Introduction

Hundreds, if not thousands, of ice hockey goalies compete hard each year to win for their respective teams. In the process, some catch the wandering eye of coveted National Hockey League (NHL) scouts who put them on a watch list. Some of the goalies who make that list eventually make their way to the official NHL Draft. Then, things really start to thin out. Although each goalie may be selected by a team come draft day, not all make the opening roster in their first, second, or third year. Some don’t ever get a chance to step on an NHL ice surface. Even fewer go on to play multiple seasons and fewer than that get a chance to play for the Stanley Cup – the NHL’s championship trophy. 

The journey of an NHL goaltender is a tumultuous one – full of mediocre highs and notable lows – it’s amazing to see that there are plenty of goalies willing to go through it all. As a goalie myself, I’ve always had a soft spot for someone playing the position. My respect for it brings me to research more about each player, their come-up into the NHL, and their potential to become one of the best. 

In my journey to find visualizations that highlight a goalie’s time in the NHL, from draft to championship, there hasn’t been any satisfying display of such a story. So, I’ve decided to go out and make one myself – a dashboard that would show the entire process. This dashboard isn’t meant to satisfy my own desires of acquiring hockey knowledge – but a tool for others to know and empathize with goalies. Not many people know why goalies enjoy being shot at by a rubber puck – fewer realize why they endure such hardship in a completely volatile career. I believe a dashboard showing this journey can help bring a more respectful understanding of the sport and the specific position. 

In my dashboard, I hope to share a coming-of-age type of story for the typical drafted NHL goaltender. My dashboard will answer…

1. Where are goalies mostly drafted from?
2. What do these goaltenders look like?
3. How many of those drafted goalies play an NHL game?
4. What does their career look like from their first season to their last?
   1. Are they successful throughout their career?
5. Which of these goaltenders eventually come around to win the Stanley Cup?

Inspiration

In order to create an effective story, the dashboard had to have the right components in the correct order. Considering the fact that reading left to right is a regular practice in the western world, I had to place my visuals accordingly. The story is a simple journey of progression through the ranks in the NHL – from draft to roster, to Stanley Cup. As simple as it sounds, there’s plenty of hard work and multiple failures one must endure between each step. However, my goal was to show just how this journey affected each goaltender who was drafted in the past 20 years.

Data journalism has become a recent common phenomenon in major publications. The New York Times is one of those publications that design data dashboards with neatly organized and detailed visualizations. One that I was inspired to hopefully replicate was their piece on steroids in the MLB. 

“Steroids or Not, the Pursuit Is On” was designed with an average reader in mind. The data visualization team clearly knew how their readers digested content. Starting from the left, a large Barry Bonds points up to the number “755” – both are the largest pieces of content on this dashboard. Since this is a place where readers normally start reading from, it was important to place the main insights of the dashboard in the top left-hand corner.  

From there, the details become finer and finer. The line graph in the middle of the main section helps guide the eye through a story of just how fast Barry Bonds, and other known steroid users in Major League Baseball (MLB), built a tower of home runs.

The story was extremely easy to understand and logically ordered to help readers guide through the journey. This example would be helpful in creating the ideal dashboard for showing just how NHL goalies over the past 20 years fluctuated in their careers.

Tools

Excel & Google Sheets

Both of these tools are spreadsheet applications that help store data in files that can be used for visualization efforts. In this project, Excel was used to store, clean, and organize the data. Google Sheets was used for gathering latitude and longitudinal coordinates with a formula script. 

Tableau

This tool is a free application analysts use to create visualizations from tabular datasets. I used Tableau to create, color, and order each visualization for later use in a finalized dashboard. 

Photoshop

The Adobe application is a tool to help create aesthetically pleasing designs. Although one could use Tableau to make a dashboard, I found it cumbersome and clumsy when trying to fit each visualization in. With Photoshop, I was able to freely move each graph around on an artboard and order it without making any substantial mistakes. 

Methodology

Gathering the Data

There was no singular location that contained specific draft data of goalies in one table. The draftee’s origins, career numbers, and deeper statistics had to be pulled from multiple sources. On NHL.com, I used season-by-season data points to compile how goalies performed each season. With this data, I could correspond a specific player’s success to exactly when they would start a season.

NHL.com also contained origin data that showed a goalie’s hometown, and draft year. What this set didn’t contain was the draft team, an important piece of information I wanted to use to show which teams these goalies were going to.

The rest of the data I needed was found at hockey-reference.com where I was able to find more in-depth data on goalie’s seasons. The site also contained draft data per year that I would use to combine all these datasets into one. 

Now, I had all of the data I needed located – goalie name, draft team, draft year, hometown, season-by-season stats, and junior leagues where each goalie played. Next, I had to combine the datasets into one that I could export into both Tableau, Gephi, and Carto.

Creating the Dataset

Before placing all of the data into Excel, I needed to gather the latitude and longitude coordinates for each player’s hometown. To do this, I used a script for Google Sheets that would match location data within a cell to the correct latitude and longitude coordinates. Once I gathered those points, I was finally able to create the datasets I needed in Excel.

The tables were all over the place and made it impossible to use an easy copy-paste method to combine the data. What was also troublesome was that some tables contained goalies that were drafted more than 20 years ago. To follow my timeline, I needed to also find the goalies who were drafted between 2001 and 2021. 

I then decided to split my data into two sets. One dataset would contain the origins of each goalie, the other would contain their career stats per season. That way, I could focus on only the points that were important throughout each step. 

Upon combining the data, there were no indicators on which season was the first season for each goaltender. To fix this, I grouped goalies together by name and sorted them by the season they began playing. I knew that since these goalies were all drafted between 2001 and 2021 that the first season recorded in the dataset was actually their first true season in the NHL. A goalie in this dataset could not come from a season before 2001 because they wouldn’t have been drafted then. 

Creating a Storyline

My ideal storyline was to show the journey of an NHL goaltender.

1. How did they get drafted?
   1. Where were they from?
   2. When were they drafted
2. What did they look like?
   1. How tall were they?
   2. How heavy were they?
   3. What was the average build?
3. What was their career trajectory?
   1. How many actually started an NHL game?
   2. What did their success in the season look like?
4. Which ones went on to win a Stanley Cup as a starting goaltender.

I wanted to test how this storyline played out before going through and creating the visualizations. That way, I could hone in more on what was truly impactful and important. So, I conducted a user test with a participant who works in the publishing industry as a full-time story editor. The participant would know which parts were added to the story and if the order seemed correct.

Upon pitching the story to the participant, there was a question on showing what these goalies looked like. “How does showing their average height and weight add context to the story? I don’t think it does.” Rightly so, the average build of a goaltender wasn’t a great representation showing a sort of coming-of-age story in the NHL. That section was taken off the original list and we were left with the following for the ideal storyline. 

1. How did they get drafted?
   1. Where were they from?
   2. When were they drafted
2. What was their career trajectory?
   1. How many actually started an NHL game?
   2. What did their success in the season look like?
3. Which ones went on to win a Stanley Cup as a starting goaltender?

Visualizing the Data

At first, I thought of using all 3 visualization tools we previously learned over this course – Carto, Tableau, and Gephi. However, I ran into issues that were unsolvable with Gephi where I could not properly export a network showing where most players are being drafted from. In hindsight, the Carto map for showing locations of players was merely a batch of individual stories of a goaltender making it to the NHL. The map didn’t add larger context to the story. Fortunately, Tableau had all the tools I needed to properly answer the main questions around what a goalie’s career is like coming into the NHL and competing in it. 

Draft Origins

Using the NHL Goalie Origins dataset I previously made, I used Tableau to create three separate visualizations – one showing year-over-year (YOY) changes in drafted goalies per country, one displaying a map of where goalies were being drafted from, and one sharing which amateur leagues were these goalies being drafted from. 

Map of Countries

By using the latitude and longitude coordinates, Tableau was able to identify which country each player came from. Originally, the map showed the entire world and consisted of countries that didn’t have any goalies drafted by the NHL. Instead of showing this, I decided to eliminate those countries and focus on those with actual data.

Junior Leagues

Upon uploading the data, I decided to use a bar graph to compare which junior leagues goalies were being drafted from. Instead of color-coating each bar by the league’s main colors, I used each league’s logo to help identify which bar belonged to which league. I then ordered the bards, top to bottom, based on the number of names coming from each league. This would help a user quickly understand which leagues had the most goalies coming from them.

