Monday, October 1, 2018

Exactly just how many people are catching Marlins games in Miami?

TLDR: In the months leading into the 2018 Major League Baseball (MLB) season, many changes occurred in the Miami Marlins organization. First, the team changed ownership. Then, the new owners cut costs by trading away some of their most expensive and talented players. Next, the team changed the way it calculated their home-game attendance. Lastly, the fans decided how much they truly like the Marlins and made their decisions on whether to purchase tickets or not ...

Now at the conclusion of the 2018 MLB season, reports are coming out on just how many fans showed up the Marlins home games. Taken at face value, most of these reported numbers are shocking: a 49% decline in attendance?! 

But calculated correctly, I find that the demand for Marlins games has dropped by a non-trivial 10% in 2018. Not only did Miami reveal they have the MLB's worst attendance, we now see they are trending in a very dubious direction.

The Miami Marlins, a self-purported "Major" League Baseball team, sucked this year. And while they narrowly avoided the dubious '100-loss' title, they lived up to their fish moniker and stunk. The 2018 Miami Marlins season resulted in some of the least attended MLB games in the history of their franchise, setting records for fewest fans since Montreal had a baseball team. In total, the 2018 Miami Marlins saw only 811,104 fans: a 49% decline in attendance over 2017.

So what happened? Most people would point to the fact that the Marlins traded away four of their most talented, albeit costly, players by the new ownership that purchased the franchise in October 2017.

Some would also then be quick to point out that the Marlins changed the way in which they count attendance. While typically MLB teams report the number of tickets sold to a game, a concept called the paid attendance, the Marlins have historically 'inflated' their attendance numbers by including promotional (free) and heavily discounted tickets as part of the paid attendance. But, beginning with the start of the 2018 season, the Marlins now have aligned their attendance calculation with the rest of the MLB.

But if the 2017 attendance includes ticket sales that are not included in the 2018 attendance, we cannot take the 49% decline in attendance at face value.

Instead, consider that the Miami Herald reported that the true paid attendance in the 2017 season was only 820,000. This would imply the 2018 season saw just a 1% decline in attendance. Sounds like the 2018 Marlins were not so bad after all, right?

But this number is deceiving as well. What the 1% decline will not show is that the 2018 Marlins hosted the two big-ticket teams, New York Yankees and Boston Red Sox, for just two games each which resulted in 10% of the Marlins annual attendance. We are still not comparing apples to apples.

What we do know is that the decline in attendance is made up of many effects including:
  1. The change in quality of the 2018 Marlins, 
  2. The change in the quality of the schedule of the 2018 Marlins; and, 
  3. The different measure of attendance.
* Note that according to Forbes, the average ticket price did not change from 2017 to 2018.

I would argue the best way to compare the two seasons is to remove the effect of the quality of the schedule and the measure of attendance (2 and 3 from above). Then we are left with the change in the quality of the 2018 Marlins and how the fans reacted to it.

To isolate the change in the quality of the 2018 Marlins, I predict the attendance in a hypothetical world where the 2018 Marlins plays the same schedule as the 2017 Marlins. To do this, I start by estimating my regular model used to predict the attendance using game-specific factors, such as the day of the week, month, and the opponent (here are some links to see more examples of this methodology using the Chicago White Sox, the San Diego Padres, and the entire league in general). These models typically are able to explain 80% of the variation in the data, and the data, as always, comes from

Using 2018 Marlins regression, I apply the estimated coefficients to the 2017 Marlins schedule to predict the attendance of the 2018 Marlins in 2017. Below, I offer a graphical depiction of the results.

The left two bars are the official attendance numbers submitted to the MLB from the Miami Marlins. As discussed, the 49% decline in attendance is made up of too many effects that we would not normally want to have impact the results, most notably that the 2017 figure includes free tickets.

To correct for this, I compare the reported 2017 paid attendance to the model's predicted attendance in the hypothetical world where the 2018 Marlins play the 2017 Marlins' schedule. The results suggest that the 2018 Marlins would have drawn 10% less fans in the hypothetical 2017 Marlins schedule, or that the demand for Marlins games has dropped 10% in 2018!

With the change in the attendance calculation, the Miami Marlins have also shown that they are dead-last in the league for attendance. Now, under new management, it appears the Marlins attendance are trending in the wrong direction.  And with some minor league baseball teams with similar attendance figures, it appears the Miami Marlins are small fish in a big pond.

No comments:

Post a Comment