All in all, about 2,500 studies qualified. But just 185 met their full criteria, which included things like use of a standard method of sperm collection (masturbation, if you're curious), along with a standard method of sperm counting. Studies were also excluded if participants were selected based on some criteria that would muck up the sanctity of the data pool, like infertility concerns or genital abnormalities. In other words, they tried to find studies that included reliable sperm counts of generally healthy men. On those, they ran statistical analyses to see how the subjects changed over time. It took Levine a year of full-time research to deal with the data, and his team spent a total of four years putting it all together.