How Analytics Can Boost Competitiveness in Sports
A business journal from the Wharton School of the University of Pennsylvania
Wharton’s Eric Bradlow and FanDuel CEO Amy Howe discuss how analytics can benefit teams, players, and customers while ensuring the necessary data protections.
Data and analytics are increasingly being used to help maximize fan experiences and for team owners to determine the value of players, but also to enhance player safety. The recent episode of Buffalo Bills player Damar Hamlin going into cardiac arrest brought into focus the role analytics could play in spotting player injuries in real time to take remedial action. At the same time, the sports industry must ensure the security and privacy of player and customer data, according to participants in a Wharton panel discussion on January 19.
While it is common for team managers to use data analytics to evaluate player worth, it is finding uses in several other settings as well. For example, sensors could be attached to athletes while they’re training, or while they are sleeping to identify sleep patterns and how they are related to better performance, Bradlow said.
“It’s about the data. The richer the data, the better the analytics will be,” Bradlow said. “Some sports are still in the nascent stages of getting data. There isn’t the same data in soccer [as] there is in basketball, football, MLB, et cetera.”
Bradlow is also co-host of Wharton Moneyball, a weekly radio program on SiriusXM that explores how decision-makers in sports can avoid common mistakes and embrace data. The show began with interviews around baseball, but has since expanded its focus to include other sports like soccer, tennis, and even ultimate frisbee.
Data and analytics are integral to FanDuel’s business model. “Our entire business and our entire sports betting platform depends on real-time data, sophisticated analytics, and modeling,” Howe said. FanDuel’s parent company — Irish gambling holding company Flutter Entertainment — includes a risk and trading organization that has about a thousand data scientists, statisticians, and engineers, she pointed out. FanDuel uses analytics also in its CRM (customer relationship management) strategy to identify risky behavior, flag problematic behavior, “and do something about it,” she added.
Online ticket sales also offer new opportunities to use data and analytics to improve customer experience. Back in the day of paper tickets, it was challenging to identify customers, Howe noted. “Nowadays you’re getting in with your phone, and most venues at this point are 100% digital,” she said. “That’s a great example of the use of data and technology to identify customers. When you enter [a venue] with your phone we can identify — not yet one for one — a large percentage of that organization. Your ability to personalize the experience, and your ability to deal with crisis in the event there’s a security breach, or manage operations within a venue, goes up exponentially as a result of changing technology and data and analytics.”
Analytics at sports betting companies also informs their decisions on entering new markets. They use data science to determine which markets to target, how much to invest, and to evaluate customer acquisition cost relative to their lifetime value and player value “to make sure that our paybacks are within the right range,” Howe said. For instance, when Ohio recently permitted online sports betting, FanDuel customers in Michigan (where it was already legal) were concerned about whether they had to open different accounts for betting in Ohio. FanDuel used analytics “to look at some of the behavioral patterns and fix the friction before it became a problem,” she noted.
Bradlow listed two markers for the sports betting industry as it tackles the issue of data security and privacy. One is the option of “data perturbation,” a form of data mining that protects it from unauthorized use. Here, companies have to decide the level of data aggregation and granularity they want to meet their business needs while ensuring data security and privacy, he said.
The other marker is the complex question of determining who owns the data. Such a question might arise in a situation where, for example, a team gathers data from sensors on a player’s physical factors and wants to use it to determine the player’s salary, he explained. “We’re going to have an era where public policy will have a lot to say about who owns the data, and who gives up rights to use their data for our analytics engines.”
Data security and privacy is a particularly sensitive issue in the highly regulated sports betting industry, and “it’s one of the things that keeps me up at night,” said Howe. “What matters most to consumers is that they have a sports betting platform that they trust. And your ability to make sure that you can protect that data, to protect them from bad actors and cybersecurity threats, is ultimately critical.”
In states where sports betting is legalized, companies like FanDuel have to undergo a rigorous process of securing customer information that includes their name, Social Security number, email address, phone number, and bank account information, she added. “There are plenty of illegal operators out there where you don’t have to provide any of that information.”
Sports betting companies must have data governance mechanisms that ensure high integrity of the information that they gather, which in turn means obtaining that from official data providers. In addition to the insights that analytics provides, sports betting companies also rely on “quite a bit of judgment,” Howe said. “Things may change at the last minute that can impact the outcome of a game. There’s a lot of different signals that you’re taking in — hard data and science. But at the same time, part of the job of a trader in a sports betting company is also to overlay judgment on top of that.”
At FanDuel, Howe also faces variations across different sports in the richness of data, but she has to work around those to stay competitive. “Not all sports obviously are created equal if you think about sports betting. [But I have to] make sure I have the most competitive markets, the widest assortment, and real-time data feeds,” she said. So, irrespective of the level of fan engagement, FanDuel packs in sports such as golf, tennis, and of late, even pickleball, she added.
For sure, analytics does not resolve all questions. “I’ve always taken exception to the phrase ‘data-driven’ because it assumes that the data is always correct,” James said. “It leaves out the role of judgment, the role of personal information, history, culture, and all of these other things that aren’t part of an algorithm,” she explained. “It is better to be data-informed and still leverage the expertise, the culture, the relationships, and the history that only comes from the coach.”
“The perfect clean dataset with all of the variables you would like to measure has never come in,” said Bradlow. “Every situation, every business decision has incomplete information.” While algorithms can help solve prediction problems, there are many things that go into decisions on whether a particular solution is fair, equitable, and inclusive, he added.
Bradlow shared that in his work in designing analytics models for the Philadelphia Eagles, he emphasized that algorithms can be used as “a decision-support tool to support judgment,” but they cannot guarantee fairness or eliminate risk.
Howe shared the example of how during COVID, many players fell sick at the last minute and had to sit out of a game. That raised questions of how at the same time, sports betting companies could make sure that they were being fair to their customers. FanDuel responded with a product called Bad Bet Relief. “If something bad happened at the last minute that you couldn’t control and you had already placed a bet, we’d either refund your bet or we’d find a way to make you whole,” she said. “So, as much science as there is in this industry, there’s also a lot of art and judgment that factors into it as well.”
James pointed out that the Damar Hamlin episode raised questions about the safety of sports, and the protection of its athletes. Analytics and big data can be leveraged to increase the safety of athletes, Bradlow said. At the Wharton Neuroscience Initiative, experts can measure brain data at a granular level and measure an athlete’s brain function over time, he added. He said he hoped that in the future, while athletes are on the field, real-time analytics can spot injuries if a player is “not performing optimally.”
More broadly, analytics will continue to evolve and advance the sports betting industry in especially two specific areas that Howe identified. One is in risk assessment where innovation will help “bring new products to market that are elevating the sports betting experience,” she said. The other relates to the impact of changing technology, such as the metaverse and its “material impact on user experiences, and how consumers engage in sports betting,” she added.
The sports industry could also learn from other industries such as financial services for ease of payments, or from Amazon in removing friction from customer experience, Howe said. “At the end of the day, consumers want a simple and easy betting platform. They want to be able to deposit and withdraw their winnings very quickly and easily.”
In response to an audience question, Bradlow noted that even with the power of data and analytics, gut instinct is still valuable. “I use gut intuition all the time in forecasting the outcome of games,” he said. “But the first thing I want to know is what does the betting line say, what do the predictive models say. Start from there, and you can always deviate from there based on information that I may have that the model might not have, or information it hasn’t captured.”