http://xoth.net/publishing/freebies/world_of_xoth_players_guide_5e.pdf

There's only humans in the world, but they have bonuses based on culture. 

Savage: +2 Str; +1 AC; advantage on Perception; disadvantage when attacking spooky monsters
Nomadic: +2 Dex; advantage on Wisdom saves; temporary skill once per day; 20 foot walking speed
Civilized: +1 to any two stats, one extra skill, one extra feat, disadvantage against poison and disease
Enlightened: +2 Wis; reroll 1s; History bonus; disadvantage on initiative
Decadent: +2 Cha; advantage on Deception & Stealth; advantage on poison saves; disadvantage on Wisdom saves
Degenerate: +2 Con; night vision; extra critical damage; disadvantage on Deception and Persuasion

---------------------------------------------------------------------------------

I posted about this in an [url=https://www.therpgsite.com/pen-paper-roleplaying-games-rpgs-discussion/atypyical-race-class-combos/]older race thread[/url], but I feel like it got lost in the weeds. In D&D-based games, one of the big distinction of races is the ability stat modifiers -- but I'm going to posit that they don't work the way they're billed, and ask about alternative approaches. The premise here is: 

[b]Stat modifiers don't produce the implied distribution, unless using roll-in-order[/b]

I feel that originally - and in many players interpretations, stat modifiers are supposed to reflect different averages. Say humans are +1 in every stat, and gnomes are +2 Intelligence but +0 Wisdom. Intuitively, the implication is that gnomes will have intelligence on average 1 high than humans, and wisdom 1 lower. And if the PCs are generated by choosing race, rolling 3d6 for each stat in order, and then applying racial modifiers -- then this result will apply to gnome PCs. 

But let's say we do roll 3d6 six times and then arrange the stats. Generally in D&D, a +1 in the prime ability for your class is worth much more than a -1 in your worst stat ("dump stat"). So with roll and arrange, it's a clear benefit for a gnome PC to pick a class with Intelligence as their prime ability. Thus, tactical players will tend to make their gnomes Wizards. The result is that in a distribution, gnome PCs will likely have a much higher average intelligence than humans, unless the set of players are up for a challenge or just trying to buck the system. 

Either way, the distribution isn't going to be what is implied by the stat modifiers. The effect isn't giving the distribution implied. 

---

So it seems to me that 

In some point-based games including