When analyzing hockey teams statistically, sometimes less is more. In the ‘Age of Analytics,’ all sorts of fancy numbers and equations have been trodden out to try to quantify teams. While much insight has been gained from these numbers, particularly the shots-based ones (SAT% and USAT%), simpler stats in large sample sizes are also insightful.
A time-tested and reliable statistic that reveals a lot about a team’s true skill level is goal differential. This does not only apply to hockey either, where coaching geniuses like Mike Babcock will tell you of its importance, but all of the Big Four sports, particularly baseball. Run differential is plugged into Bill James’s Pythagorean expectation to predict standings and squeeze out the luck.
Here we are going to look at the Coyotes regular season goal differential since the shortened 2012-2013 season and see how it compares to the rest of the of the teams in the NHL. I chose this starting point point because the management, coach and core of the roster have been relatively the same except for the loss of Radim Vrbata.
Each point represents an NHL team and three-year cumulative goal differential is on the vertical axis:
* Coyotes in Red
** 2014-2015 stats as of 2/21/15
*** Shootout wins are not credited as goals
The Coyotes have scored 458 and let in 538 goals in the past three seasons. What can be gleaned from this large sample size of goal differential? Well, the Coyotes have been a below average team. They are not terrible, but they have performed like a lottery team. A team cannot let in 80 more than it scores and expect to sustain winning.
Thinking about the past three years, these statistics are no big surprise. Maybe the Coyotes have been considered an average team, having an outside shot at the playoffs the past two seasons, but this season's performance and the underlying numbers for the past two do not grant them that status. If the Coyotes are going to really get better they need to accept their below average status and start building for the next few seasons and beyond.