After about twenty games or so, however, something interesting happens. The numbers begin to take on more meaning as trends become discernible making it easier to determine which players may have lost a step or have become legitimately better and what off-season roster moves will work out. Obviously, it is a long season in which games and championships are usually decided in the margins, but these general trends become visible at around this point. The line between the sample size being too small and real trends becoming apparent is not altogether obvious. As Yogi Berra might say, the season gets late early sometimes.