CMSAC Goes to SSAC19: A Recap

Written by: Carlo Duffy, Shravan Ramamurthy

The 2019 MIT Sloan Sports Analytics Conference was an action-packed two-days’ worth of panels, talks and networking.  We won’t bother documenting our entire days, but here are some highlights from the sessions that most stood out to us.  Shravan will highlight Day 1; Carlo will highlight Day 2.

Day 1: Shravan Ramamurthy

Shravan with Sports Analytics and Gambling Guru Jeff Ma

Basketball Analytics: Hunting for Unicorns

In this panel, several prominent members of Basketball sat down to talk about how basketball analytics is shifting towards finding "unicorns". These esteemed members included Paul Pierce, Mike Zarren, Bob Myers, and Zach Lowe. They defined a unicorn as a “big man that can stretch the space” and both play in the post and on the perimeter. Several examples that were brought up include Kristaps Porzingis and Karl Anthony-Towns.

An interesting part of this panel were the analytics that were used to evaluate players. For instance, everyone thought that the draft of Porzingis in 2015 was one of the worst picks ever. However, Kristaps became a bona fide superstar, and had it not been for his ACL, he would have been an All-NBA player this year.

The panelists were also asked about unicorns that aren’t “big men”, but rather, players that can perform in the clutch. In this instance, Paul Pierce started talking about how he would encounter players in the regular season that he knew weren’t playoff ready. Come April, Pierce would prove everyone right, when he would propel the Celtics to a deep playoff run nearly every year. Another example that was brought up was the prowess of Klay Thompson, especially in the playoffs. Such instances include the Western Conference Finals Game 6 in 2016 and the Western Conference Finals Game 6 in 2018.

Coming out of this panel, the reasonable next steps in analytics were for players, regardless of size, that can perform in the playoffs. This is interesting as the game is shifting towards positionless players that operate cohesively rather than players that play set positions.

Elam Ending: Revisionist

Nick Elam gave a talk about the Elam Ending, a revolutionary idea about how to end basketball games. The idea is that once the game has less than 4 minutes left, the game shifts from a time based game to a score based game. For instance if Team 1 is playing Team 2 and the score is 78-69 Team 1 winning, once the Elam Ending starts, the target score becomes the leading score plus 7. This is to ensure that the game doesn’t devolve into intentional fouling to preserve the clock.

The common trend now is that when teams are losing by a reasonable margin, they resort to intentionally fouling the opposing team. Although in theory this might work, we see that in execution, this seems to fail all the time, as well as create a more disengaging experience for the fans. For instance, in a high leverage game like the playoffs, it is very boring to have to see the last two minutes devolve into stop and go play centered around free throws, when in reality, the game is much more fluid.

Another benefit to the Elam Ending is that every game ends on a buzzer beater. Every basketball fan knows that a buzzer beater is the best way to end a game. The Elam Ending makes it so that the best of basketball meets the best of pickup. Recently implemented at TBT (The Basketball Tournament), the results have been very encouraging. They see that the game is more volatile when the Elam Ending was used (the losing team came back to win), and that overall, there was a better fan experience.

Day 2: Carlo Duffy

Carlo with FiveThirtyEight's Editor-in-Chief Nate Silver

Career Kickoff Panel

Leading off the day was an informative panel on breaking into the sports industry, featuring Matt Wolf of the NBA’s Team Marketing and Business Operations Department, Michelle Wang of Ticketmaster, Erin Golden of Nike, Brandon Shore of the Miami Dolphins, and Hana Cluff of WarnerMedia.  The speakers discussed how they had done so and how others can. Wang credited being at her current job at Ticketmaster to (1) personal recommendations and (2) connecting well with her eventual boss at Ticketmaster. Both can be attributed to letting personal connections happen organically, which she jokingly likened ­­to dating in that connections “just happen.”  Golden also discussed that knowing someone in Nike, before even working there, helped to put her foot in the door. Later in the panel, Wolf and Shore shared advice on getting your resume noticed in the industry. Wolf offered something unconventional about personal recommendations: he explained why picking someone with different skills than you can provide a different perspective about your abilities, and thus make said abilities stand out.  Meanwhile, Shore recommended that your resume should tell a “story” about you; bullet points should connect to form a more vivid picture. Finally, Cluff suggested chasing not the brand but the job, since after a while the novelty of working at a big-name brand will fade. It is a job, after all.

