The First Three Problems

Let us use the big12football2015 data set located at

https://rur.kvasaheim.com/data/big12football2015.csv

Let’s start by defining a new variable, spread, which is the points scored by the team less the points scored against the team:

spread = ptsFor − ptsAgainst

In these three problems, you should model the spread.

1. Model the spread as a function of the number of first downs made by the team, D1for.
2. Model the spread as a function of the number of first downs made against the team, D1against.
3. Model the spread as a function of both variables (use an additive model).

For each model, write two paragraphs explaining your results. The first paragraph will concern the assumptions tests. The second will concern the results of the regression (if the model is appropriate). Do not just give the formula for the regression line — interpret the coefficients.

Please note that you should limit yourself to the modeling techniques we have studied. This include OLS regression— as well as weighted least squares and median regression.

The Final Interpretation

Finally, start a new page and do the following interpretation.

Mike Gundy, the head coach of the Oklahoma State University Cowboys, would like to know if he should focus more on increasing the Cowboy’s offense (increasing their number of first downs) or on the defense (reducing the number of first downs by their opponents).

In a final paragraph for this part, let him know if it is better for the OSU Cowboys to have more first downs or have their opponents get fewer.

I am looking for you to interweave the results (statistics) throughout your narrative, creating a compelling argument supported by the data (see example in §5.3.4). I am also looking for one or more graphics that tell the story of the data.

Do Not Forget

Do not forget to include your code in the appendix. I should be able to run your code and achieve the same results as you.