Dodgers Marketing Promotion

Business Problem

I want to determine what night would be best to run a Dodgers marketing promotion to increase attendance.

Data

I used Dodgers Major League Baseball data that contained the following variables:

month

the month the game was played

day

the day of the month the game was played

attend

the number of spectators in attendance

day_of_week

the day of the week the game was played

opponent

the opposing team

temp

the temperature during the game

skies

indicates whether the skies were clear or cloudy

day_night

indicates whether the game was a day or night game

promo

indicates which promotional item is given away: bobblehead, cap, fireworks, shirt, or none

Analysis

First, I created boxplots of each categorical variable to see how each compared with the attendance.

Figure 1 - Boxplots of categorical variables

Though there were months with less attended games than September and October, the maximum attendance is general much less than other months. This pattern can also be seen on games played on Monday's. Games that have bobblehead promotions have the most attendance. Next, I looked at scatterplots of the numerical variables.

Figure 2 - Scatterplots of numerical variables

There seemed to be a trend with the temperature variable. When the temperture is greater than 80∘, attendance does not reach 43,000.

Figure 3 - Boxplot of temperatures by month

The months with the highest temperatures were September and October. Next, I used R to create a linear regression model to predict attendance based on month, day_of_week, and promo. Using the model, I looked at Monday games in September and October.

Results

There was only one game played on a Monday in September, in which there was no promotional giveaway. Attendance for this game was 33,540. My linear model predicted attendance to be 33,061, which is very close to the actual attendance. I predicted attendance to be approximately 43,761 if we were to run a bobblehead marketing promotion on this day, which is a 30% increase in attendance. Similarly, there was only one game that was played on a Monday in October, with an attendance of 33,624 and no promotion. I predicted an attendance of 44,086 with a bobblehead giveaway, which is a 31% increase.

Conclusion

Based on my analysis, I would recommend running bobblehead marketing promotions on Monday games in September and October to increase attendance.