To test the model's potential, data from the 2015-2016 NHL regular season was analyzed. When comparing the projected number of points and the actual points accumulated, the average difference was a mere 3.9 points. In an 82 game regular season, where the average team earns about 91 points, a 3.9 point difference is relatively insignificant. 

Continuing to look at the 2015-2016 regular season, the model would predict 15 of the 16 playoff teams successfully. The intentions of this model was to create an accurate and conclusive model. Using previous year's data as a test, this goal seems to have been reached. The model accurately predicted which NHL teams would make the playoffs, using only 5 simple metrics. This result, though interesting to many in hockey, is most useful for coaches.

When I envisioned this model, I pictured a tool dedicated to fans. I anticipated the most interesting result being the ability to tinker with lineups, and then watch was happens to a team's predicted success level. However I didn't expect this model to help identify concrete goals. For example, the Colorado Avalanche failed to reach the playoffs in 2016. They earned only 82 points, 5 points short of the last team to make the playoffs. This model allows the Avalanche to assess what needs improvement. In this specific case, their Fenwick For % was very low compared to playoff teams. Therefore, coaches and management alike can prepare for the next season by emphasizing shot generation, and even making personnel changes to facilitate this improvement. The most notable takeaway is that the Colorado Avalanche staff can provide their players with numerical goals: Reach a Fenwick For % greater than 48.5%, raise the power play % to 19%. Both of these changes, assuming all other metrics remain constant, would yield a predicted point total to reach the playoffs.