Carlos Capetillo

M.A., BCBA

Capstone // Weather

The second page has weather information for each game during the regular MLB season. Specifically, this page looks at the mean apparent temperature, in celsius, for the day in which any given baseball game takes place, as measured by the weather station closest to the venue where the game was played. Apparent, rather than actual temperature, was chosen for this data point as it takes into account other factors like humidity and windspeed– it’s essentially what your weather app’s “Feels Like” temperature indicates.

The view presented is for each team and season, except for the 2020 season. You can change the home team on the upper-left corner, and the season in the upper-right corner.

Findings

Overall, we don’t really see a relationship between the average apparent temperature of game days and attendance to games. Looking at it more granularly, we can eek out what looks like some correlation between drops in temperature and drops in attendance during the early (read: colder) months of the season in the northernmost venues, such as the Minnesota Twins’ Target Field and the Boston Red Sox’s Fenway Park.

The actual correlation between attendance (as we’ve been measuring it, in percentage of capacity filled) and mean apparent temperature is -0.04, essentially no correlation. When doing regression analysis, we similarly find a very weak slope value of -0.000994, meaning that every increase of 1° C corresponds to a decrease of 0.0994 percentage points in percentage of total capacity attending a game. It seems like fans put up slightly better with cold weather than they do with warm weather.

Notes

  • Weather data was sourced using Open-Meteo’s Historical Weather API, which can take coordinates as an input. Coordinates for every venue are provided on each venue’s Wikipedia article (see for example Petco Park).
  • The day’s overall mean apparent temperature was chosen as the actual temperature at the time of the game was not something I could easily scrape and incorporate into the rest of my data, but once time allows for it I should be able to incorporate it into the analysis.
  • Games played at venues other than the home team’s venue, including exhibition games, international games, and games moved due to hurricanes and other weather events are not included in the weather data.

← Previous page: Fan Cost Index Next page: Matchups →