by PuckStopsHere on 09/24/11 at 04:38 PM ET
I have taken a bit of a break from my sabermetrics and hockey posts, but there are still a few points I want to revisit before the season opens. I had last been looking at goaltender point shares and posted the top 20 seasons by goaltender point shares. At the top of the list all time is Roberto Luongo’s 2003/04 season with the Florida Panthers. This season is worth 20.85 point shares. No goaltender has ever had a better season according to this system. How reasonable result?
Luongo did not win the Vezina Trophy that season. That award went to Martin Brodeur. Luongo was nominated, but he finished third in the race behind Brodeur and Miikka Kiprusoff. If Luongo had as historic a season as the point shares system says, the voters did not notice.
Luongo was the star goalie in Florida. In 72 games played, he posted a .931 saves percentage (this placed him third in the league behind Kiprusoff and Dwayne Roloson but they played 38 and 48 games respectively giving Luongo an impact in more games). He posted a somewhat middle of the pack 2.43 goals against average. Perhaps most damning to voters was the fact that he posted a losing record of 25-33-14 (ties still existed back then). Luongo set an NHL record (that still exists today) with 2303 saves made that season.
Luongo’s Florida team was not a very good one. Olli Jokinen was the only player to exceed 40 points (he scored 58). Their ice time leader on defence was Mike Van Ryn, who was also the team’s second highest scorer. This Florida team finished 12th in the East Conference with 75 points and they had Roberto Luongo as the main player to thank for staying out of the league cellar. Luongo was a very heavily worked goaltender and he did very well. This is the kind of season that the goaltender point shares system looks for.
There is a problem with the normalization of point shares to points that teams get. These numbers only approximately link up and in extreme cases, such as the most heavily worked goaltending performance ever, they tend to break down. There is too much importance given to goaltending in this extreme performance.
The problem is illustrated by this example. If a goalie makes 19 saves on 20 shots, his team probably wins a game and he deserves a significant piece of the credit (to be fair exactly how much credit depends upon shot quality - which is not properly measured). If another goalie makes 38 saves on 40 shots (this is the same saves percentage as the first example), his team is a little less likely to win as they allowed an extra goal and likely possessed the puck less in the game, but it was still a good performance from the goaltender. In the point shares system, the second goalie gets more credit than the first goalie, even though the first goalie was more likely to actually post a win. The moral is that as goalies face more and more shots, the goaltender point shares method gives them more and more credit and they are less likely to have their team post points in the standings as this happens. Roberto Luongo is an extreme example.
Roberto Luongo had a very good season in 2003/04. Perhaps he should have won the Vezina Trophy, though he didn’t. However, it is laughable to claim that he posted the best season a goaltender ever has posted in that season. It clearly is untrue. He did have the highest workload of any goalie in NHL history that season. That workload of overrated in the point shares system. This is a significant problem that must be fixed in order for goaltender rankings from the point shares system to be taken more seriously.
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? ???