With the abundance of metrics tracked in an NHL game nowadays, analyzing and understanding the data itself is a difficult task. With the fluidity in the sport of hockey, many different stats, methods, and models are applied in hopes of providing valuable insight.
The challenge in presenting data to promote meaningful discussion is that it requires a ton of trial and error before an x vs. y comparison might result in a useful correlation. The amount of work many great hockey minds have put into processing data has elevated the world of hockey analytics, and educating oneself on what makes a player or a team good has never been more accessible.
When combing through stat sheets, it is easy to get lost in the vast tables of values without really having a starting point for what comparisons might have relevance. The ability to see multiple stats at the same time suddenly allows for qualitative analysis into how and why certain things may have happened over the course of the season. We wanted to experiment with a data visualization that was modular, where each visualization can be combined with different sets to paint a better picture. Data sets with individual game results can all be plotted and compared. The hope is that unique trends or segments can be immediately spotted. We chose to begin this analysis by looking at four basic metrics and present them in in a grid format with colour coded cells. Essentially, these plots were made to provide high-level visual summaries of what occurred.
The first set of data was the win-loss record of each team. The data was labeled with the six possible outcomes of a game: a win or loss in regulation, overtime, or a shootout, and each was assigned a colour.
1. Win/Loss Record
What was neat about looking at team records this way was that you can immediately spot the different stories that occurred over the season.
1. That monster Columbus Blue Jackets win streak? You can see that.
2. The collapse of the Colorado Avalanche? You can see that too. All. Season. Long.
The next three plots provide some insight into the records that are visible in the first.
2. Man Games Lost
Looking at the total number of man games lost on a per game basis can help account for large winning and losing streaks for a number of teams.
1. When Calgary’s franchise-tying 10 game win streak came to an end, it happened to be on a day when they had a season high 4 injuries. One of them was to starting goaltender Brian Elliott who was out with the flu.
2. The Winnipeg Jets, a team that underperformed last year, should take a giant leap this year as they should be significantly healthier.
(edited: Washington and Winnipeg rows were reversed)
3. The Boston Bruins, who did not have a very successful first half of the season where they were around .500 happened to be their most injured during this stretch. In the second half, where they regained their health, they surged to a winning record and into the playoffs.
3. Starting Goalie
This plot shows which goalie started on a given night. Whether it was the starter, backup, or a netminder lower on the depth chart, it’s easy to see how each team divided up their crease.
1. One of the most intriguing things about this plot is that for the most part, the top half of the league seemed to run with one bonafide starting goalie and a backup who seldom played. There are a few exceptions like the Ottawa Senators and the Anaheim Ducks, however.
3. Edmonton Oilers goaltender Cam Talbot, who many regard as the MVP of their historic season, was clearly the busiest NHL goalie. Starting 73 games last year, the Oilers own the highest differential of starts made by their number one goalie.
4. Home/Away Split
Each team may play an equal amount of home and away games, but their travel schedules can be drastically different.
1. The Pittsburgh Penguins never had more than three consecutive games at home.
2. In comparison, the longest homestand at 8 games belonged to the Minnesota Wild where they won 6 of them.
3. The Vancouver Canucks had the fewest number of switches between home and away games with 13. They also had the most extended (more than one game) homestands with their first solo home game coming in game 81.
These are only a few of many observations that can be made. If you’re interested in seeing a more detailed breakdown of the Calgary Flames’ season, please read our post over at FlamesNation!
Can you find any other cool trends? Is there any visualization of this type that you’d like to see or additional statistics you’d like us to explore? Let us know @wincolumnblog or in the comments below!