The Antitrust Perils of Sports Data For U.S. Sports Leagues

Sports data account for only a small percentage of the revenue generated by U.S. sports leagues. However, as states legalize sports betting, the demand for such data is expected to grow. How leagues choose to meet this demand may raise antitrust issues under both Sections 1 and 2 of the Sherman Act. Two data-related theories of harm concern the collective sale of data at the league level and the leveraging of monopoly power over the market for league games to the market for data generated by or about those games. An empirical analysis of the market(s) for sports data would prove valuable to a rule of reason analysis of challenged practices.

By Gregory Pelnar¹

 

I. INTRODUCTION

The collection and monetization of sports data raises important antitrust issues for U.S. sports leagues.² One concerns the centralization of data ownership at the league, rather than the club, level in possible violation of Section 1 of the Sherman Act. Another concerns the possible leveraging of a league’s monopoly over the market for its games to the market for data related to those games in possible violation of Section 2 of the Sherman Act.

In the next section, I provide some background on the sports data industry, such as the types of data collected, who collects it, and how the data are used. I then discuss data-related theories of competitive harm as applied to U.S. sports leagues.

 

II. THE SUPPLY AND DEMAND FOR SPORTS DATA

The term “sports data” generally refers to all facts and information pertaining to a sports event or sporting competition.³ Suppliers of such data include sports leagues, individual clubs, and third parties. Some data can be replicated by third parties; others cannot. Users of sports data include clubs, the media, sports gambling operators, sports video game makers, and sponsors. Some data are bought and sold; some are not.

A. Supply

Some sports data are created by the leagues themselves, such as season schedules, or via league investments, such as wearable technology. Some are created by individual clubs, such as scouting reports, and are treated as trade secrets.⁴ Still other data are assembled by third parties, such as STATS. A key legal determinant of whether a third-party can legally assemble data that competes with the league’s data is whether the third-party is using the league’s data collection technology to assemble the data. In Morris Communication v. PGA Tour, the court concluded that Morris wanted access to the product of the Professional Golfers’ Association (“PGA’s”) proprietary technology without compensating the PGA so it could sell it for a fee to others, which would be a classic example of “free-riding,” and thus ruled that there was a legitimate procompetitive reason for the PGA’s restriction.⁵ Similar reasoning was used by the court in NBA v. Motorola which concluded that a Motorola handheld pager for displaying updated information about professional basketball games in progress did not constitute a misappropriation of “hot news” that is the NBA’s property. However, the court suggested that if Motorola had been free riding off the NBA’s data collection technology to offer its service, the court’s ruling would be different.⁶

The types of sports data can be divided into several categories: fixtures, event data, performance data, and refined data.⁷ Fixtures are the vital, yet mundane, decisions of leagues regarding such things as season schedules (e.g. when the season will start and end, how many games will constitute a season, who will play who when and where). Such data are easily replicated by third parties. Event and performance data include the vast array of data collected during the conduct of a sporting event and includes both external circumstances about the event (e.g. weather, attendance) and data regarding game performance (e.g. points scored). Much of these data can also be replicated by third parties by, for example, using trained observers to collect the information. However, not all performance data can be replicated, such as data collected from devices worn by players during the game. For example, the NFL embeds radio transmitters in players’ shoulder pads, thereby yielding data that can be used to calculate how fast a specific player runs over certain distances and which areas of the field they favor.⁸ The NFL owns the “next generation” data that are captured via the equipment of Zebra Technologies, and Sportradar, the NFL’s exclusive data distributor (in which the NFL has an equity stake of undisclosed size), sells the data, which can be used for such things as creating a heat map showing movements of an individual player. Other leagues have player-tracking systems as well.⁹

