What easily observable features statistically indicate amounts of crime likely to happen? In order.
- Maleness. More males, more crime. Given Sailer's option walking choice to pass a rough looking group of black girls as compared to a group of Asian boys, walk past the black girls.
- Youth. More youth more crime. Given Sailer's option to walk past a group of Morgan-freeman-aged (and looking) black men as compared to a group of White teens, pick the old guys.
- Wealth. Simply look at the social class of the groups of teens by means of observing their dress. Walk towards the rich group. This may or may not beat #2.
- Athleticism/likely testosterone. A bunch of kids wearing glasses, and holding MTG cards is pretty low danger, regardless of race. A bunch of burly football-player types is moderate danger, regardless of race. This mitigates 3. I'm not entirely convinced that blood-serum testosterone level isn't just the complete winner here for danger-level. Femaleness, Age, and Race all impact said testosterone level. And it may be possible that some/many black women have higher testosterone levels than some/many Asian men.
- Race. On average...which you should damn well pay attention to if you know nothing else...then for safety reasons, you should choose to dodge black kids over Indian kids...and that includes my 16yo SWTOR-playing, football-player built, dark-black stepson. You should also know that this decision, entirely correct for statistical reasons, is painful to him (if he notices) and that every time you make a decision like that, it will be noticed...and he will emotionally (correctly) observe the fear/dislike of blacks, which will (on the margin) push him towards bullshit black race-baiters like Farrakhan. Statistically aware correct behavior from you, the crime avoider...creates ripple effects that suck for everyone.