily imagine the player (regardless of gender) making the character male if they rolled over that. I could also imagine a player specifically making the character female if they happened to roll exactly 18/50, because it's such a rare roll.
I'm having more trouble imagining a player who rolls 18/00, and then consciously chooses to make the character female, with full knowledge of what they're giving up.
Yeah. I'm used to a system that has non-linear bonuses and scaling, and where the difference between 16 and 18 is much greater than, say, the difference between 10 and 14.See, that's the trick though Lanefan. You're saying that the strongest an elf gets to a human is 16 vs 18. Thing is, there's very, very little difference there. +1 to hit and damage and a bit of carrying capacity. That's it. So, I'm half your size, because a big elf and a big human are about that ratio - the biggest elf weighs 130 pounds, while the biggest human about 190. Giving up sixty pounds is only that bit of a difference in stats? The elf can carry 240 pounds comfortably, while the human is carting around 270. Hrm, doesn't seem like that much difference. I'm about half your size but, I can carry pretty much anything you can?
That's why I'm so dead set against this. The system just isn't granular enough to make this make sense.
YMy first thought is all the minimums are way too high...but then it occurred to me that you're probably starting from a point buy or array setup that (ridiculously, IMO) doesn't go lower than 8.
From my viewpoint, the way to approach this is twofold:
For initial roll-up stats, use the human 3-18 bell curve as a basis then look at each stat for each other race and see how it compares at each end. What does the lowest-strength Dwarf look like compared to a Human (probably about 7)? How strong can the strongest Elf get, relative to a Human (I'd say 16)? And so on.
Then, assuming an absolute max of 20 in each stat as the Human default, go through again but this time asking what the absolute max would be for each stat by race. This part you've basically got right other than quibbles here and there about specific numbers.
Lanefan
What does the scaling have to do with it? Any scale can be mapped to any other scale. If you want a dramatic difference between elves and humans, you can do it on a linear scale just by making the elves' cap 12 and the humans' cap 24, or whatever other values match their abilities on the nonlinear scale. And on a linear scale, it'd be a whole lot more transparent that there is a dramatic difference there, too.Yeah. I'm used to a system that has non-linear bonuses and scaling, and where the difference between 16 and 18 is much greater than, say, the difference between 10 and 14.
Which makes me wonder: maybe linear bonuses aren't the answer...
Lan-"bell curves and J-curves have their place, and it may well be here"-efan