Maybe, I’m a little crazy. However, the reason I got into statistical research of football is to see where a player’s true value actually is, and to see if, somewhere down the line, I could build a projection system that could forecast all 3-4 years of a college player’s career. Today, I started to build that system. (Even though the world ends in 4 hours)
First off, in order to do that, you need to figure out your raw data. So, what I did was gathered 18 quarterbacks, from years ranging from 1978 to the present day. The quarterbacks needed to play significant time in 4 years of a college career. Guys like Peyton Manning, Brett Favre, Joe Hamilton, among others served as just an example of the sample size I took.
The results are pretty consistent from a player’s freshman year to his junior year, rising at a pretty consistent clip. However, there was a bit of a decrease from a player’s junior year to a player’s senior year. Most of the decrease is statistical noise, however, there seems to be a plateau effect in a player’s senior year.
Here are the percentage increases in a players game from junior to senior year….
- Freshman – 7.2 YPA
- Sophomore – 7.6 YPA, a 5.6% increase from freshman year
- Junior – 8.2 YPA, a 7.9% increase from sophomore year
- Senior – 8.0 YPA, a 3.4% decrease from junior year
Again, the decrease in senior year is likely due to noise more than anything else. However, the percentages listed above will be the base for a future projection system. In part 2, I’ll look at some factors for building a projection model.