The National Hockey League Entry Draft, 1969-1995: An application of a weighted pool-adjacent-violators algorithm

We use NHL statistics for players selected in the National Hockey League Entry Draft to predict performance, which is measured by career games played, as a function of the player's position, year, and overall rank in the draft. This is done by inequality-cons trained least squares using a weighted pool-adjacent-violators algorithm. The results are used to examine the value of having the ith versus the jth overall pick, to rank the drafting success of franchises, and to evaluate the performance of European players. The number of games actually played by European players drafted in the 1988-1992 period notably exceeded their expected number, as determined by the model.