top of page
Search
  • Writer's picturehenryfarleyjohnson

World Series TV Ratings

tl;dr: It doesn't seem like team popularity has much of a relationship with World Series TV ratings. Series length and year, however, appear to be closely correlated with TV ratings.


Times are tough for a shrinking sport in a shrinking medium. Just look at this chart of World Series TV ratings by year from Wikipedia:

A common refrain this time of year is that the MLB must be pulling for big market teams in the hopes that a marquee matchup might boost the ratings. I decided to see how much support this idea has in the data.


To measure team popularity, I used two different variables: the number of likes each team has on Facebook and the population of the metro area each team represents. Let's plot those two variables against each other just for kicks:

There's a positive relationship between these two, which isn't shocking; bigger market = more fans = more likes. Those outliers on the far right are the Yankees (by far the leader in likes) and the Mets (oh nooooo). Let's see which teams have the most likes per population:

And the fewest:

OK, enough messing around! Let's see how these measures of popularity correspond to ratings. Here is the average rating, in millions of viewers per game, for each World Series plotted against the combined metro population of the two teams playing:

Shoutout to the 2000 Subway Series on the far right!

This looks like a slightly positive relationship, but it's not statistically significant. A plot of rating vs. combined Facebook likes for the two teams looks virtually identical, so I won't waste your time with it, but the same is true: whatever relationship might exist appears to be pretty weak. To the extent that there is some relationship (say, a U-shaped quadratic one), it's washed away when you account for year and series length, which are highly correlated with ratings:

In fact, I cannot find a significant relationship between TV ratings and team popularity no matter how I try: taking only the higher of the two teams' popularity, testing a quadratic relationship, using a linear model, using a GAM model. There just doesn't seem to be anything there. (And fwiw, I've tried so many methods that you'd half expect a significant result even if there weren't a true relationship...)


So perhaps the moral here is that you should feel free to root for small market teams with impunity. Those Tampa Bay Rays might not be killing ratings after all! Not that it matters this year, since the only remaining teams are the Boston Red Sox, Houston Astros, Atlanta Braves, San Francisco Giants, and Los Angeles Dodgers...

Comments


bottom of page