Western Conference Defensemen: Pairing/Tier vs. Team Production

Dimitri Filipovic
August 05 2013 09:50AM

Johnny Oduya and Niklas Hjalmarsson probably aren't the 1st Blackhawks defensemen that come to mind, but their stellar play warrants recognition.
Image via Getty Images/Harry How.


On Thursday and Friday of last week, Travis Yost made his debut on this platform with a fascinating little research project. His goal was to take a look at how individual tiers of forwards for each team performed relative to their team's overall production, using even-strength zone-adjusted Corsi. Now, I've been tasked with doing the same, but for defensemen.

Why are we choosing to use "tiers", instead of terms such as "lines" and "pairings", which you are likely more familiar with? By doing so, we're able to account for things such as injuries, trades, call-ups, all of which naturally take place over the course of a season. There are so many little variations, and so much tinkering that takes place when it comes to specific usage, that it would really be a hassle to evalute things using those more commonly used terms. These tiers usually correlate very strongly with the most-regularly used combinations, anyways (i.e. the team's top two defensemen are more often than not in tier 1, and so on).

Just remember: context is crucial, especially when looking at data such as this. A substantial injury to a crucial player thrusts an unlikely candidate into a role (more specifically a higher tier, based on more frequent usage) that you may not have expected. But that's sort of the point of this. It tells us how well teams were able to perform with the pieces available at their disposal over an entire season.

Read Article | 12 Comments

Winning a Fight Has Impact on Future Outcomes

Travis Yost
August 04 2013 09:45AM





Back in July of 2009, Gabriel Desjardins at PuckProspectus tried to identify whether or not winning a hockey fight could be the impetus behind a boost in team scoring. He found virtually nothing.

More than four years later, I've found a much more significant set of data. The numbers may surprise you.

Read Article | 10 Comments

Eastern Conference: Line/Tier vs. Team Production

Travis Yost
August 02 2013 10:22AM



On Thursday, I took a look at how individual tiers for each Western Conference team performed relative to their team's overall production through even-strength adjusted Corsi. Today, the tiers of the Eastern Conference.

Read Article | 14 Comments

Western Conference: Line/Tier v. Team Production

Travis Yost
August 01 2013 10:41AM




One of the elements of hockey analytics that I think is understudied is the midpoint between individual and team-level data. Line combinations change an awful lot over the course of an eighty-two game season. Even in forty-eight, we see coaches -- be it injury-related, performance-related, whatever -- continually juggle trios of forwards on a game-by-game deployment. For this reason, I think a lot of people unfortunately shy away from it.

Read Article | 18 Comments

Machine Learning and Hockey: Is there a theoretical limit on predictions?

Josh W
August 01 2013 09:39AM

Graph from Arctic Ice Hockey 


In my previous work I was working on predictions in hockey, at the micro level of a single game, to determine who will win and who will lose. I was using Machine Learning techniques (or algorithmic modeling) and one thing I've noticed from my work, and from reading papers on other peoples material, is that for every sport there seems to be a prediction limit (that is different for every sport). This is what I want to explore; how much of the results that we see are based on luck, and is there a theoretical prediction limit in the NHL? 

In this post I will be focusing on the first question, what part of the standings is made up by luck?

Read Article | 7 Comments