Regression to the Mean and the St. Louis Rams

I stumbled upon a interesting article or two this afternoon, the first was an article at AdvancedNflStats.com.  Which is an incredible and informative site and if you do not read it, you are not getting the full NFL bang for your buck.  Whatever that means.  Either way, read what they have to say over there, it is great.

There argument is thus: Coaches are fired all the time between seasons and 68.8% of the time, the team’s record improves the year after.  (The data is from this profootballreference.comarticle from 2006, which is another enlightening site.)  Owners and GMs see this high number and will decide to change coaches because they know that 2/3rds of the time, their record with improve.  (Now, with the Rams there is a very good chance and we ALL hope for our sanity that the Rams will get 3+ wins in 2009.)  They base this theory off something called regression to the mean, which in NFL terms means that teams with 16-0 records in seasons that nothing went wrong, will have something go wrong and lose games the next year.  This also applies in the other direction. 

The mean is the mean for a reason and it pulls teams up and it pulls teams down.  For a team to have a great season, a lot of things have to go right.  And things do not all go right every year.  Bounces of the football and personnel moves and injuries all do not go wrong every year as well.  Which is why regression to the mean works just as well in football as in math.  The bounce of a fumble or the ricochet of an interception are random events that oftentimes change the course of a game or a season.  The regression to the mean law tells us that these random occurrences will start going in the Rams way.

The Advanced NFL Stats article uses the Lions as an example, but we will use the Rams.  The Rams went 3-13 and 2-14 in successive years.  The Rams are primed and ready for a regression to the mean.  How far should the Rams regress to the 8-8 mean or beyond?  No one knows.

The Profootball Reference Article from 2006goes a little farther to study if historically teams did better if they replaced their coach, or kept their coach.  They broke it down into 5 groups and the Rams sadly fall into the 0-3 win category.  But, there is good news.  14 out of 15 times the team improved their record in the next year under the new coach.  The obvious change for the Rams is the head coach, but the Rams will have also changed GMs and plenty of players across both sides of the ball. 

So, what have we learned?  The Rams are getting ready for a mathtastic 2009!  In all seriousness, the Rams should increase their record in 2009 thanks to not only the regression to the mean, but a new head coach, a new GM and hopefully a refreshed, more talented and younger roster.  The Rams will need all the help they can get and we Rams fans will be happy to see them regress back to 8-8 and beyond.

EDIT: The Profootball Reference Article also let’s us know that the Rams got the short end of the stick as they had a 1 in 10 chance to have a worse record in 2008 than 2007 and the Rams nailed that bad boy on the head.  So, after pulling the short straw in 2008, the Rams have an even better chance to improve their record in 2009 – hopefully substantially.


Enjoyed this post?
Subscribe to Rams Gab via RSS Feed or E-mail and receive daily news updates from us!

Submit to Digg  Stumble This Story  Share on Twitter  Post on Facebook  Post on MySpace  Add to del.icio.us  Bark It Up  Submit to Reddit  Fave on Technorati

4 Responses to “Regression to the Mean and the St. Louis Rams”

  1. […] below average in both fumble recovery and red zone rushing in 2008, and those stats should have a strong regression to the mean and show big dividends on the scoreboard. Tags: football […]

  2. […] a whole different post, but the Rams have taken regression to the mean and spit in its face.  In February I wrote an article that looked at some stats that said that the Rams had a 1 in 10 chance to go from 3 wins to 2 wins.  They took care of […]

  3. […] a whole different post, but the Rams have taken regression to the mean and spit in its face.  In February I wrote an article that looked at some stats that said that the Rams had a 1 in 10 chance to go from 3 wins to 2 wins.  They took care of […]

  4. […] February before the 2009 season, I looked at the Rams and some of the regression math based on teams that changed coaches and concluded that after the Rams 2-14 finish in 2008 the Rams […]

Leave a Reply