Taking some time to participate in #SportsVizSunday
Typically for me, this time of year is a lot of fun from a sports point of view. Formula 1 started the season this past weekend, March Madness tips off tonight, we're getting close to playoff hockey, and baseball is almost back!
Growing up in St. Louis, I've always been a baseball fan. When I saw that the #SportsVizSunday guys were having a baseball themed month, I figured I should try to participate. I'd been bouncing around an idea for a MLB viz, focusing on the difference in performance during Spring Training and how the season typically turns out.
As fans, we tend to put a lot of weight on evey game our teams play, even Spring Training games. I think it's human nature to root for our teams even when the outcome isn't all too important. However, I wanted to definitively prove that the outcome of Spring Training isn't well correlated with the outcome of the regular & postseason.
A quick note on the viz. The methodology for creating the sigmoid curves was actually quite simple. With a bit of data scaffolding, and the well documented sigmoid function, I had the curves drawn in about 5 minutes. In the interest of not replicating internet data, I won't go into the process - Google 'Sigmoid Curve Tableau' for one of several tutorials
The data prep on the other hand was a bit more time consuming. I scraped the standings from ESPN.com, got rid of some of the columns that I didn't want, and came up with my methodology for establishing the standings for both Spring Training and Regular/Postseason. Spring Training was the easiest - I ignored the Grapefruit and Cactus league boundaries and settled on win percentage to determine the standings. For the Regular and Postseason, I ranked the World Series Champions as 1st in the standings and the runner up as 2nd. From that point, I ranked the remaining 28 teams by win percentage.