Data Science eSport with Live Commentary

Watch and learn, as two competitors dive into a data science problem with live commentary and explanation! See how the players frame the problem, deal with tricky data, and push to get a result, all within a one hour playtime limit. (2 hours of total time)

Events are 2 hours long:
0m - 30m: Pre-game introductions

• Players first see the problem (they have 30 minutes to read and plan, but cannot start working yet).
• Commentators introduce themselves
• Commentators introduce each of the players and their data science toolkit
• Commentators discuss the problem, the data, and likely approaches

30m - 35m: Pre-game check in

• Check in with each competitor to hear about their strategy for tackling the problem

35m - 55m: Framing the problem and exploring data

• Players start with data exploration and writing code.

55m - 60m: Break 1

• Commentators debrief each player on what they’ve learned during exploration.

60m - 80m: Feature engineering and pipelining

• Players are suggested to spend this time cleaning data, engineering features, and creating machine learning pipelines.

80m - 85m: Break 2

• Commentators debrief each player on where they are, and what they hope to accomplish in their final 20 minutes.

85m - 105m: Machine learning and algorithm tuning

• Players are suggested to suggested to spend this time training and tuning machine learning algorithms

105m - 120m: Final debrief

• Results are revealed for each player
• Audience votes for their favorite competitor
• Commentators and players discuss how they feel they did and what they would have done if they had more time
• A winner is selected

Player coding screens will be visible (commentators will flip between the two). Commentators will discuss their take on the data science skills being applied, as well as occasionally going into detail about beginner/intermediate topics.

(Do you want to get involved as a competitor, commentator, or behind the scenes? Reach out to the organizer, Aaron).