Tuesday, November 22, 2016

How Predictive Analytics is Shaping the Future of Hockey

            Over the past decade, big data and predictive analytics has taken over the landscape of the sports, most prominently baseball, basketball, and football. One thing that all three of these sports have in common is the vast amount of raw data that each game produces. This makes it relatively easy for analytic specialists and mathematicians to create algorithms that produce actionable data. A sport that does not fall under that category is hockey, a sport contingent upon toughness and teamwork. Those two variables are extremely hard to tally as the article states that “The business of sports, however, can’t quantify the toughness of its players”. Nonetheless, hockey teams are attempting to implement big data analytics in their front office operations, just in a different form than most other sports. Teams have started hiring people from all backgrounds to head their analytics departments, while also hiring outside consultants. How teams are collecting data and what data they are seeking was the most intriguing aspects of this article in my opinion.
Unlike most sports, who collect and interpret data in similar ways, hockey teams take varying approaches, depending on how they value certain immeasurable intangibles. Kyle Dubas the assistant general manager of the Toronto Maple Leafs, who assembled one of the first “research and development” staffs in the NHL. His staff consists of people from an array of backgrounds including a chemical engineer. Their team (and many others) values blocked shots, which they incorporate with other stats including hits to compile ‘grit’ stats or shots to derive ‘real shots on net’ stats. Along with in house statistician teams, teams will hire outside consultants, some with no hockey background to assist them strive for analytical success. Teams have hired industry juggernauts such as SAP analytics to assist them with compiling relevant data and analyzing it to produce actionable results. Also, hockey is adopting analytics in a more broader sense than just at the team level.
As the NHL has always been the little brother to the NFL, MLB, and NBA, the league has taken measures to attempt to have competitive ratings with the big 3 sports leagues in America. The league is valuing fans input more than ever as fans are being presented with surveys and other forms of questionnaires to express their opinion on how the league to become more popular. Also, the league compiles data based on viewership to see which teams, match up and time slots are most likely to produce high ratings. This, along with the data from the fans can be analyzed by the league office to produce predictive results that can increase popularity and hopefully make hockey more popular throughout the world. I believe predictive analytics will become more prevalent throughout the hockey world as it attempts to cement itself as one of the most popular (and profitable) sports worldwide.  



  1. Technology and sports has become highly interesting topic in the past couple years. There have been people saying they can use data to help with the training of the players to prevent injuries, but I think this alternate route is a great idea for hockey. Even though hockey is considered a more aggressive sport, I believe analysts can collect data like they do in football because that is just as rough, if not more. There are many ways analysts or technological staff can implement this into the sport, it’s just a matter of getting people to notice hockey further and get the players to buy into the idea.
    As Kris mentions in the end of the post, they are trying to get players to take surveys, but many people dislike taking the time to fill put paperwork, so I think this could develop a bias within the data the analysts collect. Additionally, what if they could maybe track the skating paths of the players, or maybe add technology to the skates. If they put wearable technology in the skates, they would be able to see how much pressure they players put on the skates to change direction and maybe they could design a skate to better support the ankles of the players. Additionally, I think they could uses data to measure the force of the players when they get hit and hopefully design a safer arena, maybe with walls that could absorb the impact of the players. Also, I think analytics could play a huge part of the goalie position. They could use data to measures where the shots are placed, where they are taken from and the angles the goalie was placed at the time the shot was taken. I don’t exactly know how they will do this, but I think this could improve the skills of the players.
    Data is associated with many issues, such as security, storage capacity, maintaining qualified staff and updating the technology when it is needed. The biggest concern I would have for the NHL is how they would find funding for these project and upgrades to the program. Obviously, the NFL and NBA are programs that are extremely popular and highly funded, but would the NHL be able to compete with them. Some may ask why the funding needs to go to the NHL, which will be a huge obstacle. If they do not have the funding they will not be able to upgrade the software, which will be necessary due to Moore’s Law. Also, they would need to have a complex security system to protect this data from hackers and would have to purchase how amounts of storage space. I think this would be a huge benefit for the NHL, but if they lack viewers now, how will they find a means to pay for all this technology?

  2. Having grown up playing the great sport of hockey, Kris’ article is extremely interesting because of the way the teams, as well as the league itself, are employing data analytics practices to improve their own organizations, but also grow the sport throughout the U.S. and the world. I can certainly attest to Kris’ claims that from a pure statistics standpoint, hockey is unlike any other sport in which it is extremely difficult to translate aspects of the game into actionable data. However, I find the concept of “Grit Stats” intriguing and very exciting. Hockey players are arguably the toughest professional athletes in the world, historically playing through punctured lungs, broken bones and shattered jaws. The idea of tracking blocked shots, hits, 50-50 pucks wins, etc. in the hopes of deriving some sort of value from them is a phenomenal idea. These results will give front office executives insight into middle of the roster players. Guys that aren’t going to go out and score 30+ goals in a season, but will work their tails off as a role player and grinder to help their team make a deep playoff push.

    The second part of the article explains how the league is beginning to use various forms of analytics to advance the sport here in the U.S. and around the globe. League executives have begun to see a value in building interactions with the fan base past the in arena experience at weekly games. The league is sending out surveys and various other quantitative research methods to provide a platform in which fans can express their feelings and opinions on their current experience, and provide advice on how the league can make improvements. By engaging in these techniques, the NHL is positioning themselves to make huge gains in terms of popularity in the domestic market. With NFL ratings are at an all time low due to multiple factors regarding the league itself, the NHL is being presented with an outlet to propel itself into the upper echelon of sporting leagues and its implementation of big data analytics is what’s going to help them get there.

  3. Kris, I found this article quite interesting as I have been looking at how sports have incorporated the use of big data and how to fill the gaps in the output. As you said, hockey is different than the MLB, NFL and NBA because of the physicality of the sport and the way stats are recorded. Things like passing accuracy, breakaways and big hits are extremely important in hockey but they are hard to record. One thing that I think is missed is how big hits affect the play. Hits are recorded in hockey like a tackle is in the NFL. However when you look at pure data it does not show the effect of that hit like it would in football by a stoppage of play at a certain yard line. Hockey is a continuous game that moves quickly much like basketball and for hockey the players toughness is a massive part of the game. Players like Shawn Thornton made their career by being the toughest guys on the ice and through fighting. Aspects of the game that are so mentally driven like toughness and big hits will be ignored because it is not possible to measure them accurately. It is possible to look at a player and know their toughness by watching them play and speaking with coaches. That is why scouts exist and in order to actually see the full picture of a player and what they can bring to the team, we cannot ignore aspects that play major roles in the game.


Note: Only a member of this blog may post a comment.