Since joining the Tableau Public team three weeks ago as its first Sports Data Analyst, I have had the enviable job of coming up with interesting visualizations about football data, to herald the beginnings of the NFL and NCAAF seasons. As I am still myself a novice Tableau user, I thought that in this post I would walk through the steps I took in undertaking this project, hoping to provide some practical fodder for those of us interested in using Tableau to analyze the sports that we love. I will segment my post into ten steps, constituting the first of what I hope will become many Dash Board Top 10 blog posts.
- Identify area of inquiry: Every project must have some burning questions that drive its direction. As I contemplated the beginning of the college and pro football seasons and my all-important fantasy football matchups, I realized that I could explore the intersection between college football and fantasy football in the NFL - which college teams and conferences produce the best fantasy football players - a relationship about which I knew little. This explorative desire motivated me throughout the project, culminating in my finished dashboard: Best Fantasy Factories
- Identify what types of data you will need: Once you have figured out a broad area of focus, you need to then start thinking about the nuts and bolts of your research project – namely what types of data will you need to do your analysis. For me, this step was pretty clear: I needed NFL fantasy data and college football data that were linked together to allow cross-comparison.
- Find necessary data sources: This step in the exploratory process is often the most arduous and time-consuming, as rarely can you find all the data you need in one place or in the appropriate format. Often you will have to scrape data from different websites and then jigger into a suitable format in Excel or a similar data aggregation program. For my project, I needed a couple of different data streams: I needed NFL player data with yearly fantasy point production. I needed biographical data on those same NFL players, such as where they went to college, where they were born, how tall they are, etc.
- Download datasets and get them into a ‘Tableau-able’ form: Here is where your Excel skills come into play, as often it takes a lot of grunt work to get your data into a usable Tableau-friendly form. For me, the process was as follows:
- To acquire the NFL data I went to pro-footballreference.com and found their yearly fantasy point leaderboards. Then all I had to do was download all the relevant years and compile them together in an excel spreadsheet.
- Acquiring the biographical data was slightly more complicated, as I had to scrape it from the player cards on profootball-reference.com by using an external scraping application called Import.io that I found to be efficient and effective. It is highly recommended for all your scraping/extracting/crawling needs!
- Remove all extraneous Excel columns/rows (no blank rows/columns!) and connect your dataset to Tableau.
Thanks for following along as I worked my way through my first ever Dash Board Top 10. If you are interested in NFL vizes like me, check out some of these recent Tableau Public author creations visualizing different elements of the NFL:
- Matt Chambers Bigger, Stronger, Faster: Visualizing the NFL Combine
- David Seawright’s Why Campaign's Should Advertise During NFL Games
- Bill Petti's Fantasy Football: Projections and Consistency 2014
- Matthew Cobb’s College Football…Big Business