Since the All-Star Break is coming soon, it seemed appropriate to write something related to the two halves of a baseball season. Some teams like the Astros have had great first-half seasons, and other teams like the Phillies have had tough first halves. This raises the question:
What impact does mid-season success have on the team home attendance for the remainder of the season?
I would think there is a connection between success and attendance — teams with winning records in the playoff chase should have a rise in attendance, and teams with losing seasons should experience drops in attendance. We’ll explore this question using Retrosheet game log data for the 2014 season.
I download the game logs for the 2014 season. I create a new variable
Half indicates the half of the season (I define the first half as games before July 1). Then I compute, for each team, the mean home attendance in the first half and the mean home attendance in the second half.
In the following graph, I plot the average home attendance against the percentage change
PERCENTAGE.CHANGE= (ATTENDANCE.2ND.HALF – ATTENDANCE.1ST.HALF) / ATTENDANCE.1ST.HALF. (I call this a mean-difference plot since this similar to the plot of this name popularized by John Tukey.)
Several interesting things we see from this graph:
- The big market teams (SFN, NYA, SLN, LAN) experienced little change in attendance from the first half to the second half.
- A couple of Texas teams (TEX and HOU) experienced 10% drops in attendance in the 2nd half.
- A number of small market teams (CLE, KCA, OAK, SEA) experienced big (over 10%) increases in attendance
To address our original question, we collect the number of wins and losses for each team in each half of the season. The next plot graphs the teams success in the first half (the difference between the number of wins and the number of losses on July 1) against the percentage change in attendance.
This is interesting. The two Texas teams had losing records at midseason which could explain the 10 percent drops in attendance. In contrast, practically all of the teams with winning records at midseason had gains in average attendance after midseason. (There are two exceptions, Atlanta and San Francisco who had slight drops in attendance despite a winning record on July 1.) Generally, there is a positive relationship between midseason success and attendance growth. Also, Major League Baseball should be happy to see the rise in attendance for most of the teams — it seems that the new wild card system has kept more fans interested in seeing their teams play.
If you’d like to reproduce these graphs on R, the code is available on my gist site.