Social Media in MLB: A Data-Driven Approach to Optimizing Strategy

Milwaukee Brewers VP of Strategy and Analytics Mike Schwartz presented data-driven research on how to get as many people as possible to engage with the Brewers’ social media accounts.  Conducting this analysis required the mundane task of scraping posts on Facebook, Twitter and Instagram, and determining how to record agreed-upon operational definitions (e.g. if we agree to include a variable on how “funny” a post is, how do we determine how funny it is?).  Schwartz and his team had to be careful as some posts were hard to quantify.  Success, measured in terms of in-platform total interaction (likes, retweets, etc.), was modeled as a function of many independent measures that were difficult to observe.  They found that of the three platforms (Facebook, Twitter, Instagram), the Brewers had the least impact over controllable engagement on Twitter, yet the most on Facebook. They can use this knowledge to drive social media strategy going forward.

Making the Modern Athlete: A Conversation with David Epstein and Malcolm Gladwell

This was an engaging conversation between two highly respected authors.  Here the goal was to discuss whether early specialists or generalists have more athletic success.  To motivate this debate, Epstein and Gladwell opened by contrasting Tiger Woods’ background to Roger Federer’s.  Woods’ road to fame began very early: he first showcased his golf skills at five years old.  People who point to Woods’ success argue that mastering a sport requires hyper-specialization, entailing intense, repetitive training.  This has been the classic argument to achieve superior athletic ability in a particular sport. But pointing to Federer’s background, Epstein presented an alternative, perhaps better, path.  In contrast to Woods’ focus on one sport, at an early age Federer trained in other sports besides tennis, such as basketball and badminton. Thus others, like Epstein, have argued it might be better for athletes first to try many sports in order to broaden overall skill development.  This discussion then extended to optimal youth programs, and even contexts outside of sports such as medical school and training in academics and music.

Sports Mythbusting

Towards the end of the day was a fun, rapid-fire panel hosted by Freakonomics’ Stephen Dubner and The Ringer’s Jason Concepcion.  It turned out to be a live Freakonomics radio podcast. The hosts quizzed U.S. Men’s National Soccer Team’s Darcy Norman, WNBA point guard Sue Bird, DraftKings’ Jason Robins, ESPN’s Pablo Torre, and the Boston Celtics’ Mike Zarren on the accuracy of various sports myths.  With Norman, the hosts asked about athletes’ risk mitigation and the correlation between athletic performance and injury probability. Bird was asked to discuss gender systematic bias and bridging the difference in competition by gender. Robins then came in and dispelled any myth of illegality behind the likes of DraftKings and other (approved) sports betting venues.  Meanwhile, Torre followed by discussing a set of studies that have examined whether dating a Kardashian truly diminished athletes’ performances [1].  This topic’s serious academic discussion drove my attention away from what Zarren discussed to close the panel.  Perhaps that is part of the “Kardashian effect.”

Overall

Sloan was a great experience that broadened our awareness of sports analytics and the sports industry.  Researchers, speakers, and others we networked with gave us confidence that if we truly want to be in sports, we certainly can do it.  Thank you to those who donated to our Crowdfunding Initiative that provided the means to send the two of us to Sloan this year. We look forward to sending other deserving members to next year’s conference.  

Note

[1] Torre mentioned this blog post on the Kardashian effect by Harvard undergraduate Matty Cheng: http://harvardsportsanalysis.org/2019/03/does-dating-a-celebrity-affect-athletic-performance-a-closer-look-at-the-kardashian-effect/

Leave a Reply