I mean it's almost all young men from other ethnicities committing violent crime too, although youth violence is typically slightly more pronounced in black demographics than others.

It's all percentages relative to other demographics so if you're cutting down to only young men of black ethnicity then you should cut down to only young men of other ethnicities, at which point you're back to ~13% of the total sample size.

If not more like 15+% at that point, since black birth rates have been higher for the last couple of decades as well as doing a worse job of surviving into old age, so they have an overall younger age distribution.