By: James Eby, Suvrath Penmetcha (@savvysuv), Arvind Pendurthi (@arvindpendvrthi), Achyuta Burra, Maksim Horowitz (@bklynmaks), and Steven Silverman (@Silver_Stats)
Below we recap an incredible second day at the Sloan Sports Analytics Conference:
Participants were given frame-by-frame player tracking data from the 2016 Under Armor All-American Football game. They were given several lines of inquiry to pursue, including visualizing wide receiver routes and quantifying how offensive lines control the line of scrimmage. All the presentations were very impressive for such a short time frame (about six hours), and the team at ESPN Stats and Info did a great job in setting up the event. CMU Student and TSAC Vice President Steven Silverman was one of three finalists in the Student Division.
Basketball Analytics: Hack-a-Stat
A common theme throughout the panel is the idea of wearable technology and how it can be used to enhance player training, injury prevention, and player management. One example the panel mentioned was how wearables measured the distance Rajon Rondo ran in a playoff game. This is extremely helpful for player training since it can help trainers figure out how to condition players so they can sustain the rigors of the season. The panel also discussed how wearable technology needs to be allowed in NBA games to have a greater impact. This is because tracking player movements during practice is not indicative of the intensity of a game.
Modern NBA Coaching: Balancing Team and Talent
A panel of former NBA coaches talked about how the popularity of “space and pace” in the NBA and how the success of this style of play is personnel dependent. The coaches spoke to tailoring offensive schemes to their best players. For example, the Spurs adjusted to run a slower paced offense because of the addition of LaMarcus Aldridge even though over the past few years they were know as a fast paced team. The panel also showed evidence of the work to still be done in regards to analytics in basketball. The coaches were asked about their use and reliance on analytics over their gut and it was clear that these NBA coaches did not value or understand analytics at a high level.
Optimal Ticket Pricing: The Frontier of Quantitative Analysis with Primary and Secondary Market Data
Drs. Robert Stavins and Dr. Matthew Notowidigdo shared their experiences from helping professional teams determine optimal pricing models for their stadiums. They used an econometric model with dozens of inputs (e.g. secondary market price, seat location, gameday weather, opponent) to determine a price for every seat in a stadium on a given day. Through this analysis, they have helped teams maximize attendance and ticketing revenue.
Business of Sports
The panel began by discussing the difference between American sports leagues, which engage in revenue sharing, and European football leagues, which do not. Because of this arrangement, American teams have incentives to share business best practices with each other, while European soccer teams do not.
They also talked about how to increase fan engagement in efforts to build a “global brand.” Fan engagement, both in person and online, can help teams learn more about their fans’ needs and create new fans. The NBA has been making a concerted effort to reach fans in countries like China and India to build a global market, much like soccer has.
A quick note from the end of the panel: both of the NBA executives on the panel, Steve Pagliuca and Tad Brown, admitted that it’s only a matter of time before we see sponsors on NBA jerseys.
How do you stop Leicester city? Advanced tactical analysis in the English premier league.
Dan Altman, founder of North Yard Analytics, revealed why Leicester’s current exceptional performance is not as surprising as it seems to the casual observer, even when compared to last season’s performance which almost ended in relegation:
1) Leicester’s analytics department used Opta soccer data to assess which players may make a contribution to the team, and spent €50 million in new players for this season, including Kante, Huth and Fuchs.
2) Leicester city plays a direct game with high pace. Generally possessions with fewer passes and direct progress on the pitch are more successful. The players purchased for this season are some of the most important passers, strengthening the team’s play style.
3) Leicester also signed Claudio Ranieri as manager, whose long career has had many successes.
These key changes in Leicester’s organization, along with a little luck, explain much of why Leicester has been so successful this season.
NFL ticket pricing
David Highhill, director of the NFL Club Business Development Group, provided a detailed insight into some of the facets of ticket pricing for NFL organizations. Most of the revenue from ticket sales for NFL teams not only happens very early within a season — most tickets sold within a month after they go on sale, before the regular season– but it is largely composed of season ticket holders. Therefore there is little to do when it comes to single game sales and variable pricing. David proposes a few key points to consider when pricing tickets:
1) Primary Market
Understanding the venue you are working with is an essential way to optimize ticket pricing. One good strategy for pricing is to get to know the venue in person; blueprints and plans, even in 3D, don’t give the full story of a seat in the venue. Understanding the geographical nuances of fields, such as where there is shade or the view from a particular section is, can be useful when tweaking prices to maximize profit, attendance, and fan retention. One main variant in ticket prices within a season is the opponent a team is playing; rivalries and strong match-ups allow for higher ticket prices.
2) Secondary Market
You also need to analyze the secondary market for tickets; the relative price of re-sold tickets compared to original price varies greatly by row, with both positive and negative differences. Understanding the resell value of seats allows for further optimization in pricing. However, these changes should not be equal to the difference in original and re-sell prices. Ticket holders often purchase tickets because of the great value in re-sell so it will not be effective or fair to the consumer. Small changes in prices over several years can be a more effective way than a large increase.
3) Amenities and Benefits
Often ticket holders will be more willing to purchase a ticket or season tickets when it comes with benefits, and knowing what those can be or creating more benefits should allow for a more optimal pricing system.
If there is only one thing Sloan has taught us it is that communication is key in the analytics industry in order to successfully apply solutions drawn from your results. If others cannot understand something, it is unlikely they will apply your results. Getting to know how you should communicate with each department will be imperative to your success as an analyst. You can’t communicate with the sales department the same way that you communicate with the executives as they have different goals and skills. Possibly more important is to create interact with all departments as a whole, since every decision made with regards to pricing ultimately affecting the “spider web” of all the different departments in a business
Research Paper finalists
The winning research papers were from four tracks (one each): baseball, basketball, other sports, and sports business. The baseball paper introduced a new way of looking for the “hot hand” or streakiness effect, finding a much larger magnitude of such an effect than most previous literature. The basketball paper classified ball screens through video analysis, splitting them by how the defense responded (going over, going under, trapping, or switching) and examining how each strategy affected the offense’s points per possession. The other sports paper (which was named the overall winner) developed a model for win probability in tennis at a shot-by-shot level, incorporating such factors as the style of play, the length of the rally, and environmental variables. Finally, the business paper looked at cross-brand engagement by using geolocation services, attempting to quantify how fan bases of certain teams interact with other products to determine how to customize advertising at each stadium and for each fan base. All of the papers were very impressive and exemplify the leading edge of sports analytics research.
One-on-One with Nate Silver
With Daryl Morey interviewing the audience had a chance to ask questions to FiveThirtyEight’s Nate Silver. Focusing mostly on politics, Silver talked about his different political projection systems and how predicting politics is much more difficult than sports due to small sample size and the large number of predictors. Silver also spent time talking about the GOP candidates and what he expects to happen for the Republican Party. On the sports side, he discussed some new interactive features FiveThirtyEight is adding to its website especially in regards to Men’s College Basketball.