Want to get involved? We're always looking for ideas and content for Weekly Challenges.
SUBMIT YOUR IDEAThis is my proposed solution: I performed association analysis using Pearson's correlations. Next, I manually selected those fields and modelled the complete dataset using only those predictors - and I explicitly disregarded any multicollinearity, assumptions checking etc. Subsequently I used that model to score the teams of interest and added two more fields: [Projected Losses] = Round(162 - [Projected Wins], 0)
[Games Back] = Round([Max_Projected Wins] - [Projected Wins], 0)
This shouldn't be the best way to perform this type of analysis as I'd prefer working with stepwise elimination and VIF and checks for multivariate normality... However, my results match with the solution... I'd suppose I should get a badge for this
Solution attached
Great Challenge for learning the predictive tools!