With expansion plans in the works, the NHL and the teams that comprise it are under pressure to increase revenue. Sports are entertainment and in the entertainment business, fun is money. It’s not rocket science to link a rise in the number of fans to a rise in the amount of money coming in. Every team in the NHL wants to win. Even the teams that have been rumored to have tanked for attractive draft picks want to win, albeit just a little bit farther in the future. All things considered, winning, or even the hope of winning, is good for the bottom line.
Hockey teams are businesses comprised of human beings. Most human beings, like it or not, allow their feelings about a person to dictate how much latitude and gratitude they are willing to give. If a player is well liked, works hard and listens to the coach, the coach will often like that player. That player will get more latitude for the mistakes he makes and more gratitude when he does something well.
The coach’s feelings about a player will often direct his eye away from the player’s deficiencies and toward his strengths. This works the opposite way as well. If the coach is not very fond of a player for whatever reason (preconceived ideas of the player’s skill level, work ethic, personality, etc…), he is more likely to see that player’s mistakes or shortcomings than his talents. This is often referred to as “confirmation bias” and it is just basic human nature inside and outside of sports.
The problem with this part of human nature is that it leads to decisions based on skewed information. Looking at this through the lens of hockey as a business, this can have a negative impact on the bottom line. The business of sports is about gaining a competitive advantage and confirmation bias often hinders this. Gaining a competitive advantage when management and coaching are not on the same page can be a monumental chore.
If team management thinks a player is a good investment, but the coach is not willing to play him, there has to be a reason. If team management thinks a player is a poor investment, but the coach favors him, there has to be a reason. Perhaps it is that the player just isn’t as skilled as the GM thought and his play after being acquired shows that, or perhaps the coach has formed an opinion of the player that is skewing his judgment of that player’s strengths and weaknesses. Whatever the reason for the disconnect, something is being lost in translation between team management and the coach.
You’d be hard pressed to find a coach in the NHL who doesn’t want to win. I think it is safe to proceed under the assumption that every coach is doing what he thinks is the best for his team. Team management obviously has the burden of financial constraints to worry about, but I think it’s safe to assume they want to do the best they can for the team given what they have to work with in terms of a budget.
If everyone involved wants to do what is best for the team, then why is it that the coach seems to favor using Player X when management traded for Player Y with the goal of filling a deficiency on the team? Every team has an example, or ten, of this situation each season. When the gamble taken on a player is small, this kind of disconnect often isn’t that big of a deal. When the contract that player holds has a more substantial hit on the salary cap or the player costs a significant price in terms of draft picks, prospects or roster players, the disconnect is a big deal.
Regardless of the reason behind this phenomenon, teams cannot afford to have this happen very often in a hard salary cap league. When competitive advantage is getting slimmer with the passing of every season, teams must work to be as efficient and creative as possible in getting value from the players on the roster and in the system. The negative impact of human nature’s built in biases need to be minimized to do that. As we all know, statistics, in their many and varied forms, can help accomplish this.
At this point in the evolution of hockey, teams that are not known to have invested in statistical analysis are often looked at as being behind the times. Teams that have publicly announced investments in statistical analysis are often viewed as smarter than the others.
Here’s the issue: regardless of how advanced, predictive, unique, or amazing the statistical analysis out there may be, if the coach or management do not get a sense of its usefulness or value, it won’t make any difference to the team.
A team can hire the smartest, most renowned statistical thinker and all of it will be for naught if the coach, GM and analyst aren’t speaking the same language. This is by no means meant to suggest that coaches and GMs are incapable of understanding statistics or analytics or whatever the favored term of the day is, or that statisticians are incapable of understanding the game of hockey. What it means is that the people in these roles often speak different languages.
Much of the predictive modeling and cutting edge mathematics based work being done in hockey is being carried out by people who specialize in those fields. Building a predictive model is not easy and probably not something that is in the every day skill set of many of us. Mathematics, probabilities, and the like are their language.
In hockey, the coaches have often spent much of their lives learning and living hockey every day. They have built relationships with players to understand how to motivate them. Their knowledge of the game and managing players is often not in the every day skill set of many of us. Hockey is their language.
Hockey and statistics are two different languages. People are far more likely to be persuaded by information they can grasp quickly than by something they have to read or hear repeatedly to get. It has nothing to do with intelligence; it has to do with comfort and familiarity.
You may have learned that War and Peace is an acclaimed novel, but if someone gives you a copy of it printed in Russian and you speak English, you aren’t going to get much of Tolstoy’s message. You could, of course, take the time to learn Russian in order to appreciate the book, but you could also just get a copy of it in your native language.
In order to exploit the competitive advantage that using statistical data can provide in hockey and minimize the impact of confirmation bias (human nature), the data has to be translated into actionable information. The data has to be put into a practical, usable form. Simply telling a coach that Player X is better than Player Y is not going to persuade him to change his thinking or challenge his opinion of a player. It is also not going to help the coach determine what the player can improve upon, identify a player’s strengths and weaknesses, figure out a way to maximize or minimize their impact, or decide whether a player’s use within a system needs to be adjusted. It is not going to help a GM identify players who will provide value to the team or identify needs the team has that must be addressed in order to improve.
Statistics, as a competitive advantage, are only an advantage if they are useful to the team. The whole purpose of statistical analysis in terms of coaching is to give the coach the ability to make informed decisions. If the coach can’t envision a way that the information will create tangible results on the ice, he is not going to be persuaded. The coach has to be able to trust that the analyst knows what he or she is talking about if there is to be any hope that the coach will trust the information.
Combining the statistical language and the hockey language into something actionable requires the analyst to not only understand both languages, but to have the ability to derive and translate the meaning from one language to the other. This will go a long way to establishing that trust and as a result, providing a real competitive advantage.
There are many possibilities for using statistical data in hockey. Provided the knowledge is there, the only real limits lie in the time the person has to put into it and his or her creativity. Every day incredibly smart people are applying their knowledge of mathematical and statistical theory to hockey. This is the way that our knowledge base will expand and grow.
Experimenting with different metrics and theories helps identify predictive and useful data about the sport. When coupled with the people using their knowledge and creativity to come up with ways to track different aspects of the game, the result is a set of tools that organizations can use to analyze and evaluate the team, players, prospects, systems, trade targets, and even their opponents.
The use of statistical analysis in hockey isn’t an attempt to reinvent the wheel. Statistical analysis gives us a set of tools we can use to analyze, evaluate and improve the wheel. The more the practical value of these tools is made apparent, the more coaches and GMs will be persuaded to use it to the advantage of their teams and thus, their bottom lines.