Executive Summary
Steward Roddey, the general manager of Oakland A’s baseball team is faced with the decision of whether or not to give a hike to Mark Nobel, the second best pitcher in the American League. Nobel’s agent was commanding a contract fee in the region of $600,000 per year owing to his performance statistics from the 1980 season. One major argument presented by Nobel and his agent is that Nobel has the ability to attract crowds and thereby increase attendance to the games and drive ticket sales. The agent quoted a figure of $105,650 as the amount lost as revenue when Nobel did not start.
To answer the question of whether or not the presence of Mark Nobel increases ticket sales, we have performed a regression analysis of the ticket …show more content…
Hence, this variable (O4) is very significant in determining the no. of tickets sold.
Model 3: Multiple Linear Regression Model between No. of Tickets Sold and all the identified predictor and dummy variables (as classified in table 1)
Exhibit 4 shows the detail analysis of this regression model.
The regression equation we get is as follows: Ticket Sold = b0 + b1 . POS + b2 . GB + b3 . TEMP + b4 . NOBEL + b5 . WKEND + b6 . PREC + b7 . DAYTIME + b8 . TV + b9 . PROMO + b10 . DH + b11 . O1 + b12 . O2 + b13 . O3 + b14 . O4 + b15 . O5 + b16 . O6 + b17 . O7 + b18 . O8 + b19 . O9 + b20 . O10 + b21 . O11 + b22 . O12
Though the R-square value obtained from this model is quite significant (0.7939), p-values of many predictor variables fall outside the significance level (95%). Hence, we further need to optimize the model by eliminating the variables with high p-values and taking into consideration only those variables whose p-values are within the acceptable range.
Model 4: Forward Selection …show more content…
This shows that the addition of Nobel variable has essentially lowered the R-square value and hence the addition of Nobel to model is not significant.
It could also be noted from the very high p-value of Nobel (0.45644), which is quite high and more than 0.05 (Significance level assumed to be 95%). Hence, this further confirms the analysis that variable (Nobel) is not significantly determining the no. of tickets sold.
Model 6: Using Interaction variables
We tried to model the regression using interaction variables. For this we considered the interaction between the variable NOBEL and each of the other predictor variables, including dummies (Exhibit 9). However, as our data did not contain any continuous variable except for TEMP, GB and POS, we did not find any relationship between the interaction variable and the dependent variable. Hence this approach did not look very
Steward Roddey, the general manager of Oakland A’s baseball team is faced with the decision of whether or not to give a hike to Mark Nobel, the second best pitcher in the American League. Nobel’s agent was commanding a contract fee in the region of $600,000 per year owing to his performance statistics from the 1980 season. One major argument presented by Nobel and his agent is that Nobel has the ability to attract crowds and thereby increase attendance to the games and drive ticket sales. The agent quoted a figure of $105,650 as the amount lost as revenue when Nobel did not start.
To answer the question of whether or not the presence of Mark Nobel increases ticket sales, we have performed a regression analysis of the ticket …show more content…
Hence, this variable (O4) is very significant in determining the no. of tickets sold.
Model 3: Multiple Linear Regression Model between No. of Tickets Sold and all the identified predictor and dummy variables (as classified in table 1)
Exhibit 4 shows the detail analysis of this regression model.
The regression equation we get is as follows: Ticket Sold = b0 + b1 . POS + b2 . GB + b3 . TEMP + b4 . NOBEL + b5 . WKEND + b6 . PREC + b7 . DAYTIME + b8 . TV + b9 . PROMO + b10 . DH + b11 . O1 + b12 . O2 + b13 . O3 + b14 . O4 + b15 . O5 + b16 . O6 + b17 . O7 + b18 . O8 + b19 . O9 + b20 . O10 + b21 . O11 + b22 . O12
Though the R-square value obtained from this model is quite significant (0.7939), p-values of many predictor variables fall outside the significance level (95%). Hence, we further need to optimize the model by eliminating the variables with high p-values and taking into consideration only those variables whose p-values are within the acceptable range.
Model 4: Forward Selection …show more content…
This shows that the addition of Nobel variable has essentially lowered the R-square value and hence the addition of Nobel to model is not significant.
It could also be noted from the very high p-value of Nobel (0.45644), which is quite high and more than 0.05 (Significance level assumed to be 95%). Hence, this further confirms the analysis that variable (Nobel) is not significantly determining the no. of tickets sold.
Model 6: Using Interaction variables
We tried to model the regression using interaction variables. For this we considered the interaction between the variable NOBEL and each of the other predictor variables, including dummies (Exhibit 9). However, as our data did not contain any continuous variable except for TEMP, GB and POS, we did not find any relationship between the interaction variable and the dependent variable. Hence this approach did not look very