Hockey on the Side: Shocker, great teams get great stats
Mar 24, 2014

Hockey on the Side: Shocker, great teams get great stats

I've had some fun slipping the occasional non-politics related post into this politics blog like some of my old piano recordings or the newer Tech on the Side series. Covering politics and politics only doesn't always do good things for my blood pressure, and so here goes for the inaugural Hockey on the Side. Besides, the emphasis on this blog is Canadian politics, so surely Hockey should be included.

Hockey commentary and opinion pieces are, far too often, trash. The problem often comes down to picking one's pet factors that explain the outcomes of hockey games, with little real basis. 

Dan Rosen's piece here prompted this response, and in it he details five reasons why the St. Louis Blues (who have clinched the playoffs in the West) are doing so well. This is an interesting topic, trying to determine what particular factors make a particular team more or less successful. However, it is quite a bit trickier than the article would suggest for a simple reason: teams at the top tend to do well over a wide range of statistics. Hence, simply picking a few pet factors where they do well in don't help to illustrate in a meaningful way why the team is so successful. 

Let's begin with his final point since it is the most egregious. The Blues apparently score the first goal more often than any other team, and manage to win the game given that they scored the first goal 3rd most often in the league. These are undoubtedly true stats, but do they tell us anything at all? Teams that are doing very well - teams that score lots of goals and let in few goals - are going to be more likely to score first and are going to be more likely to win having scored first. These stats are a consequence of the Blues being a dominant team, but they do nothing to explain - as the article ostensibly aims to do - what it is about the Blues that make them a dominant team. This is correlation does not equal causation 101. 

The article opens with the claim that the Blues crush in defence. It is true, they are third in the league for goals-against. But they are also third in the league for goals-for. As we would expect with a top team, they crush a lot of stats, and cherry picking one side of that doesn't help us. A lot of the defence related stats mentioned are all highly correlated and don't help defend the asymmetric preference for the one side. A team which is good on defence gets a good goals-against, but it also gets individual defensemen with high shot and corsi differentials.

 It also relates to goal tending where, because good goal tending is going to give good stats for defence and good defence is going to give good stats for goal tending, it becomes hard to sort out exactly which factors are dominating. Add in the laughably short 10 games for Ryan Miller that doesn't have the significance to determine the outcome of the team, and the whole pair of points on defence and goal tending just doesn't really identify exactly where the strengths of the Blues are by way of quoting statistics about the success of defence and goal tending (while entirely ignoring offence). The same type of points are true for power plays, his third reason. 

This isn't to say that there are not legitimate differences in playing styles where a team can focus more on defence (which involves the way the offensive lines play too) or on offence (which involves the way the defensive lines play too). Chicago scores by far the most goals, and ends up in 6th. Los Angeles leads the way in goals-against but third from the bottom in goals-for placing them in 8th. Those are examples of teams where on can really identify a unique (and opposite) playing style that has resulted in success for both but with a very distinctive focus. The Blues, however, are not one of these teams. They do well in both ends of the rink, and end up being second because of it. Boston is the same way (second in both measures, first over all). 

In general, hockey is a relatively well rounded game. There will always be asymmetries where some teams are more defensive focused vs offensive, some relatively better or worse in powerplays than five on five, some more reliant on a few key players (in particular goal tending) vs a wide bench, etc. But it is pretty hard to top the standings without being very good in wide number of different aspects of the game. So if you are sitting at #1, as the Blues are for the West, you are going to be likely to come near the top in quite a number of these stats. Picking out your favourite few stats and saying that is the reason for the success doesn't help us understanding anything. 

Thoughts on this post? Comment below!

Share this post:

Tweet It! Facebook Add Feed Reddit! Digg It! Stumble Delicious Follow

Post a Comment

Frequent Topics: