The first game of the 2016 World Series is tonight — a historic matchup between two teams, the Indians and the Cubs, that haven’t won the World Series for a long time. We all know that Indians and Cubs had respective win/loss records of 94-67 and 103-58 during the regular season. In this post, I’ll explore the pattern of winning and losing for each team during the 2016 season, look at lengths of winning and losing streaks, and see if these streaks may have impacted the home ballpark attendance of each team.
General Winning and Losing Patterns During the Season
One can obtain the game-to-game results of each team from Baseball-Reference — the datasets are easily downloadable and read into R as csv files. (For example, the Cubs game-to-game results can be found here.) In the graph below, I use the ggplot2 package to plot the game-to-game results for each team (a win is a 1, a loss is a 0), and then use a loess smoother to explore the general pattern (the red lines correspond to the overall team winning proportions).
We see that the Cubs had a very strong start, followed by a slump in the middle of the season, followed with a strong winning spurt, and finished the season playing pretty well. In contrast, the Tribe had a mediocre start, playing .500-ball, but had a great winning spurt in the middle of the season. The Tribe fans should be encouraged to see that they finished the season in an upward trend.
Streaks and Slumps
I have a package BayesTestStreak that has a function find.spacings that will quickly find the lengths of all of the breaks (the lengths of losing streaks) between winning games. This is also used to find the lengths of all winning streaks. Below I have used a jittered dotplot to show the lengths of winning and losing streaks for the two teams.
With respect to winning streaks, both teams had notable long streaks in the 2016 season — the Tribe had a 14-game streak and the Cubs had streaks of 11 and 8. (The Tribe’s winning streak was in the middle of the season.) With respect to losing streaks, the Tribe was notable that its longest streak was only 3 games, while the Cubs had two 4-game losing streaks and one 5-game streak.
Impact on Home Attendance
It is well-known that winning and losing has a big impact generally on ballpark attendance. So I thought it would be interesting to look at the pattern of attendance for the two teams for their home games. The two teams have different size ballpark capacities, so I reexpressed each variable by a percentage (100 would be a sell-out). Since there might be a day of week effect, I indicate by color if the game was a weekend (Fri, Sat, Sun) or weekday game.
- I think the Cubs draw well in attendance even in losing seasons (partly due to a famous ballpark), so it not surprising that the Wrigley Field attendance has hovered near capacity for most of the season. There was some available seats early in the season, especially for weekday game.
- The pattern of Indians attendance is more interesting. Early in the season, the attendance was pretty mediocre, even for weekend games. But the long winning streak appeared to have some impact on attendance, although weekday attendance was much smaller than weekend attendance. Surprisingly, the weekend attendance dropped off a little towards the end of the season. (Clearly, the Indians publicity folks need to figure out how to boost attendance at weekday games next season.)
- I can tell you from personal experience that the Tribe’s attendance will be good for the World Series games — I was unsuccessful in getting tickets. (Anyone have extra tickets for Tuesday or Wednesday games?)
Exploring Baseball Data with R 