Visualizing Wikipedia: First 2015 Iron Viz Qualifier Submission

Tableau announced this year’s first contest for the Iron Viz competition. For this visualization, I was only allowed to use data from Wikipedia.
Check out my submission here.

The Idea
In the midst of March Madness, how could I not build a viz around the NCAA Basketball Tournament? I knew I wanted to build a bracket, but I wanted it to tell a different story than just basketball.

The Data
Collecting the data was tedious. For every school in the 2015 NCAA Men’s Basketball Tournament, I collected the school’s student population size, endowment fund and campus size (in acres) from the school’s Wikipedia page. With the contest only being limited to Wikipedia, some of the metrics were unable to be found. Unfortunately, some metrics were measured by different standards on each Wikipedia page. Take campus size for example. Some schools have multiple campuses and only recorded the acreage for all of the campuses. Some schools broke up the acreage to show main campus and total. Unfortunately, one school did not record any acreage. Since the data was not recorded in the same way for each campus, I had to make a judgement call. I decided to take the largest number I could find on the Wikipedia page. I was happy to see the University of Georgia win something, but I have to admit that the Athens campus isn’t quite that large. I was surprised to see how many acres in Georgia they do own though. So I may be biased, but I believe that deserves a prize.
I also went to every school’s basketball Wikipedia page to get the school’s mascot or logo and collect the school’s colors.

The Build
Every time I build a new viz, I learn a little bit more. When I set out to build this viz, I had two goals: Master the Path Feature and utilize the cool new features in Tableau 9. With the visualizations I create for my job, there’s little need for the path feature in Tableau. I wanted to learn more about it, so I made sure to incorporate it in my dash. Then, it turned out to be the main feature of my viz. I’ve come to really love the feature and can’t believe I’ve barely used them before now. From other blog posts and a lot of trial and error, I restructured my data to visualize the bracket on my viz. My vision for the viz was pretty simple: who would win the bracket if this wasn’t a basketball tournament? Instead, I wanted to determine the winner by other metrics that schools take pride in, such as student population size, endowment fund and the size of the campus. Determining ONLY the winner would be pretty simple, just return the max of the metric chosen. My simple vision turned a little more complex…I also want to determine how far in the bracket each team advances.

For each game, I needed to determine who the “winner” would be. The new Level of Detail (LoD) Calculation in Tableau 9 was perfect for this. I “fixed” my calculation on the Division the team is in and who that team will play.

The first LoD calculation
{fixed [Division], [seed groups]: max([DetermineWinner])}
Seed group is a calculation that determines the other opponent for each team. For this, I used a case/when calculation. Seed 16 and seed 1 gets the same value, seed 15 and seed 2 get the same value, and so on. The DetermineWinner calc uses the parameter to return the different metrics to determine who the winner will be. The LoD calc returns the winner’s metric. I continue this type of calculation until we get to the championship game. I roll all of the teams that are playing together and return who the winner is for each round.

Next, I needed to determine the path. Since the bracket is built on Cartesian coordinate plane, I was able to use the x-axis to determine where each round ended.
The next calculation:

//Champ Game
if abs([X])>0 and [test 5]=[DetermineWinner] then 5
//Final Four
elseif abs([X]) >= 2 and
([test 5]=[DetermineWinner] or [test 4] = [DetermineWinner]) then 4
//Elite Eight
elseif abs([X]) >= 4 and
//Sweet Sixteen
([test 4]=[DetermineWinner] or [test 3] = [DetermineWinner]) then 3
//Round Three
elseif abs([X]) >= 6 and
([test 3]=[DetermineWinner] or [test 2] = [DetermineWinner]) then 2
//Round Two
elseif abs([X])>=8 and
([test]= [DetermineWinner] or [test 2] = [DetermineWinner]) then 1
//Start of Tournament
elseif abs([X])>=10 and [test]!=[DetermineWinner] then 0


The Design
Now, I needed to clean up the viz. I wanted to make sure I kept a sleek design with dynamic functionality. When I put the teams name as a label, the viz started to look messy. I used the dual axis feature to add the team’s logo as shapes. I also used the Actions feature so that the bracket and the logo functioned in unison. When a team’s logo is selected, you are able to see how far the team advanced in the bracket. Here is an example:


When I clicked on Oklahoma’s logo, you can see how far Oklahoma advances in the bracket. Because the path stops, we know that Oklahoma lost in the “Sweet Sixteen game”. To make the viz, visually appealing, I added reference bands for each round. The darkest grey is Round 2, next is Round 3, then the Sweet Sixteen, followed by the Elite Eight, the lightest grey is the Final Four and the white space is the championship game. The logo in the middle is the winner of the tournament. I also added the Wikipedia logo to give the directions of the viz. When you hover over the logo, it provides directions to interact with the viz.

The Edit
I initially published my viz with an all black bracket. After I continued to play with the viz and others gave feedback, I realized that it wasn’t exactly intuitive that the path actually shows how far the school advances in the bracket. I initially wanted to “fill out” the bracket by showing the logos for each round. This ended up looking really busy and taking away from the simplicity of the viz. After playing around, I found out that the school colors are a great way to show how far the school advances.

I also wanted to make sure I showed the exact links from Wikipedia where the data was gathered. I used the URL action feature to link to the school and basketball website. Let me know what you think of the new features.

Please check out the viz and give me any thoughts you may have. I look forward to hearing from the users. Many thanks to my friends for troubleshooting and helping me with the idea. 

Tagged , , ,

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: