>>513313285It ought to be pretty easy to convert this into the % likelihood of a black murdering a white versus vice versa. Assuming the stats are accurate. For example, in the year of the peak of black on white murders, 1980, there were 3677 black on white murders, and 942 white on black murders.
I am assuming that this was before white was split into hispanic and non hispanic, so in 1980 there were 83.1% whites, and 11.7% blacks. 3677/11.7 = 314.27, and 942/83.1 = 11.34. Therefore in 1980 a black was 27.7 times more likely to murder a white than vice versa.
We can also calculate the murder rates per 100,000. With a 1980 population of 29,986,060, blacks had a murders-whites rate of 3677 * 0.0033, or 21.2 whites killed per 100,000 blacks. (approximately the homicide rate of brazil or mexico).
With a 1980 population of 188,371,622, whites had a murders-blacks rate of 942 * 0.00053, or 0.49 blacks killed per 100,000 whites. (approximately the homicide rate of italy or south korea)