May 28, 2007

Luongo Should win the Conn Smythe

“Player Contribution is a method for allocating credit for a team’s performance to the individual contributors on a hockey team. More precisely, it is a way of allocating a team’s wins to individual players. It puts offense, defense and goaltending performance on the same page, in the same currency.

Defense has always been the toughest part of the game to isolate. The real power of Player Contribution is its ability to get at defensive contribution in a meaningful way. The way to do that is to isolate the component parts of the game. If you identify offense and goaltending, the rest must be defense. Study situational play to break defense down into smaller, more homogeneous pieces. Identify defensive responsibilities by position. Once you have done all that, the pieces become easier to understand and defense becomes easier to quantify.” - Ryder on Player Contribution


My Version
Ryder Version
If someone wants a more detailed explanation of the method there's a 71 page page document [PDF] explaining everything. I'll try to summarize some of the key concepts. There is one key decision made in the process of doing player contribution, and that is deciding how to split up the different parts of the games: offense (goals and assists), defense (minuses) and goaltending (save percentage). How important is the goaltender? Ryder argued that hockey has 5 players on each team who played offense 50% of the time and defense 50% of the time and the goaltender was always defensive. So he split it: 42% offense, 42% defense and 16% goaltending. I have problems with this line of thinking, as I find it confusing to think of a goaltender helping the team defensively when they've trapped the opponent at the other end of the ice. I also consider the fact that not all the forwards support their defenseman in the zone as there is often a forward high in the zone waiting for an outlet pass, which suggests that even when the team is trapped in their own zone offense is a primary part of some players at all times. I therefor prefer a 55% offense, 36% defense and 9% goaltending. Whether you agree with me or Ryder doesn't really matter, as I provide both data sets. I find my results produce better results that correlate better with what GM's pay for a give player's services.

2007 Player Contribution [xls - 550kb] [modified June 2, 2007]
2 for my versions of the regular season and playoffs.
2 for Ryder's versions of the regular season and playoffs.

The player who tops the list [Ryder Version] in total wins contributed is Luongo with an amazing 4 wins (Canucks won 5). Canucks scored at mostly marginal rates and their defense wasn't stellar, allowing a lot of good chances on Luongo, however they managed to win a series and a game against the Ducks. The performance was so great that it is unlikely that another player will catch Luongo in the remaining 4-7 games as the closest player in line is Giguere with "only" a score of 64 (3 wins).

I should note here: this stuff all suffer from the classic line problems associated with the plus minus statistics and even in goal scoring as a player cannot get assists if no one can score.

One win = 20 points.

Ryder's top 10:
1.Luongo: 81
2. Giguere: 64
3. Kiprusoff: 55
4. Miller 49
5. Hasek: 46
6. Emery: 38
7. Lundqvist: 29
8. Pronger: 29
9. Lidstrom: 29
10. Turco: 28

Therefor, as the most valuable individual player in the playoffs, Luongo should win the Conn Smythe. Of course this suggestion is crazy and there's no way a player who lost in the second round wins such a prestigious trophy, but certainly this shows how important Luongo was to Vancouver (and how pathetic the rest of the team was).

No comments: