Royal Statistical Society Professional Statisticians’ Forum Webinar
A free-to-attend, virtual meeting for professional statisticians
Monday 21 November 2016, 2pm-3pm (GMT)
Working as a Data Scientist in sport: are empirical models good enough and how do we measure good enough?
Presented by Dave Hastie (Sporting Data Science)
In this webinar Dave will discuss his consultancy work within the sports industry as a data scientist at Sporting Data Science. Having started in this area in 2003, Dave has worked with professional sports teams and confederations, broadcasters, companies building fan engagement products and organisations linked to sports betting. To date his his work has focussed on football, tennis and cricket but the approach can apply to many other sports.
The talk will provide a flavour of Dave’s work, giving examples covering the whole data science pipeline, including data wrangling, data analysis, visualisation and statistical modelling. The webinar will cover some of the challenges working within the industry. Specifically, Dave will focus on communicating complex ideas and receiving buy-in amongst traditional sports executives and fans, who rarely have a statistical background but are considered to be experts in their domain.
The final part of the talk will look at the specific question of how empirical based models developed by domain experts compare to the parametric or non-parametric models that are the basis of the statistical modeller’s toolbox. As part of this question, Dave will also address the issue of model assessment and some of the difficulties in comparing models within this domain.
The presentation slides will be available from www.rss.org.uk/psf prior to the meeting.
Dial-in details
Phone number: +44 (0) 330 336 6002
If you are outside of the UK please visit www.rss.org.uk/psf then follow the link for a local rate number you can use.
Passcode: 476 766 1144
Web address: https://rss-uk.globalmeet.com/RSS
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|