Is there a good way to measure the success of General Managers at drafting players into the National Hockey League? Games Played alone can be a flawed statistic. Your results will tend to be skewed by what part of the draft order each GM has been making his picks. Steve Tambellini can look like a far better drafter than most because he's made more 1st overall picks than any of his counterparts. You have to be careful not to reward GMs who built bad teams, lost a lot of games, and made higher draft picks.
Players drafted by bad teams also have a better chance of making the team on his entry level contract, versus a perennial playoff team that has fewer roster openings. Some GMs believe in rushing prospects to the NHL, others believe in letting them mature in the minors. You can have two identical players drafted by different GMs at opposite ends of the rush/ripen philosophy spectrum who are equally good by age 23 but one has played twice as many NHL games. In this case one pick was not better than the other, but using only GPs one appears to be twice as good.
For my money, the best way to determine the value of the draft pick is the average annual salary of players after their entry level contracts. It's easy to estimate the expected future non-entry level average annual salary of each NHL Draft Pick, and it looks something like this.
For each individual player, we can estimate how much they are above or below the expected value of one drafted in that spot. Each season after a player is drafted, their value changes based on performance. How can we estimate changes in future value each season? Special thanks to the Hockey News annual NHL Future Watch magazine. I've got the complete set going back to 2006. The expected future non-EL salary of a player ranked #1 in FW is just under $5M. The 75th ranked is just under $1M. And so on and so forth. For players on entry level contracts I estimate their future value, after that their current AAV salary is their value. Since some of them are still active, I used their most recent value up to the age of 26.
Now that I've explained the method, it's time to ask how much the draft picks of each NHL General Manager have exceeded or missed expected value from 2005 to 2013. Salary data is current to the end of 2015/16 (I'll update my spreadsheet with this current season's stats when it's finally over at the end of June). The ranks are based on total sum of value minus expected value, divided by the number of draft picks. Only GMs who made at least 20 draft picks are included. Note that goalies and players drafted over the age of 20 are not included in these numbers.
|RANK||NHL GM||Picks||Beat E[V]|