NCAA Rankings: Road to the Final Four




NCAA Rankings

I live in Kentucky…the Northern part, just south of Ohio.  Even though I’m 15 miles from the University of Cincinnati and 16 miles from Xavier University, we learned at an early age to root for the Kentucky Wildcats (70 miles away).  I love Kentucky basketball (who, as of me writing this, are ranked number 4 in the nation and just demolished the #1 Tennessee Volunteers) and watch most of their games (but I’m not one of THOSE fans).  I enjoy the sport in general, I love watching other teams play, and boy do I love tournament time. 

If you love sports and love data, you can have an absolute hay day.  Sports data is so rich and so accessible.  It’s probably why Simon Beaumont, Spencer Bauke, and James Smith’s SportsVizSundayis so popular.  Recently, I’ve spent most of my Tableau time with sports data; just look at several of my most recent vizzes: A Giant of a Season, Streaking, and now a viz on NCAA Basketball.

Okay, let’s chat briefly about the NCAA Rankings viz.  I’ve been using Tableau for about 12 months and to be perfectly honest, I’ve never created a bump chart.  I decided that I would like to do just that and would need to find some data to utilize.  Then one evening I was watching a UK game, saw the Top 25  ranking next to their name, and thought that NCAA Rankings over time would make a perfect bump chart.

I created the bump chart pretty easily, after all it’s really just a line chart with a dual axis to show a shape.  However, when I looked at the results, I thought it was a bit confusing.  There were 25 times for 12 weeks (at that time) which meant it looked like a rainbow had thrown up.  I decided to change it up a bit by using logos, team colors, and to allow users to choose two teams to compare (while fading other teams to the background).   The downloading and cleaning up of all the logos and matching of team colors took forever...


I couldn't resist a little baseball movie humor.

When I was about 90% complete, I saw Jeff Plattner tweet out a very similar NCAA Rankings Viz.  It was well done, was on a black background like mine, and certainly looked similar.  His was a very good viz and I admit, I was a bit bummed that he beat me to the punch.  However, they certainly had their differences and I wasn't going to let it discourage me.

As I continued to finish up the last 10% of this viz, I decided to bail on the straight lines and implement some nice, smooth curves...Sigmoid curves to be exact.  Jeff Shaffer was the first person to utilize Sigmoid curves in Tableau.  He used it to recreate a curvy slope chart, which he wrote about in two blog posts:  A Graph Recreation in Tableau and the follow-up regarding How to Build it.  Since then, Sigmoid curves have become a staple for Tableau developers.  I used Jeff's techniques along with a Sigmoid Bump Chartblog post from Rody Zachovich.  I also downloaded Matt ChambersPath to the Playoff viz.  I then created the Sigmoid curves to connect the dots…and boy did it make a world of difference.

I’ll soon be writing more about the Sigmoid Bump Chart, but for now, I present NCAA Basketball: Road to the Final Four.  (And yes, I will be updating this visualization through the end of the season). 



As always, if you ever have any questions or comments, please feel free to contact me at any time.


Kevin Flerlage, February 26, 2019 | Twitter | LinkedIn | Tableau Public








2 comments:

  1. Hi,

    How do you add team logos to the ranking chart as mark? And how can you create a filter showing the logo/best rank/worst rank on the side?

    Thank you so much!

    ReplyDelete
    Replies
    1. Hi Pat. The logos are just custom shapes. Learn more about them here: https://www.tableau.com/drive/custom-shapes. The filters are not filters, they are parameters. You can read more about those here: https://help.tableau.com/current/pro/desktop/en-us/parameters_create.htm#:~:text=A%20parameter%20is%20a%20global,%24500%2C000%20and%20otherwise%20returns%20False.

      Delete

Powered by Blogger.