In my last two articles, I've compared total points to what I consider to be a superior statistic, expected points, for all NFL **offenses** and **defenses**. The idea is that if we know how many points an offense should have scored or how many points a defense should have allowed based on how well they actually played, we can figure out how lucky (or unlucky) they got by examining their actual points scored/allowed.
If an offense ranked 18th in points scored but ninth in expected points, then there's a pretty good reason to believe that they were unlucky. That's actually exactly what we see with the Carolina Panthers, a team that played quite well on offense in 2012, and one the numbers suggest could improve in 2013. It's not that the Panthers are guaranteed to experience better-than-average luck – every NFL team is likely to experience an *average *amount of luck going into a season – but rather that since they were so unlucky in 2012, they'll probably score more points in 2013 even if their luck is the same as the league average. That's the basic idea behind regression toward the mean. Random results tend to congregate near the average.
Reading the Charts
In each analysis of expected points, I posted a chart that showed how each team's actual rank compared to their expected rank. Here's a small example:
Before I examine the trends for the Cowboys' division rivals, I want to mention that these lists aren't meant to rank teams in terms of overall talent. I don't think that the Panthers have the best offense in the NFL, or even close to it. Rather, I simply think they're most likely to improve their point total in 2013. Even if they play exactly as well as they played last season, they'll score more points this year.
Below, I combined the differences for all teams on both sides of the ball. The Panthers had an expected difference of 9 on offense, as you see above, and 1 on defense. That's another way of saying that, if the Panthers play exactly the same as they did last season, they'll probably finish close to nine spots higher in total points and one spot higher in total points allowed.
Again, we can view this as a chart for the "likelihood of improvement." The teams at the top are most likely to improve upon their offensive and defensive ranks (and thus their records), while those at the bottom are most likely to regress.
It's worth noting that it's easier for some teams to improve or decline based on where they finished in 2012. The Patriots scored the most points in the NFL last year, for example, so there was no room for improvement for them in terms of expected points. They finished first in that category as well, but they couldn't possibly have improved. That means when you see a team that was already good in 2012 listed near the top (or a team that was bad near the bottom), their fortunes are really likely to change in 2013.
Using Expected Points to Predict Changes in the NFC East
Let's look at the expected overall change in total points scored and allowed for the teams in the NFC East:
- Eagles: 8
- Cowboys: -1
- Redskins: -4
- Giants: -11
Remember, these numbers are based solely on a regression of luck, not changes in personnel, coaching staffs or health. That's great news for the Cowboys; despite all of their injuries in 2012, Dallas still finished 8-8. When you consider that the 'Boys are highly likely to be healthier this season, there's a really good chance that they'll improve upon their 8-8 record.
The Eagles are the only team in the division that got unlucky in 2012 (in regards to how many points they should have scored and allowed). Based on how they played, Philadelphia should have finished three spots higher in points scored and five spots higher in points allowed. Part of their expected improvement is due to them just being a bad team last year, however.
The Redskins had a small amount of luck in 2012, but nothing that greatly affected their point totals. However, take a look at the Giants. New York finished sixth in points scored, but they played well enough only to finish 10th. On defense, they allowed the 12th-fewest points, but they should have finished seven spots worse. None of that is a guarantee that the Giants are going to be worse than 9-7 in 2013, but it suggests they are highly unlikely to once again post a 85 point differential.
In some divisions, the projected luck-based changes might not matter all that much. In the hypercompetitive NFC East, however, all four teams have a legitimate chance at taking down the division. As teams get closer and closer in talent, the role of luck increases. The Cowboys, Eagles, Giants and Redskins are all so evenly matched that the teams' fates could once again be decided by something as simple as an odd bounce of the ball.
