Basketball Data Madness 2020: Data Challenge Kickoff Event

By |

BASKETBALL DATA MADNESS CHALLENGE 2020

The Michigan Institute for Data Science (MIDAS) and the Exercise & Sport Science Initiative (ESSI) are excited to announce the 2020 Basketball Data Madness Challenge for students with an interest in data science. Student teams of up to 4 members will be given access to two datasets on Michigan Basketball player performance loads and overall game statistics.  Teams will have one week to analyze the datasets (on your personal computers) and answer predefined questions from the challenge organizers. Final questions will be similar to the following examples:

1) How do performance loads relate to game outcomes (e.g., win/loss)?

2) How do performance loads relate to game dynamics (e.g., tempo)?

In the spirit of March Madness, 4 winning teams will be selected by a panel of U-M data scientists and will each receive a $400 “Final Four” prize.  They will also get to present their results at the Sports Analytics Conference, hosted by the Ross School of Business on April 15, to leaders in sports analytics from the NBA and Michigan Basketball.  The judging will be based on both the scientific content and the communication of results. We look for well-designed and well-communicated projects at any level of expertise.

Both graduate and undergraduate students, at all levels of data science expertise, are welcome to participate.  One team member should register the entire team by Friday, March 13.  You may also register as an individual and MIDAS will help you form teams.  Only 16 teams will be accepted to compete.

Kickoff Event will be held in Ross #420

BlueJeans:To join the meeting on a computer or mobile phone: https://bluejeans.com/604498818

One Touch Dial-in:
+13122160325   604498818#

Connecting directly from a room system?

1.) Dial:
– 3-1841 (Registered with U-M Video Cluster)
– 199.48.152.152 or bjn.vc
2.) Enter the Meeting ID: 604498818 or use the pairing code

Just want to dial in?

1.) Dial:
+1 734 763 1841 (Last 5 digits from campus)
1.408.614.7898 (US or Canada only)
+1.312.216.0325
International Callers (http://bluejeans.com/numbers)
2.) Enter the Meeting ID: 604498818

Want to test your video connection?
https://bluejeans.com/111

Challenge Timeline:

March 13 – Team Registration Closes

March 16 – Data Challenge Kickoff Event – Learn about the challenge datasets.

March 23 – Results submissions due via PDF or slides and a Jupyter Notebook

March 25 – Announcement of Final Four teams

April 8 – Final Four project summaries due

April 15 – Sports Analytics Conference

8:00-8:30am: Final Four team meetings with industry representatives

12:40-1:00pm: Final Four presentation for conference attendees

Questions about the data challenge?  E-mail MIDAS Education Program Manager, Trisha Fountain (tvfount@umich.edu).

U-M students make strong showing at Michigan Datathon

By | Data, Educational, Events, General Interest, Happenings, News

University of Michigan students won first and third places in the Michigan Datathon held Nov. 4, 2017 in the Michigan Union and hosted by Citadel LLC, Correlation One, and the U-M Statistics Department.

1st-place winning team from the University of Michigan:

Ruofei (Brad) Zhao, Statistics Ph.D. student

Zheng Gao, Statistics Ph.D. student

You Wu, Master’s in Applied Statistics student

Kevin Zheng, Sophomore, Computer Science

 

2nd-place team:

Zi Yi, Statistics Master’s student, University of Chicago

Tian Gu, Biostatistics Ph.D. student, University of Michigan

Shuo Zhang, Statistics Master’s student, University of Chicago

Shiyang Lu, Robotics & Naval Architecture and Marine Engineering Master’s student, University of Michigan

 

3rd-place team from the University of Michigan:

Hanbo Sun, Master’s in Applied Statistics student

Xinghui Song, Master’s in Applied Statistics student

Tuo Wang, Master’s in Applied Statistics student

Hang Yuan, Master’s in Applied Statistics student

 

For more, see https://lsa.umich.edu/stats/news-events/all-news/graduatenews/MichiganDatathonWinners0.html