2018 MIT Sloan Sports Analytics Conference Recap

Written By: Scott Steinberg, Nathan Jen, & Ansh Hirani

There are many reasons why the MIT Sloan Sports Analytics Conference generates so much hype each year. Houston Rockets general manager and basketball analytics leader Daryl Morey is one of the co-founders of the conference. Shane Battier and Brian Burke, among others, can be easily be spotted in conversation or strolling down the hallways of the convention center. Representatives from almost every sports team can be found at this yearly gathering, making it a hotspot for college students to make a name for themselves. The CMU Sports Analytics Club was fortunate to send Vinay Kukutla, Scott Steinberg, Nathan Jen, and Ansh Hirani to Sloan this year, and we are incredibly grateful to everyone who contributed to our crowdfunding campaign. Here’s a recap of some of the panels and research presentations we went to:

Hockey Analytics on the Fly
by Nathan Jen

This panel, which featured Dan Bylsma, John Chayka, and Chris Snow, was mainly centered around the growth of analytics in hockey, and how NHL Front Offices use analytics to try to gain a competitive advantage. To illustrate, a lot of NHL teams use analytics to understand why their team is trending in a certain direction. Front offices of teams on a winning streak will constantly try to use analytics to understand what area of the game the team is excelling in, and how the that team can continue excel in this area.

One interesting point brought up in the panel is the different responses that each player has to analytics and how there are right and wrong times to use analytics. They pointed out hat there are players who will use and ask for analytics to help them improve areas they are weak in. For example, if a player sees that his face off win percentage is steadily decreasing, he will then use more practice time to work on his face-offs to help bring that percentage up.

On the other hand, there are also players who despise analytics and will never want to know about their stats in depth. This can be due to a variety of reasons. For instance, players might not want to know what analytics is saying about them because they are scared that it might mess with their head. If a goaltender believes that he is having a really good season because of his high save percentage, it might not always be beneficial to show the goaltender that he is facing very few high danger shots. That way, that goaltender will hopefully continue to play at an elite level due to his sky high confidence level.

Finally, the panel discussed the future of hockey analytics, and the probable next steps. One interesting technology that is on the horizon, and can revolutionize hockey analytics is player tracking. Player tracking would be incredibly beneficial to a sport like hockey because it would allow teams to see how fast players are with and without the puck. It can also help show teams why they are struggling in certain areas, such as the penalty kill, where they can see where the holes in the defense are created.

NBA 2.0
by Nathan Jen

The NBA 2.0 panel, focused on how basketball can be improved by adopting new rules, and featured Ben Falk, Evan Wasch, Mike Zarren, and Rafael Stone as speakers. With the NBA being as popular as ever, this panel really challenged the notion of whether or not it is necessary to think of new rules for the next NBA. However, the decision that the panel decided on is that this is the perfect time to innovate the product, because there are very little drawbacks to making the NBA as good as it can possibly be (Whether that means decreasing the amount of dead time, or increasing player safety).

It is awesome to see the NBA take innovation and new ideas so seriously because in an age like today, everything changes so rapidly and the NBA will need to make constant adjustments to keep growing its popularity. This culture of innovation can already been seen through the changes to the draft lottery. With the NBA viewing the draft as a process that balances the competition within the NBA (instead of a way to find homes for young, talented basketball players), the old draft system gave the teams at the bottom the best chance at receiving the top pick in draft. Under the new system, the odds at the bottom to receive the top pick are flatter, meaning that teams at the bottom will not have a much chance at receiving the top pick in the draft compared to a team for instance that has the 5th worst record.  This should be beneficial to the NBA because without as great of an incentive to tank to the bottom, the intensity of each game will hopefully increase as each team will play hard to get a win each time they step onto a basketball court.

The discussion that I loved most during this panel, however, was the conversation about how to increase the flow of a basketball game and limit the times when play is stopped. One interesting point brought up was after some testing done by the NBA, the three factors that deter people from watching the product more was commercials, time-outs, and free throws. While not much can be done to cut down time during the commercials or time-outs, there are a lot of interesting ideas to make free throw shooting more efficient. One idea brought up was to only allow one free throw shot, and having that shot worth as much as the foul is. So that one shot would be worth 3 points if the shooter was fouled shooting a 3, and 2 points for a shooting foul in the post. Since this idea is so drastic and deters too far away from the fundamentals of basketball, it will probably never be implemented. But I can’t help but support drastic ideas getting tossed around because that is the only way that the NBA is going to get better. With Adam Silver developing a new culture centered around innovation, I am excited to see how big the game can grow, and what the NBA will look like in 25 years.

