I got a couple of links from previously unknown (to me) stats curmudgeon Rory Boylen, who has a post up that starts with some some boilerplate:
"Statistics, of course, don’t tell the whole story. The only way you can get a real sense of how good or bad teams are is to sit and watch them play out their games."
But then he goes into full-on Jayson Stark mode, giving us all kinds of crazy stats with zero context:
- "The Vancouver Canucks don’t trail after the first period often, but it seems when they do, they end up losing"
Sample size = 8 games, 4 of which they tied up
- "The Senators have gone beyond regulation eight times this season, but haven’t won a single one of those games"
Welcome to small sample sizes, my amigo
- "The Sens are so bad beyond regulation they haven’t even scored one goal in the shootout. What is strange, though, is the Tampa Bay Lightning. With five wins in the shootout, the Lightning are one off the league lead, but their shooting percentage (25 percent) is ranked 25th."
The Sens performance was "so bad." The Sens are not "so bad." As for Tampa, again small sample sizes
- "Faceoff winning percentage is a key stat. A year ago the Canucks were seventh in the category, but the acquisition of Manny Malhotra has helped shoot them to the top of the league this season. But improved faceoff stats don’t always translate to winning: four of the top 10 teams in faceoff percentage last season missed the playoffs and half of the top 10 this season are on the outside looking in."
At least we have a true talent here. But remember, the gap between the best faceoff team in the league and the worst is something like 3 wins. Maybe 2. So it's no surprise that faceoff percentage doesn't drive making the playoffs.
- "If I were to tell you the Edmonton Oilers lead the NHL in giveaways, you probably wouldn’t be surprised. If I told you they had the most giveaways last season you’d probably have the same type of reaction. But would you be surprised to learn the Oilers have been one of the top two teams in giveaways (finishing second twice) each season since 2004? Even in the season they went to the Cup final, when they finished eighth in the Western Conference, the Oilers still led the league in turnovers by nearly 300."
On to rink bias. Edmonton did not lead the league in turnovers by nearly 300 in 2005-06. They were observed by human eyes to have led the league in turnovers. Perhaps sitting and watching games is not the best way to figure out how good a team is?
- "Who is the most physical team in the NHL? The Flyers? Bruins? Flames? How about the New York Rangers. Currently leading the league in hits, the Rangers have finished tops in that category twice in the past three years and were second last season."
More rink bias. The data from the people who watch the games at MSG is so bad we may as well not use it.
- "What’s surprising is the Washington Capitals and San Jose Sharks (the No. 1- and No. 4-ranked offenses from last season) have been held scoreless seven times each this season. The two teams, combined, were shut out only three times last season."
Back to small sample size.
- "Defense may win championships, but it has to get you to the playoffs first. The 11th-placed (in the West) Los Angeles Kings and the 15th-placed (in the East) New Jersey Devils have allowed the fewest shots-against per game (27.4). Other non-playoff teams Calgary and St. Louis sit in the top five in this category as well. At the other end of the spectrum, the Anaheim Ducks, Atlanta Thrashers and Boston Bruins (all playoff teams) allow the most shots-against per game. No wonder their goalies are in the running for the Vezina Trophy."
Saving the best for last, of course. Score effects, score effects, score effects. I wonder what Rory would do if he found out that the Edmonton Oilers were outshot every year that they won the Stanley Cup?
So if you've got a few minutes, give Rory a hand and let him know what's up...