by PuckStopsHere on 06/14/12 at 03:45 PM ET
Yesterday I began my sabermetrics and hockey series for this summer by posting the top 20 players by adjusted +/- ratings last season. At number one on the list is Patrick Sharp of the Chicago Blackhawks. He posted a +24.2 rating to beat out Patrice Bergeron of Boston by 0.4 points.
Sharp had a good season last year. He scored 69 points and was the second highest scorer on the Chicago Blackhawks. He is a responsible defensive forward, though he is a step behind the serious Selke Trophy candidates. What exactly does it mean that he leads the NHL in adjusted +/-?
Sharp is clearly a good player, but this stat does not show that he is the best player in the league. In fact it isn’t expected that the top adjusted +/- rating in the league should be the best player in the league. It should be a good player who succeeds in his role, but it is a relatively rough indicator of player quality. There are several factors that can influence results. The biggest is the nature of +/- ratings themselves. A +/- event occurs whenever an even strength or shorthanded goal is scored. This happens a handful of times per game. A lot of information exists in hockey games that is lost by only collecting a handful of events per game. The information that is collected is the players who are on the ice when a goal is scored. All players on the ice on the scoring team get a +1 and all players on the ice on the team that allows the goal get a -1. It is clear that all players are never equally responsible for every goal, but this is the safest way to record the information without introducing a scorer bias ( Dave Staples tries to score the players who are responsible for the goals but I think his scorer biases make these numbers harder to draw conclusions from than the raw numbers).
As a result of the lack of events and the fact all events are treated as equal, when clearly they are not, it is expected that there should be some “experimental error” in these numbers. The top player in the league may not have the top ranking. Likely the top ranked player is the top ranked player because of some of the factors not properly accounted for in +/- ratings. This Patrick Sharp ranking is an example of this.
Sharp plays on a Chicago team with a good group of frontline forwards. They have Jonathan Toews, Patrick Kane, Marian Hossa and others dividing up the ice time with Sharp. More often than not Toews, Kane and Hossa play as a time together. This line faces the top opposition for the Hawks. This frees Sharp to play against lesser competition. I think that is the main reason that he gets the top adjusted +/- rating in the league.
Sharp is a top player on a good team. With an adjusted +/- system he is compared to an expected value for a Chicago player. Sharp plays against easier opposition than the other top Chicago players. Thus he will get a better +/- than they will if he is equally good (assuming all other things are equal). It is that advantage that makes Sharp the top adjusted +/- rating in the league instead of his having a good rating as one might be more likely to expect. Clearly it shows that Sharp succeeds in his role, but it also shows that his role is easier than that of the other top Chicago forwards.
Add a Comment
Please limit embedded image or media size to 575 pixels wide.
Most Recent Blog Posts
About The Puck Stops Here
Who am I? A diehard hockey fan.
Why am I blogging? I want to.
Why are you reading it? ???