Mark Cuban, Big Data and Rugby. Intriguing.
I wrote recently about Opta and AIG doing something interesting with stats. Well, the final announcement has been made today.
It’s mostly along the lines of the original snippet – an AIG-sponsored competition to develop the best rugby apps using Opta’s data from all of New Zealand’s internationals in 2012, with the winners getting a chance to get their hands on a prize fund totalling $85,000 including access to Opta’s stats feeds for two years – but with some interesting additions.
The judging panel has been given some real heft with the inclusion of IRB CEO Brett Gosper as well as tech entrepreneur Mark Cuban. Cuban, now owner of the Dallas Mavericks NBA team, is a fan of using analytics in running his franchise and talked about that at length at the Sloan Sports Analytics Conference in Boston earlier this year.
Cuban might now be associated most readily with basketball but he played rugby at Indiana University (N.B. link slightly NSFW) as well as afterwards for the Pittsburgh Harlequins, Dallas Harlequins and the Dallas Reds. With that in mind he’s probably one of the more prime candidates on the planet to judge a tech/rugby/analytics competition.
When asked today about his motivation for getting involved, Cuban said “I’m a huge rugby fan plus I’m a huge fan of sports analytics. It was a great opportunity I couldn’t pass up”.
Developers entering this competition will get access to data feeds from 2012 including on-the-ball actions categorised and logged with an x/y coordinate attributed to where it took place on the pitch. That sort of data in terms of both quality and scale doesn’t usually get into the public domain and, even though it’s restricted to competition use only, it’s got potential to open things up a bit for sports analytics enthusiasts.
Getting sports data into the public domain requires investment, someone to foot the bill on behalf of the stats provider. In this case that’s insurance giant AIG who according to their head of sponsorship Daniel Glantz want to “present the sport of rugby in new and interactive ways”. But getting this data into the public domain (in whatever form) can provide real value within sports itself as new ideas are produced by those outside the very small, isolated pool of analysts employed by professional teams.
In Major League Baseball there is a system called PitchF/X that tracks the speed and trajectory of a pitched baseball. Mike Fast was an amateur baseball analyst who latched on to PitchF/X as being of huge potential value. His analysis of PitchF/X data led him to write for the widely respected analytics website Baseball Prospectus and ultimately to Fast ditching his previous career as a semiconductor engineer and being hired to work full-time for MLB’s Houston Astros in 2012.
Back to rugby, the “Overall Grand Prize” for this competition is described thusly (and attributed a value of $57,640)…
- $7,500 cash prize
- Tickets, round-trip flights (flying Air New Zealand), and accommodations for 2 to an All Blacks rugby match in New Zealand (Total retail value: $13,568)
- Live data Rugby feed from Opta Sports for 2 years (Total retail value: $30,000)
- Promotional exposure for your app for one week on Rugbydump.com and Allblacks.com (Total retail value: $3,750)
- One-hour consulting meeting with Opta Sports (Total retail value: $1,000)
- New Zealand All Blacks and USA Rugby swag (Total retail value: $1,822)
Apart from that there are a few other prizes also available including a “Rugby Education” prize for the best app for educating people new to rugby about how the sport works. Considering the link to the USA in this competition – access to some of the USA’s 2012 rugby data feeds as well as the presence of US Rugby head Nigel Melville on the judging panel – that’s quite an interesting one.
Bottom line? The winning apps will need to have the capability to get more people interested in rugby (or to get rugby fans even more interested in rugby), whether that’s checking stats on their favourite players, checking match stats before/during/after games or in some other neat way that the clever developers will think up. The details are now up at http://aigrugby.challengepost.com/ and those techie-boffins interested in entering will have until October 9th to submit their entries with the judges announcing their final decisions in mid-November.
While this app competition might generate plenty of publicity for AIG’s rugby sponsorship, it might also be a way to find rugby’s Mike Fast. One thing is for sure – progressive rugby scouting departments will be looking closely at what this competition throws up.
I downloaded that data this evening. There’s certainly a comprehensive amount of it and learning how to sift through it quickly would be the first step.
Disappointingly there doesn’t appear to be detailed player information on rucks – as in who hits them etc. This would seem to be a major hole in their data.
Hi Mike,
do you still have a copy of the data? I’d love to get a copy of it. I’m studying for a masters at present and there’s no x/y data available publicly for rugby.
It’s much later but I’m looking for a rugby dataset to dig into. Anybody have it? I’m a working data scientist in the health care field but I love rugby and am looking for something interesting to do.
Hi Mike. That’s interesting, thanks. Know what you mean – ideally you’d love to get data that, in the right hands, could recreate movements of ball and players for an entire 80 minutes. Suppose player info in rucks doesn’t amount to “on the ball actions”?