Although hockey is my passion, I foresee models like the one I've created this summer being used as effective scouting and coaching tools for a variety of sports. With the domain expertise of coaches in a certain sport, similar models can be created, as long as the data is available to do so. Given the increasing data and access surrounding sports, collegiately in particular, such models are realistic possibilities for the athletic programs on Bucknell's own campus!

As the team statistician of the Bucknell Women's Volleyball Team, I collect mass quantities of data used selectively by coaches. However, I now realize the potential of said data is creating overarching models to analyze the future success of the volleyball program. This season, my goal is to generate a predictive model for the volleyball team, in order to establish long term goals in terms of team statistics. The model may not improve the team's chances of winning instantaneously, but concrete goals can provide encouragement for long term improvement.  

Similarly, I'm familiar with the biostatistics tracked for the Bucknell Field Hockey team, and I am confident a similar project could aim to find optimal performance measures for the team. Collegiate athletic programs are ceaselessly searching for an edge over their opponents, and I believe optimizing performance models may provide this advantage.

As great as this may sound, I have experience first hand the difficulties of integrating analytics into sport. First, players and coaches alike can struggle to grasp the importance of analytical findings. Transparency and thorough explanations are effective ways to communicate the noteworthiness of one's findings. Additionally, coaches can be hesitant to accept new methods of analyzing play, and for good reason. Many coaches have spent years finding ways to get out the best of their players, so presenting data to coaches must be done gently. One way to go about this is to allow coaches to explore findings on their own. As long as they trust that the model is accurate, which can be emphasized to them, coaches can use the model to reach their own conclusions.  Explore below, and see what conclusions you can draw about NHL teams, and their keys to success!

Interact with the model itself

Below, readers can try the model themselves! Use either the sliders or the dialogue boxes to alter each of the five metrics. The resulting points and playoff likelihood will be the output of the model.