In Europe, some leagues have Elit and Junior divisions. The main difference between these types are age and special rank. Sometimes, a player can be granted special access to an Elit league even if they don’t hit the age requirement. Those players end up being way better than their age groups and look to compete with higher-caliber players. These leagues still share the same logo, for Finland and Sweden, so to differentiate between Junior and Elit, I changed the color of the specific bars. The Junior leagues would have a light grey color and the Elit leagues would share the same dark grey color as the rest of the graph.

YOY Change in Countries

For this graph, I wanted to show the change each country compares YOY to how many goalies compete in the NHL. I decided to use a stacked graph and order each graph stack based on the number of players competing from each country. Countries with low amounts of goalies competing in the NHL would show up at the top and those with high amounts of goalies competing in the NHL would show up at the bottom.

How did they fare in the NHL?

For goalies, many hockey analysts argue about statistics that help identify what makes a superior goalie. The debate continues to develop as analysts continue to advance the technologies that can record data points per game and per season. However, most of the success metrics remain around the likelihood a goalie can stop a shot and the number of wins they can bring to a team. In these two graphs, I decided to use Save Percentage (S%) and Win Percentage (W%) over each goalie’s season to visualize their journey in the NHL.

• 

• 

To help understand just how much a goalie’s career fluctuates, I placed an average reference line for both SV% and W% on the corresponding graphs. This would be an index a viewer could use to see when goalies performed above or below the average. To add more context, I added in a goalie’s average age to each season number. This helps viewers understand that goalies don’t normally start when they are drafted at the age of 18. It also helps show just how old goalies are at each point in a season.

The lines for each graph also show how many goalies reach each season. The color gradient and size change depending on how many goalies play through multiple seasons. Viewers can see that the amount clearly thins out over time, which points out just how difficult the position is and how hard it is to maintain a steady career in the NHL.

Making the Dashboard

Before putting the pieces together, I wanted to conduct one last user test to see how viewers would ideally like to see these points represented. Considering that this dashboard would be more enticing to see for sports fans, I chose a former sports radio show host who also works in data science to help me finalize the representation of this dashboard. 

I showed the participant two separate ideas for creating the dashboard. One was an almost direct outline of the New York Times dashboard mentioned earlier, the other was a new approach designed like a funnel to accentuate just how a career looks for a goalie.

• 

• 

Out of both types, the participant selected the first option which was the funnel version showing just how many goalies made it through the ranks. The participant said that the order helped emphasize the understanding of what each visualization gave to the overall goal of the project.

Once I downloaded the visualizations from Tableau as PNG files, I placed them into Photoshop to start organizing the tiles and progression. Before each section, I shared a number of just how many goalies were included in each set of visualizations. This way, a viewer could follow along with just how many goalies were in each dataset and see more data on how the sample sizes got slimmer and slimmer based on career progression. At the end of the dashboard, I identified images of goaltenders who won the Stanley Cup and were a part of both datasets.

The Final Product

Reflection

Within this dashboard, a viewer is able to see just how volatile an NHL goaltender’s career is. We see that their numbers rarely progress over consecutive seasons – rarely are they ever over the average in SV% and W% two seasons in a row. We also realize those that get drafted never get a chance to play any minutes in the NHL. It shows just how a career as a goalie is not a straight arrow. Some goalies end up playing professionally in Europe or other smaller leagues for the rest of their careers. Others get too bogged down by injury to continue playing the game. For the select few who do make it, only a handful win a Stanley Cup as a starting goaltender. 

This was possibly one of the toughest projects I conducted within the design. There are so many methods to take on when trying to display information correctly – one has to take into account everything from how a person reads, to the psycho-physics that affect a human’s vision. Making a dashboard was the most difficult because of how each visualization was different. They were complex puzzle pieces that weren’t always the same fit. It took some user research to understand what a viewer would expect to see and how they would ideally run through the information. 

This dashboard is not perfect. It is far from the quality of the New York Times dashboard that I hoped to somehow replicate. I also think that using a tool like Photoshop, Illustrator, or After Affects could help amplify the default designs Tableau affords its users. To continue developing this visualization, I look towards using such tools to integrate the data and design a more complete, and aesthetically acceptable display of this information. 

With the dashboard how it is now, people can begin to understand and empathize with what a goalie goes through to make it in the NHL. When someone sees a visualization like this that explains statistically you are unlikely to be successful as a goalie, it’s interesting to note those who go through with this career knowing the likely outcomes. That’s just part of loving the game. For hockey players, accolades and success aren’t the primary focus. Each player focuses on continuing their career so they can keep having fun with the game they love.