In a previous blog post, I discussed in detail several Tableau hacks I used to recreate my Steph Curry dashboard. While these new techniques did contribute to the originality of this item, I feel it is primarily the design and layout of this dashboard that makes it stand apart from my other work. Trying to find this design balance was an iterative process that forced me to find compromises among data visualization, user experience, and design.
Style vs. Substance
Capturing the initial attention of the audience is important, but Tableau authors should still prioritize substance over style. This advice is especially true for informational dashboards, which rely on data visualizations to communicate accurate and easily digestible information.
Although fun and intuitive interactivity may seem like style and fluff, it’s an essential element needed to create a good experience for your audience. A well-thought out user experience keeps your audience interested and engaged, which increases the time and effort they are willing to invest when exploring your dashboard. Don’t give unnecessary reasons for people to leave your dashboard, especially not before they have an opportunity to interact and discover interesting insights.
The tile map and histogram technique, such as the one shown above, was created by establishing relationships between unique identification codes in the underlying data. This process is explained in more detail in the previously mentioned blog post. Essentially, these data relationships allow for interactivity to occur between the grid and corresponding histograms. The resulting interactivity permits users to flexibly explore and analyze the same data from different orientations.
Unique chart types and chart combinations draw viewers in and encourage them to investigate dashboards more closely. Tile grids with marginal histograms are uncommon enough to help capture initial attention. Novelty always seems to get bonus points from the Twitter community.
For more info about marginal histograms, please check out this excellent blog post by Steve Wexler.
Navigation / User Experience
As you can see above, the left sidebar contains four separate menus. Each menu is a different Tableau dimension that categorizes each shot made by NBA star Steph Curry in 26 games during the 2015-2016 season. When users choose an option from the menu, they can see how each selection impacts both key metrics: shot accuracy and points per shot.
The menus are independent from the others, which allows for targeted drilldown selections. Users can hold down CTRL (“Command” if using a Mac) and select multiple options within the same menu, allowing them to find answers to any unpredictable ad hoc questions that may arise.
The top bar provides users with key metrics, additional filters, and a help icon. In total, 10 dashboard elements are packed into these two small areas of the dashboard. The “Shot Type” menu is an efficient use of space, because it acts as both a filter and as the color legend. Organizing these elements into their relative places on the dashboard was extremely important, because it’s easy to create a confusing mess when adding so much interactivity.
Since the charts and navigational techniques used in this dashboard are so complex, the design had to be disciplined. In contrast, an aimless design approach would either overwhelm or confuse users about how to interact with the data. Each individual design decision had to support the overall design goal of keeping everything as clean and simple as possible.
During the process of building this dashboard, I was going through an overly enthusiastic Adobe Illustrator phase. Although I was never a horrible offender of chart junk, I relaxed some of my opinions concerning this topic at the time. I could not, however, let myself relax when making a specific design decision that could have potentially ruined the overall purpose of the visualization.
I created a custom hardwood basketball court image in Adobe Illustrator, and it turned out perfect. Unfortunately, the overall dashboard design broke when trying to include this image as a custom background. It added too many unnecessary colors and textures, which distracted attention away from the data.
Despite the several hours I spent making this neat graphic, it needlessly reduced essential white space without adding any true benefit — so I scrapped it. Since my overall design goal was to keep the presentation as clean and simple as possible, I was forced to make the correct decision when it mattered most. This situation reminds me that even a seemingly small design choice can make or break the success of a dashboard.
Overall, only three colors were used: blue, gold, and white/grey. This color selection helps viewers associate this dashboard with Golden State, while not detracting from data viz’s best practices.
Luckily, the team colors for the Golden State Warriors are naturally easy on the eyes. The header is baby blue, the navigation bars use two different tones of light grey, and the canvas is white. These different color layers reinforce the separation of sections: header, navigation, and chart area.
As I’ve said, the ultimate design goal was to avoid any elements that could distract users from the main features of the dashboard while also adding fun interactivity. The principal design focus should be “do no harm” when supporting complex chart types and data visualizations. The decisions I made throughout the design process for my Steph Curry dashboard balanced interactive analysis, user experience, and design.