There are strong sex based differences in propensity for youth violence in WEIRD countries. There's less of a strong correlation outside of those countries but it still broadly applies when analyzed.

It's more complex than just being "hardwired" or "testosterone" but the differences are statistically significant and repeatable.

There is a strong correlation between intelligence and being a criminal. It's not predictive, but if you have low intelligence you're part of a vulnerable group that is far more likely to be imprisoned. This could also be that smart criminals don't get caught, but large case reviews with low intelligence offenders often find a lack of foresight of what happens after commiting a crime, and a lack of understanding about how serious it can be. Both of which strongly correlate with criminality. It's a fancy way of saying they don't think ahead and don't understand the magnitude of what they're doing.

Within the literature the greater male variability hypothesis is not controversial, but well established fact.

The rest of my opinion is derived from a statistics 101 level of understanding of bell curve comparisons between two groups and what happens to the tails and outliers (which is also backed up piecemeal by my sources in different areas indirectly.)

I'm willing to add more sources as needed but I think it's time you did some searching to refute my position if you still think it's suspect.