While professional football might never get the Moneyball-type benefit that baseball has, I still think there are undervalued "situationalists" in the NFL that teams could exploit to their advantage. This simply means someone that could excel in a certain underrated skill that adds value to a team, similar to how on-base percentage has for MLB teams.
One of those situationalists is the kickoff responsibility. To state the obvious, kickoffs are necessary in football to start either half or after scores; they're usually exciting and filled with hope for the receiving team but also violent and dangerous, both to the player's health and to the team kicking off. So it'd be great for a team kicking off to limit the number of chances that you give the other team to return one for a touchdown or to get a great starting position as well as limiting the amount of number of chances for someone on your team to get injured trying to cover the kickoff. Thus why touchbacks are so important to NFL football.
I love the accessibility and usability of PFF stats, so I wanted to get some data behind the value of having someone on your roster that had a strong leg on kickoffs and got more touchbacks. Below is a scatter plot of NFL kickers that played 8 or more games in the 2013 season, with their PFF kickoff value (the higher the value, the better the kicker was at kickoffs) on the Y axis and the average distance of their kicks in yards on the X axis (PFF doesn't clarify if this includes touchbacks or not).
The linear trendline describes the overall picture, namely that the farther a kickoff is, the better the kicker is at kickoffs. This shouldn't be a surprise. The R-squared value of the trendline is the "fit" of the trendline for the points, so a higher R-squared value means the trendline better describes the data.
Below is a graph of the kickers PFF kickoff value on the Y axis versus the average starting yard line of the other team on the X axis.
This trendline shows that, generally, the lower the average starting position of the other team the better the kicker is at kickoffs. However, this also has to do with how well the kicker's team is at covering kicks. That's why it makes sense that the average starting yard line trendline's R-squared value is lower than the average distance's trendline, because how far a kicker kicks if more of an indicator of how good that kicker is at kickoffs than the average starting yard line, since the latter isn't fully under the kicker's control like the first.
Now let's see how touchbacks affect how good a kicker is at kickoffs. Below is a graph of the kicker's PFF kickoff value on the Y axis versus the percentage of their kickoffs that were returned (assuming a kicker with a higher percentage of touchbacks would have a lower percentage of their kicks returned).
This trendline's R-squared value is much higher than the previous two other characteristics we looked at, meaning that the lower the percentage of kickoffs that were returned is a better indicator of a better kicker on kickoffs than the kicker's average kickoff distance or starting position of the other team.
All of the above should be pretty obvious to most followers of the sport, but it's good to have some data behind assumptions. So now let's look at what was referenced in the title, the case of Chip Kelly and Eagles recently picking rookie Cody Parkey over veteran Alex Henery as their kicker heading into the season. The first obvious reason is he was cheaper, as a rookie makes much less, meaning the Eagles had more money to spend elsewhere. But let's move beyond that.
Chip likes to think out of the box when it comes to how he runs his team and so the Eagles trading a backup RB to the Colts for a rookie kicker to battle the entrenched veteran, something most teams wouldn't do, is another example of this type of thinking. But Chip and the coaching staff obviously thought there was a chance they can improve their roster, so they brought in competition.
Undoubtedly the most important quality a NFL kicker has to have is making field goals, so for the Eagles to pick Parkey over Henery they likely are at least equal when it comes to that. But I believe a large part of Chip choosing Parkey has to do with his prowess on kickoffs compared to Henery. Parkey led the NCAA in the number of touchbacks last year, hitting roughly 70% of his kicks for touchbacks. Henery only kicked about 40% of his kicks for touchbacks last year. Granted the footballs they were using were different in shape, both college and NFL kickers kick off from the same yard line so Parkey can be considered a much better kickoff kicker than Henery, touchback-wise.
Assuming the Eagles have about the same number of kickoffs this coming season as Henery did last year (100 or so), switching to Parkey would result in about 30 more touchbacks than Henery! That's 30 less chances for the other team's returner to take one to the house and 30 less chances for one of your players to get hurt trying to tackle the returner. That is a huge advantage to the Eagles for switching out their kickoff kicker. Assuming both are comparable field goals kickers (this is a big assumption, I'm not downplaying it), Parkey would likely add much more value to the Eagles than Henery would have. This decision was another reason Chip Kelly is a great coach.
Now if a team already had a kicker who was excellent at field goals, but struggled on kickoffs, I could see a team keeping a kickoff-specialist on the 53-man roster in addition to that kicker as opposed to a 3rd string LB or TE. I hope to prove this in a later analysis.