By Request, Football

For some reason I completely forgot to follow up my post about the most important stats in football. A friend reminded me over the weekend, so I went ahead and did some analysis this morning. Just in case y'all forgot what I was doing, I'll paste from the original post: "I thought about it for a while, and I came up with two answers: a. starting field position, and b. percentage of offensive series resulting in a first down or a score." Thanks, again, to one of my devoted readers for pointing me to the drive stats listed on footballoutsiders.com. I can't believe I didn't think to check there for the stats; I mean they wrote Pro Football Prospectus 2007, which I read cover to cover. Anyway, here are the results of my analysis (all stats through week 11):

I realized while I was compiling and sorting my data, that in order to be thorough I would have to do hours and hours of research. I really should look at offensive and defensive stats for the past few seasons, and then try to come up with a reasonable way of combining my two stats, because they certainly don't contribute to wins and losses evenly. Instead, all I did was look at offensive stats for this year, ranked the teams one to 32 in each of the two stats, and then I added up the rankings and ranked the sum.
For example, the New England Patriots' average starting field position is the 33.94 yard line, second best in the league. Their drive success rate is 84.3%, best in all of football. One plus two equals three, the lowest sum out of everyone, so they get the best ranking.
For one more example, the Kansas City Chiefs' average starting field position is the 28.35 yard line, 26th in the NFL. Their drive success rate is 59.3%, second worst in the league. 26 plus 31 equals 57, the hightest sum out of everyone, so they get the worst ranking.

Next, I decided to compare their rankings to their winning percentage. Here's where I ran into trouble, because luck plays a big role in wins and losses. One bounce of the football can mean the difference between 6-4 and 5-5; between a .600 winning percentage and a .500 winning percentage. Az additionally, I elected to look at Pythagorean projected winning percentage, which takes into account only points scored and points given up. For example:

The San Francisco 49ers have scored 113 points and given up 223. Inputting those numbers into the Pythagorean formula looks like this: 1 / (1+(223 / 113)^2) = .204. Their actual winning percentage is .200, so that's a very good projection.

Anyway, I determined that the results of the analysis were inconclusive. My system projects the Patriots, Steelers, Cowboys, Colts, Giants, and Buccaneers to be in the top nine. This agrees with their Pythagorean projections. However, my system loves the Browns (4th overall), Bengals (7th overall), and Redskins (8th overall), and their Pythagorean projections are 15th, 20th, and 18th respectively.

Next time I'll look at turnover differential and see what that data will show. I'm looking forward to comments.