As I continue my summer sabermetrics and hockey series, today I am looking at the top offensive zone starters. These are the players who most frequently are on the ice for faceoffs in the offensive zone as opposed to the defensive zone. These players are ranked by offensive zone starts minus defensive ones. This group of players are given a significant number of offensive chances and this will help them boost their offensive numbers. Unless they succeed to post significant offensive numbers this is likely a failed opportunity. These players may be weak defensively and are being protected from defensive responsibility or they may be strong offensive players given further chances to boost their offensive numbers.
This is useful in order to understand the context of how players are used in the NHL. This context will be used to adjust Corsi ratings. Players who start their shifts in the offensive zone are more likely to have positive Corsi events (attempted shots) in their favor than those who start their shifts in the defensive zone. This is a necessary adjustment to Corsi ratings in order to make them an individualized ranking of puck possession.
Here are the top 20 players in 2013/14 by excess offensive zone starts:
Dion Phaneuf was a star as soon as he entered the NHL. He was a Calder Trophy nominee in the strong class of 2005/06 where he beat out Henrik Lundqvist to earn a nomination alongside Sidney Crosby and Alexander Ovechkin. The next season he made the First All Star Team and was nominated for the Norris Trophy as best defenceman in the NHL. His career was then on a Hall of Fame track but he never arrived as a Hall of Famer. He failed to progress. Perhaps he had come too far too fast but he developed a bad habit of abandoning his defensive responsibilities to take offensive risks that rarely panned out. As Phaneuf became less defensively responsible, he was forced into a more significant defensive responsibility. Phaneuf played a tougher role and he played it less responsibly.
This is shown when we look at the worst players by team adjusted Corsi. Phaneuf has the third worst rating of -295.2. It doesn't help that his usual linemate of Carl Gunnarsson is worst in the league at -332.2. The fact that Toronto's top defensive pairing were among the worst puck possession players in the NHL last season helps to explain why the Leafs missed the playoffs.
Several days ago I listed the top 20 players by team adjusted Corsi. Today I want to look at the flip-side and see the 20 worst players by the same measure. Corsi is the difference between attempted shots (shots on goal, missed shots and blocked shots) for and against a team when a given player is on the ice. This is a measure of puck possession on the team level. In order to try to make it a more team independent measure, the first thing one can do is subtract off a team adjustment. That is what is done here.
This list has a problem where several players on the list got traded during the course of the season. It is not uncommon for players who are playing poorly (and hence have a poor Corsi) to be traded or waived during the course of a season. In that case, the team adjustment was weighted so that if a player played 60% of his games with team A and 40% with team B, his team adjustment would be 60% of team A`s and 40% of team B`s.
Here are the worst 20 players in the 2013/14 season by team adjusted Corsi:
Yesterday I posted the top 20 players by team adjusted Corsi ratings. This is a group of players who were among the best in terms of puck possession in 2013/14. Anze Kopitar of Los Angeles leads the pack, since I have already written about him; I will focus on the number two player in the list Marc-Edouard Vlasic of the San Jose Sharks and write a case study on him.
Vlasic has long been listed as one of the most underrated players in the game. It isn't clear if he should be listed as underrated anymore now that he won a gold medal with Canada in the 2014 Olympics. Vlasic was more of a depth defenceman in the Olympics, but given the depth of players available to Canada this is impressive.
Today I continue my summer sabermetrics and hockey series by looking at the top 20 players in 2013/14 by team adjusted Corsi rating. Corsi ratings are strongly team dependent. Players on good teams are more likely to have god puck possession numbers than players on bad teams. In order to team adjust a Corsi rating, you must subtract a team adjustment with is the team Corsi rating for the player's team divided by five since five position players contribute to this number. This is similar to the adjustment that Reif and Klein use on +/- ratings. The idea is to remove team effects and make the number an individualized ranking.
This number is not a final puck possession ranking because we have not yet adjusted for the role a particular player plays on his team. Those that play in defensive situations are less likely to have good puck possession numbers than those who play in offensive situations. That is still to come.
For the most part, the top players by team adjusted Corsi rating are similar to the raw, unadjusted top 20 although shuffling of rankings and some new players are introduced.
Here are the top 20 players by tam adjusted Corsi ratings in the 2013/14 season:
In the 2012/13 lockout shortened season, the New Jersey Devils had the second best Corsi rating in the NHL despite missing the playoffs. When Ilya Kovalchuk left to play in the KHL, many people picked the Devils to do quite poorly. I bucked the trend and picked them to have a solid season with the addition of Cory Schneider in goal. That didn't exactly happen. The Devils again missed the playoffs. They missed by five points last season and they again had a top Corsi rating. The Devils had the fourth best team Corsi in 2013/14. New Jersey seems to be the one team that most bucks the trend that good puck possession (as shown by a good Corsi rating) leads to success on the ice. Why is this?
