Low Nielsen Ratings

I'm tired of hearing about this being the lowest rated world series.  So was last year's.  And the 2000s have been lower than the 1990s.  That's not news.  But, to show this, I did a quick analysis using data from ESPN (see data below for years 1996-2006). I did a multiple regression regressing either Rating or Share on year (coded 1 = 1995 to 11 = 2005) and on # of games played. I deliberately excluded 2006 to pick up the trend by year to see where you'd expect 2006 to fall. The outcome is very simple:

Year is a highly significant predictor in both regressions despite the small sample.  # of games makes no difference.

Ratings are dropping 1/2 a point per year.

Share is dropping 1 point per year.

The upshot:  Thhe 2006 viewing results are right at what you'd expect given larger patterns in TV viewing (i.e., more shows, more use of the internet, etc.).

The aberrant World Series was 2004, which was well above the trend (technically, it's "residual error" is 1.6 standard deviations above what would be expected).  Thus, our Cardinals were in the most unexpectedly popular World Series in recent years; of course, credit for the level of interest is shared with a large Red Sox fan base and the mystique of an 80 year old curse. (It wasn't due to an exciting series.)

I'll go out on a limb and predict that next year's share will be about a point lower than this year's.  

The ESPN data are included below.

Year            Rating  Share   #Games
2006 (Fox)    10.1    17    5
2005 (Fox)    11.1    19    4
2004 (Fox)    15.8    25    4
2003 (Fox)    12.8    22    6
2002 (Fox)    11.9    20    7
2001 (Fox)    15.7    25    7
2000 (Fox)    12.4    21    5
1999 (NBC)    16    26    4
1998 (Fox)    14.1    24    4
1997 (NBC)    16.8    29    7
1996 (Fox)    17.4    29    6
1995 (Fox-ABC)    19.3    33    6