When looking at whom to draft come mid-summer, an important factor to consider is looking at the market of the league you hope these prospective players will one day be a part of. This line of thinking applies as well when you are looking to acquire players through free agency and trade. Everything comes down to value and production so you need to ask questions about the modern market that will help direct you towards the most ideal way to acquire assets. One of the common indicators of value in sports is position scarcity. While hockey doesn't have the amount of positions of a sport like football or even baseball, positional value does exist. At Hockey Prospectus, we've looked at the goaltending market at a pretty in-depth level, so I think I can safely discount goalies from this research as their value is at a low point due to how inseparable most of the market is from one another. Today though we will be looking at forwards versus defense and which positions are scarcer at which production points.
The common thought process nowadays is that the NHL has essentially leaned towards being a forwards-league so I decided to look at production by position to see if this was true. This graph shows how many players at each position produce at a given GVT/Game. It is also normalized for the fact there are many more forwards than defensemen in the league. The data used is using this current season and the 2009-10 season combined and there is a 30 games played minimum. There is a three period moving average because some points on the X axis had a very small N.
A couple of observations:
At the highest production points in the league, forwards stand alone. That is where players like Crosby and Stamkos produce and there simply isn't a single defenseman over the last two seasons who has had a GVT/Game over 0.30 while there have been nine forwards. While the sample is small, the difference between the top forwards and defenders is significant. You simply won't get the highest possible GVT/Game production from players on your back-end.
After the elite level when we start getting into the perennial all-star candidates and the top line/top pairing players there is a larger number of forwards than defenders. However the better way to look at that would be to see that defenders are scarcer at that production point than forwards. Putting it simply, it is easier to get a forward who produces 0.20 GVT/Game (Giroux, Burrows, Sharp, Clowe) than it is to find a defenseman (Lidstrom, Letang, Doughty).
As we move towards the meatier part of the population at a much smaller production point, defensemen significantly outnumber forwards. This is for league-average players such as second- and third-line forwards and mid-pairing defenders, amongst others. While forwards are usually the best route to go if you want high-end production, league-average production is easier found through the blue-line and is a scarcer resource when found in forwards.
For players providing zero or negative value, forwards outnumber the D again. This is likely explained through the fact that this production point is where most of the league's enforcers and guys who ride the bench reside and that role has become reserved exclusively for forwards. Coaches can't afford to have a defender ride the pine all game.
So if you wanted to look at this graph from a very simplistic view of how to draft, the answer in regards to which position to draft and when would be answered by looking at Iain Fyffe's column on Peak GVT by pick.
The first selection or two generally should be a forward if you can project them as possible elite players as you won't find that kind of production from defenders. If there is a defender available in the top handful of picks who you'd project as a top-pairing quality player, select him as the projected Peak GVT after the fifth pick drops below the two win above replacement level. Then from that point on in the first few rounds just go forward-crazy as most likely you will be getting an average player at best if anything.
This of course is not how you'd approach a draft as there are many more variables involved, such as a team's current system/organization, the draft class and the player's individual talent level and projected development path. Young forwards also tend to produce at a higher level quicker than defenders during their pre-UFA years. There also is the market factor in terms of how the positions are paid. All these variables are subjects to be examined further, as this was simply an introductory piece.
Corey Pronman is an author of Hockey Prospectus.
You can contact Corey by clicking here or click here to see Corey's other articles.