The median is a more robust statistical measure than the mean. If you have a team that won by 50, won by 5, won by 3, lost by 1 and another that won by 8, won by 5, won by 3, and lost by 100, they would both have a record of 3-1 and both have a median point spread of 4. It’s a measure that doesn’t reward or punish blowouts. In the language of statistics, the median is less affected by outlying values.
I’m introducing the idea to step up to some later ideas, more “out there”, but for now, let’s leave you with some stats from the 2010 season, and the notion that for a 10-6 team, Green Bay had a pretty nice median point spread.