Monday, September 19, 2005

A Loss is Not A Loss

For the last few years in the NHL, there have been teams given a point for losing hockey games (as long as they lose them in overtime). For those teams, a loss is as good as a tie (its worth the same number of points). As a result, there are a few more points available in the season. Sometimes, teams can finish above ".500" and actually have more losses than wins. For example, in the 2003/04 season, the Buffalo Sabres had a record of 37-38-7 for 85 points (with 4 OT losses) and the Minnesota Wild had a record of 30-32-20 for 83 points (with 3 OT losses). These teams in the NHL propoganda claim to be better than .500 teams when they lost more times than they won. Ideally, statistics should show reality and not hide it to make you think your team is better than it actually is.

This season, it is going to be even worse. Now goalies are not going to be creditted with losses when they lose a game in overtime or in a shootout. Goalies too can be above .500 and lose more often then they win. We can be deceived into thinking our goalie is better than he actually is due to the misleading propoganda statistics. It no longer matters to the NHL that players will get more due than they deserve, with cost certainty, they cannot actually get paid more money. The NHL can deceive fans into thinking that their goalie is better than he actually is, and it will not cost them anything at contract time.

Trying to justify this move, in a TSN story they have a quote from Martin Biron of the Buffalo Sabres

When you look at some guys' win-loss statistics sometimes, it doesn't truly represent what he did for his team.

This statement is entirely true, win loss records do not reflect how good a goalie is. Goalies on good teams tend to win more than goalies on bad teams. It is quite possible that a goalie could have a season good enough for a Vezina nomination and lose more games than he wins. Roberto Luongo did this in 2003/04. This has nothing whatsoever to do with overtime and shootout losses. It merely shows that win-loss records are a poor statistic to rate goalies. One attempt to rate goalies is shown here. This method is not perfect and does have to use wins and losses to some degree, but that is more representative of the fact that rating goalies statistically is a hockey sabermetrics problem. Making goalies win-loss percentage slightly increase by including overtime losses (which are not actually losses in the new treatment) only serves to further obscure the problem and it does it in a way that makes goalies look better than they are with no cost to the NHL.

Comments: Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?