Is AI The Answer?
by Scott Steinberg

Artificial intelligence is a central part of any conversation about technology today. However, what exactly is “AI”? As defined by Patrick Lucey, the Director of Artificial Intelligence at STATS, AI is “technology that emulates humans at a particular task.” This panel, which also included former Philadelphia 76ers general manager Sam Hinkie, discussed the possibilities of sports applications in artificial intelligence for now and the future, from practices as simple as detecting pick-and-rolls in basketball to “downloading” brains of quarterbacks to gain understanding of their decision-making on the field.

“If you feel like the world has changed a lot in the last 24 months, just wait,” stated Hinkie. At the end of the day, the panelists agreed that communication will be key to pushing AI to greater limits. “The more and more complicated [the numbers get], you get to a smaller part of the funnel of the people who understand,” Hinkie later added. “It’s the job of the people--if you really want to be effective, [you need] to communicate it in a way that [everyone] can take it in.”

Competitive Advantage
by Ansh Hirani

At the SSAC, there was no shortage of interesting “competitive advantage” talks to learn from. Research relating to player health and performance analytics were at this year’s forefront.

Leading one of the more interesting talks (in my opinion) was Dr. David Martin, who spoke on “wearable technology”. In Martin’s research, Martin addresses the problem of deciding when an injured player is eligible to return. He references instances in the NBA and NFL where injured players are participating in practice, but not necessarily playing in the games. The ultimate question being asked is “When is the threshold point for players to return to play games?”. Martin’s research addresses this question through the use of wearable sensors that monitor a player’s health during practices. The purpose of these sensors is to keep track of muscle, ligament, and tissue development as the players go through rehab and practice. Using this sensors provide a great tool for health preservation since it can precisely monitor what routines cause health improvement and what may cause setbacks. As a result of this, these sensors help better understand player health and when it is appropriate for them to return to the game. However this technology still needs development before it can go to the masses. Currently, Martin’s lab is working on data preservation and minimizing errors through the sensor’s reportings.

Other talks worth mentioning are Dr. Aaron Baggish’s talk on physiological doldrums and Dr. Meeta Singh’s work on biological clocks. In Baggish’s talk, Baggish conducts research on human physiology to optimize athletic performance. Baggish spoke of how the human body has an “intensity” threshold for physical activity, where if worked beyond this threshold, the body exhales more CO2 than it inhales oxygen. In Baggish’s research, he tries to find this person-specific unique threshold and builds a workout routine to maximize athletic performance. In Singh’s talk, Singh talks about innate biological clocks and how people fall into one of “morning”, “midday”, or “evening” clocks. Depending on the type you are, that is when you will perform the best. Singh’s work is designed to categorize people and then build sleep schedules and workout routine for players in sports like football and baseball where travel is an important factor in player performance.

Trust the Process? Team Building and Rebuilding in the NBA
by Scott Steinberg

The final panel of the conference featured varying perspectives from both on and off the basketball court. Hinkie and 2-time NBA Champion Chris Bosh were both featured on this panel, as well as former Cleveland Cavaliers general manager David Griffin, Los Angeles Clippers Executive Vice President of Basketball Operations Lawrence Frank, and Boston Celtics co-owner Steve Pagluica.
Hinkie’s controversial rebuilding strategy with the Philadelphia 76ers, popularly known as “The Process” (a phrase Hinkie denied ever using), put him at the center of attention on this panel. Griffin, when asked about the success of Hinkie’s strategy, called it a success, complimenting Hinkie’s methods of using past NBA rebuilding models to maximize the probability of building a championship contender.

“The most important thing we do in all sports is mimic the people that are successful and figure out what they did to be successful and try to recreate that,” said Griffin.

Pagluica, on the other hand, disagreed. “I don’t think by definition the process as defined by losing all you can for four or five years in a row is actually the way to get to the promised land,” he said.

The 76ers are currently enjoying their first winning season since Hinkie originally took over as general manager in 2013. Whether or not their rebuild will be recognized as a success in the future is still up in the air, but given the assets Hinkie acquired and the position he put the team in before resigning in 2016, the 76ers are primed to at least be a major contender in the Eastern Conference for years to come.


Leave a Reply