How much weightage should we give some of the new analytical stats for judging the performance of players?

I would not give too much weight to the new basketball player analytical statistics (stats) for judging the performance of players. One of the more commonly used stats is called Player Efficiency Rating (PER). It is supposed to measure a player’s per minute productivity but many misinterpret the stat and believe it can be used to determine a player's ranking in the NBA from best to worst.

Another commonly used stat is win shares, designed to estimate a player’s impact on their team's win total. Both PER and win shares are based on complex mathematical formulas and I simply don’t believe they should be given a lot of weight.