The numbers are based on surveys of population conducted in 1993 and 1994 by the Centers for Disease Control. The results were confirmed by a 2000 study in the "Quantitative Journal of Criminology" (Vol 17). The results of these findings were then expanded to the total expected for the entire US population. This is not an unusual technique in statistics. Even if we reduce these results by a factor of 5 to account for conservancy and then say that only 50% of those were an actual prevented death the pro-defense side still comes out ahead by a factor of roughly 20 to 1 based on 2010 data (~11,000 homicides).