The exploration of linear best fit should look at questions like:
- how do you define "best fit"?
- are there multiple ways to "best fit" a set of data points?
- what are the importance and implications of this exploration and the answers to the questions above, on prediction and "predictive power"?
