This summer in my look at sabermetrics and hockey, I have been looking at the Corsi Number. They are a good judge of puck possession that is much like +/-, although it comes with different strengths and different weaknesses.
I have ranked the best 20 and worst 20 player adjusted Corsi Numbers. This ranking is done treating Corsi as a counting stat accumulated over the season. It can also be looked at as a rate stat and ranked per minute of 5 on 5 ice time played. To that effect, I have listed the top 20 Corsi rates from 2008/09. Today, I look at the worst 20.
This summer in my look at sabermetrics and hockey, I have been looking at the Corsi Number. They are much like +/- ratings, except since they include all shots directed at the goal (including missed shots and blocked shots) they give much larger numbers with less random fluctuations in them, They show who is driving puck possession (those players with good Corsis) and remove the element of goaltending from +/-. When players play defensive roles that do not drive puck possession, their Corsi is not always a good method of rating their play (with the best example of this being Jan Hejda). Thus far, I have only looked at Corsi Numbers treated as counting numbers and not as rates (per minute). This is how they are treated at behind the net.
Here are the top twenty players (with 50 games played or more last year) ranked by Corsi per minute of 5 on 5 play:
Probably the least logical results I have found in this summer’s look at sabermetrics and hockey surround Jan Hejda of the Columbus Blue Jacket. Hejda was among the Columbus Blue Jackets leaders in ice time with over 22 minutes played per game. He is not an offensive star as he contributed only 21 points last season. Most of his value is defensive.
If we look at the top 20 adjusted +/- ratings from last season, we find Hejda’s +19.4 is good for sixteenth best in the league. That is a clear sign that he has been a good player. However, if we look at the worst 20 adjusted Corsi Numbers from last season, we find Hejda’s -155.8 is nineteenth worst in the league. That suggests he has been a bad player and clearly contradicts his +/-. How can that be?
In this summer’s look at sabermetrics and hockey, I have been looking at the Corsi Number as an alternative to +/- ratings. Today, I am listing the worst 20 adjusted Corsi ratings from 2008/09. The adjustment is done in the same manner developed for +/- ratings in The Hockey Compendium by Jeff Klein and Carl-Eric Reif. I discuss this adjustment method here. In this method, a team adjustment is calculated from team Corsi Numbers. Since five players are on the ice, team Corsi Numbers are divided by five to give a baseline team value that is treated as a “zero” for that team. All individual player Corsi Numbers have the team adjustment subtracted off.
Here are the worst 20 adjusted Corsi Numbers from the 2008/09 season among players who played at least 50 games and with only one team:
When I listed the players with the worst 20 Corsi Numbers last season, the list included a lot of defencemen who play on bad teams against top competition and usually start their shifts in the defensive zone. This of course is a contributing factor to their poor Corsi numbers. Copper N Blue notices this and points it out. He criticizes me for listing the players with the worst Corsi Numbers, suggesting that it will provide ammunition to the “anti-numbers” crowd. There are people who do not want to believe that statistics and math can be used to understand hockey and would just like it all to go away. It is obvious that statistical analysis has made significant in roads to hockey, so it isn’t going away. It also isn’t clear that hockey can ever be solved on the level of baseball, but that is a goal - even if it is unattainable. This aspect of the game has little effect upon your viewing of a game, unless you want it to. There is no reason you have to know anything about statistics to view a game.
Because of their financial situation, the NHL is running the Phoenix Coyotes. This creates a situation where any Phoenix move can be interpreted in two different ways. It can be interpreted as having been done to suit Phoenix interests and it can be interpreted as trying to suit NHL interests. It is impossible to know for sure how important NHL interests are in Phoenix moves (but conspiracy theorists will maintain that they are quite large). As a case in point, we can look at Phoenix’s recent trade for Radim Vrbata. This trade was made with the Tampa Bay Lightning and cost the Coyotes Todd Fedoruk and David Hale. Vrbata is a player who had success in Phoenix. In 2007/08 he scored 56 points with the Coyotes. This prompted Tampa Bay to offer him a three year contract worth a total of $9 million. Vrbata was not given much of a chance in Tampa before the Lightning gave up on him. They got out of his contract by letting him go to the Czech League to play with Mlada Boleslav BK.
In this summer’s look at sabermetrics and hockey, I have been looking at the Corsi Number as an alternative to +/- ratings. Today, I am going to list the top 20 adjusted Corsi ratings. In this case I have adjusted them as a counting stat in the same manner that was developed in the Hockey Compendium by Jeff Klein and Carl-Eric Reif. I discuss this adjustment method here. In this method, a team adjustment is calculated from team Corsi Numbers. Since five players are on the ice, team Corsi Numbers are divided by five to give a baseline team value that is treated as “zero” for that team. All individual player Corsi Numbers have the team adjustment subtracted off.
Here are the top 20 adjusted Corsi Numbers in the 2008/09 season among players who played at least 50 games and with only one team:
The NHLPA is making news in the hockey world by suggesting that players skip the on ice portion of the Olympic orientation camps that will be held by the Canadian, American, Russian and German Ice Hockey Federations. Their public reason is that it is not clear that the players have insurance coverage should they be hurt during the camp. I think the private reason is because these camps are essentially useless. The Olympics will occur in February 2010. The camps occur six months early in August 2009. They are far too short and far too far in advance to teach players a new system (assuming that is a smart move with the best players in the world) or get different players used to playing together. They are not needed to get players in shape for the Olympics. NHL hockey players will be in shape and anything they do six months in advance has little impact. They have developed out of the administrative desire to do something, even if it is useless, to try to gain an advantage in the Olympics. In theory orientation camps may sound like a good idea, but because of the NHL season, they must be held six months in advance, which is too early to be useful. That doesn’t stop them from happening. They are an administrative way to cover the hockey federation. If a country fails to live up to expectation in the Olympics, they cannot be blamed for not doing enough to orientate players.
In this summer’s look at sabermetrics and hockey, I have been looking at the Corsi Number as an alternative to +/- ratings. Its appeal is that it includes a lot more events than +/- (as it is the difference between shots directed at the goal for and against when a player is on the ice in 5 on 5 situations). The drawback is that it includes missed shots and blocked shots as being equivalent to goals. Any systematic reason why a player or team might take or allow an excessive number of shots that miss that net (or some other anomaly like that) would skew the ratings. Yesterday, I gave the team Corsi Numbers from last season and it led to a very good discussion. Some teams have Corsi Numbers that do not correlate well with their regular season success. For the most part this can be explained (the comments to that post contain some of this discussion). For example, Boston has the 13th best team Corsi Number at +49 despite finishing first overall
In this summer’s look at sabermetrics and hockey, I have been looking at the Corsi Number as an alternative to +/-. Corsi Numbers are the difference between shots directed at the goal (shots on goal, missed shots and blocked shots) for and against when a player is on the ice in five on five situations. The benefit is that they encompass a lot more events than +/- does. However, whether or not it is a better or comparable series of events is somewhat of an open question.
I have calculated the Corsi Numbers for all 30 teams in the NHL. These can be compared to team +/- ratings.
About The Puck Stops Here
Who am I? A diehard hockey fan.
Why am I blogging? I want to.
Why are you reading it? ???