Hi there was a fair bit of data cleansing to start.
Then it was a case of merging by player name; I used a Left join in SQL speak on the Overall Pick table - just in case a player did not exist last year (a rookie).
then we used a formula to trim down some columns into one such as age, team, position and name.
Lastly, I disagreed with the method they used to rank the players. Rather than doing too heavy DQ, my philosophy was to sum the rank whether it was hitter or pitcher by team and player type (new field). then Sum of rank / count of player for a given type of player would get you an average score by team by player type; then sum the score for the hitters and pitchers; and sort from lowest to highest; tagging each team with sort order ID.
This was re-joined with the clean dataset to give the 2015 team rank.
Cheers,
Philip
