DSAU25, my cohort within the Data School, just finished our first presentation on the Friday of our first week in the office. Tasked with making alterations to the visualization we created for our initial Data School application, we put our heads down as early as 9am in the morning and got stuck in as the rain outside billowed and wailed in an endless downpour.

My attitude to the task at hand, while initially full of optimism and excitement at the thought of redesigning my basketball-based dashboard, quickly began to resemble something more akin to pathetic fallacy – a nightmarish, hailstorm of problem after problem pelting me as I struggled to make Tableau bend to my will.


Above is the link to my initial visualization, and while I was happy with it at the time, there were a number of changes needed that I felt helped tidy up the format and make it more presentable, as well as allowing for easier exploration.

The first thing I changed was the size of the text and pictures, as it was formatted for Tableau public and thus wasn’t correctly spaced on Tableau desktop. I also changed the font of the title of the graphs to ‘Helvetica Neue’, which is the same font as the rest of the text on the dashboard. Additionally, I removed the Average Shot Distance table, as it didn’t provide much important information and was too fiddly to alter, with total numbers changing if anything was touched.

The 3 different line graphs in my first section had various differences (one had bars, while one was an area), but they all tracked variables on a dual axis, so I made the decision to create parameters and a calculated field which would allow each y-axis to change depending on which variable is selected.

The data itself was split into 5 different excel sheets, which meant that the queries took forever too load. I also didn’t know too much about relationships and joins, which meant the tables weren’t connected in the most efficient way, limiting the way some of the data could be presented. This also meant that to have the two main NBA seasons (the 2003-2004 and 2022-2023 seasons) in my dashboard presented on the same graph, I had to create different parameters and filters that allowed a graph to switch between the data of the two seasons. As shown in the image below, I created a calculated field that allowed the parameter to switch what is shown on the graph between the two seasons when selected – I also altered the colours of the ‘Shots Taken and Made’ graph to coincide with the ‘Zone Range’ legend.

The last section with the scatter plots of the different NBA teams was largely left the same, however I did change the colour of the dots to a neutral gray so as not to confuse with the colour legend of the previous graphs, as well as moving the ‘Wins’ slider to below the plot.

The only change I made to the corresponding information below was to adjust the colours of the ‘Make/Missed Percentage by Zone’ graph so they matched with the ‘Zone Range’ legend.

I also removed the shot chart I created that visually displays the location of each individual shot a team took over the course of the season – while I really enjoyed creating it, what it showed was pretty much the same thing the rest of the graphs on the dashboard showed.

There were a few more changes I wanted to make, but I ran out of time as the hour to present crept closer and closer. I accidentally left the x-axis off the scatter plot, and I didn’t have time to rearrange where I would put the ‘Zone Range’ colour legend for the graphs below the scatter plot.

Beth had a few pointers after the presentation as well – she recommended making the ‘Total Shots’ table and the ‘Most Popular Shots by Position’ graph a bar chart to help visualize and accentuate the difference in shots. She also recommended altering the colours of the ‘Zone Range’ to different intensities of one colour, so it’s understood they are different ranges of the same thing. The final things she mentioned was to change the ‘Team’s Shot Zone Frequency’ pie chart to a bar graph, and to make more of a mention that you can click the scatter plot points to explore the stats of each team and that the green and red highlights for ‘Team’s Offensive Numbers” indicate above and below league average respectively.

This was a great chance to go over what we created when we knew nothing about Tableau and add some new features that we learnt this week.


Nicholas Seah
Author: Nicholas Seah