Corsi is not the only thing required to have a winning team. Anyne who suggests this is probably making a strawman argument against Corsi ratings or totally misinformed. Corsi only measures puck possession at even strength. It does not look at special teams at all. Power plays and penalty kills are important but are not covered at all with Corsi. Corsi stats are only taken in 5 on 5 situations. Corsi also does not measure goaltending or shooting percentage. It only measures the number of shots attempted. We can get a look at goaltending and shooting percentage in part by looking at PDO - this is the sum of saves percentage and shooting percentage for a given team. New Jersey had the fourth worst PDO in 2013 and again they had the fourth worst in 2014. PDO is generally taken as a measure of luck as it is not repeatable on an individual player level. On a team level this isn't so true. A team with poor goaltending will have a poor PDO. New Jersey is an example of this. In 2013, their goaltending of Martin Brodeur and Johan Hedberg was the worst in the NHL. Last season, things were better. Brodeur played fewer games (though still nearly half of the season's games) and was no better than he had been the previous year. Schneider was an improvement on Hedberg and he played a larger percentage of games. This was an improvement, but the Devils remained below average in goaltending thanks to Martin Brodeur.
Yesterday I posted the team Corsi ratings from 2013/14. This is the difference between attempted shots (shots on goal, missed shots and blocked shots) taken and allowed in 5 on 5 situations. It is a measure of a team's puck possession.
The Los Angeles Kings led the NHL with a +630 team Corsi. Thus they attempted nearly eight more shots than their opponents per game. They were the best puck possession team in the NHL. The Los Angeles Kings also won the Stanley Cup. How much of a success of Corsi is this? How much of the Kings Stanley Cup came from their puck possession ability?
My next step in my summertime sabermetrics and hockey series is to look at team Corsi ratings. This is the difference between shots attempted (shots on goal, missed shots and blocked shots) by a team and their opponent in 5 on 5 situations. It gages the puck possession ability of a team, which is one of the more reproducible aspects of play if rosters remain the same. This is not a gage of how good a team is in that special teams, goaltending and ability to create good scoring opportunities are not measured, but if a team has a Corsi very different from their finish in the regular season it is quite likely their regular season result is not repeatable.
These team Corsi ratings are used partly to rank teams and partly as team adjustments to decouple individual players from team effects and thus make Corsi more of an individual puck possession ranking.
Here are the team Corsi ratings in 2013/14:
Yesterday I posted the worst 20 players in the 2013/14 season by their raw, unadjusted Corsi rating. This is the difference between shots attempted by their team (shots on goal, missed shots and blocked shots between a team and their opponents in 5 on 5 situations when a given player is on the ice. It is a strong gage of puck possession on the team level. In order to make meaningful comparisons on the individual level we must adjust for team effects and to a lesser degree the role a player plays on his team.
It is expected that the worst Corsi rating in the NHL will go to a poor puck possession player on a poor puck possession team. This is exactly what happened in 2013/14 as Carl Gunnarsson of the Toronto Maple Leafs was worst in the league with a -480 Corsi. This means that his opponents took six extra shots per game when Gunnarsson was on the ice than Toronto did.
Gunnarsson's "reward" for this season was a trade out of Toronto. He has been traded to St Louis along with a fourth round draft pick (Ville Husso) for Roman Polak. It's a trade fitting of the player with the worst Corsi in the NHL. A draft pick had to be added to him in order to get a defenceman who will be used more for depth than in a frontline situation.
I have started up my summertime sabermetrics and hockey series by listing the best 20 players in the 2013/14 season by their Corsi rating. This gave a list of players who were on teams that had strong puck possession when the player in question was on the ice. Because of the team aspect, the list was limited to seven NHL teams and many placed several players on it. The players were strong players who helped their team control the puck, but the question of how much is the individual player and how much is the team remains.
Today we will look at the bottom of the barrel. When these players were on the ice, their opponents controlled the puck. Since this list does not have any adjustments for team effects, it will be limited to a handful of poor puck possession teams and we will again have the question of how much is the team and how much is the player. That is a question I will attempt to answer in the future.
Here are the 20 worst players in the 2013/14 season by raw unadjusted Corsi rating:
About The Puck Stops Here
Who am I? A diehard hockey fan.
Why am I blogging? I want to.
Why are you reading it? ???