I had forgotten that. The info is still useful, but may be weighted a bit toward that players first half. If the guy wasn't dramatically different from the first half to the second, the data would be more reliable.