Rankings and Systematic Error

The ranking system in football. It is the most bizarre system in modern sports, and the most prone to inaccuracy. However, before I simply lambaste the system and try to figure out a way to exploit it, I need to consider many of the relationships that the ranking system is intertwined with.

Rankings influence the point spread. The point spread is influenced by oddsmakers and crowd perception. Team rankings are also based on past history and performance. Some of that history is so far past that a current team benefits from games played years in the past between none of the players currently involved in the most recent game.

So todays games affect the spread of games in the future between those same "teams" even though all the current players will have graduated.

Also, market size affects the spread and rankings. Bigger markets means more exposure, which means a good performance is duly rewarded.

Remember, a strong finish to the season, which some big names returning, means a high preseason ranking. A weak finish, or big names leaving, means a lower than true ranking. However, the value players in college football are typically offensive linemen, nose tackles, etc. The skill positions are less valuable, especially on offense.

Very similar situation to how points are valued in contracts in the NBA, but don't correlate perfectly with winning.