D3boards.com

D3soccer.com => Men's soccer => Topic started by: ctdennis11 on July 31, 2018, 04:20:55 PM

Title: College Soccer Analytics
Post by: ctdennis11 on July 31, 2018, 04:20:55 PM
Hey there... I'm new to the group but thought others might benefit from some work I did with respect to college soccer analytics. At this point it is mostly New England / New York / D3 focused but am hoping to find time to include additional states & conferences.
https://public.tableau.com/profile/charles.dennis#!/vizhome/D3SoccerAnalysis/CollegeSoccerRosterData (https://public.tableau.com/profile/charles.dennis#!/vizhome/D3SoccerAnalysis/CollegeSoccerRosterData)

Essentially I've downloaded and normalized roster data and rankings from 2017. The tableau visualization allows users to filter on conference, position, height, major, roster size, home town, and a few other dimensions. Some of the data is sparse, meaning not much data on major for example... or some schools don't include height/weight... etc. Clicking on anything will help filter the data so, for instance you could filter on teams with roster sizes between 25-35 players who were graduating a high number of juniors and seniors or determine how many midfielders in NESCAC are between 5'8" and 5'10". Anyway, it is posted to Tableau Public and available to all. Hope you can find some interesting data. Best if viewed from a larger monitor vs a cell or tablet.

Also... stuff i'm working on like determining if average taller or heavier have are statistically more likely to finish higher in the conference. Same for teams with larger rosters vs those with smaller. More to come.
Title: Re: College Soccer Analytics
Post by: blooter442 on July 31, 2018, 05:00:35 PM
Great stuff. Given the (informally labeled) "arms race" that some have observed in recent years in terms of NESCAC teams seemingly getting "taller" on average, I wasn't all that surprised that it was the "tallest" conference on average out of those listed. However, it was somewhat surprising (at least to me) that its average height is greater than that of D1 conferences ACC and America East. And while I figured that Bates was the "shortest" team in the conference, I was also surprised that Amherst and Tufts (and four other schools) have the same average height — it would appear to me that, in addition to Amherst, Tufts and the other four have had an eye on height when recruiting. Obviously correlation does not equal causation, and squad average height does not determine the success of a given team, but interesting data points nonetheless.

Update: Wow — after looking at it in more detail, Tufts is "taller" than Amherst — 71.846" vs. 71.828" (2nd and 3rd, respectively). Colby first at 72.107". Again, there are many caveats, but I guess Amherst isn't the stereotypical "size" NESCAC team anymore!
Title: Re: College Soccer Analytics
Post by: ctdennis11 on July 31, 2018, 05:08:43 PM
If I could find time to add a few more years data, it would be interesting to see the trends. Glad you're finding some value in it!
Title: Re: College Soccer Analytics
Post by: NEPAFAN on August 01, 2018, 12:15:02 PM
following , cool.
Title: Re: College Soccer Analytics
Post by: ctdennis11 on August 01, 2018, 03:49:57 PM
Added town data to the map, zoom in by scrolling. Select multiple towns or other data points by holding down shift or control on your keyboard.
Title: Re: College Soccer Analytics
Post by: ncac_dad on August 30, 2018, 07:34:17 PM
Great job!

I use Tableau at work. I used it during the recruiting process for my sons.

My boys loved it.

Great work in using the application as it was intended.