Clubs also collect performance and biometric data on their players. For example, MLB clubs use wearables like Readiband (which, e.g. can track how much sleep a player gets), pressure plates to track players’ explosiveness and strength, and pitch-recognition software to improve hitters’ ability to quickly identify balls as they leave an opposing pitcher’s hand.¹⁰ The line between a player’s privacy and a club’s need for at least some performance and biometric information about that player is blurred. Kristy Gale defines athlete biometric data (“ABD”) as “[a] measurable and distinguishable physical characteristic or personal behavioral trait used to recognize one’s identity, including but not limited to name, nicknames, likeness, signatures, pictures, activities, voice, statistics, playing and performance records, achievements, indicia, data, and other information identifying a particular athlete” and argues that ABD is usefully characterized as intangible property belonging to those who contribute the ABD.¹¹ However, in CBS Interactive v. NFL Players Association, the court ruled that the operator of a fantasy football website that used professional football players’ names and statistics did not violate any right of privacy.¹²

Another way to categorize sports data is raw versus refined.¹³ Raw data refers to the single event or performance data collected. Refined data uses raw data as an input to produce aggregated information in the form of statistics. Ownership of sports data has been the subject of litigation, with leagues asserting that certain data are their intellectual property (“IP”) and therefore subject to federal (e.g. copyright) and state IP protections. One area where this issue has arisen is sports gambling. Some leagues have argued that they have an intellectual property right to sport gambling proceeds and lobbied states legalizing sports gambling to mandate payment of an “integrity fee” equal to 0.25 percent of the wagered amount by the bookmaker to the league. Marc Edelman argues that state-sponsored sports gambling does not infringe the leagues’ federal and state law intellectual property rights.¹⁴

B. Demand

Demand for sports data comes from clubs, the media, the video gaming and trading card industries, fantasy leagues, sponsors, and the gambling industry, among other users.¹⁵ Clubs use sports data to analyze athlete and team performance. The print media uses line-ups, scores, and league tables, while the broadcast media uses all those plus real-time graphics. Makers of video games, such as FIFA 14 and Madden NFL 20, both produced by EA Sports, incorporate sports data into their software. Trading cards include player statistics. Fantasy leagues use sports data to assign points to participants based on actual player performances in games. League sponsorship agreements provide sponsors such as IBM with access to sports data. The sports betting industry relies on fixture lists as well as event and performance data to create bets. Companies like Sportradar specialize in the collection and distribution of live sports betting data and sell it to online betting companies such as Bwin and Betfair.

The demand for sports data has been recently spurred by the U.S. Supreme Court’s decision in Murphy v. National Collegiate Athletic Association (“NCAA”) that held that the Professional and Amateur Sports Protection Act (“PASPA”) violated the tenth amendment of the U.S. Constitution, thereby preventing the federal government from using PASPA to block states from legalizing commercial sports gambling.¹⁶ At least 12 states have legalized such gambling, with three – Tennessee, Illinois, and Michigan – going so far as to mandate that bookmakers buy and use official league data to determine certain wagers.¹⁷ The NBA and MLB had lobbied for such a requirement, arguing that only if bookmakers use official data can leagues ensure the integrity of sports betting and maintain consumer confidence in the game, but most states chose not to mandate the use of official league data.¹⁸ An attempt by Senators Hatch and Schumer to pass a federal law (the Sports Wagering Market Integrity Act of 2018) mandating the use of official league data for sports betting purposes failed.

According to a joint report by the Nielsen Company and the American Gaming Association, the direct and indirect impact of legal sports betting on the four major U.S. leagues will be more than $4.2 billion per year in additional revenue.¹⁹ Most of the revenue increase ($3.28 billion) is projected to come from an indirect impact on media rights, sponsorships, merchandise, and ticket sales. The second largest contributor to the revenue increase is direct TV advertising from gaming services ($596 million), followed by direct sponsorship revenue from gaming services ($267 million). The additional revenue from the sale of data and video is projected to be only $89 million. The NFL is projected to experience the largest revenue increase (in absolute terms): $2.3 billion, of which only $30 million will come from league data and product revenue for third-party gambling services. The corresponding figures for the three other leagues are: MLB ($1.1 billion, $28 million), NBA ($585 million, $25 million), and NHL ($216 million, $6 million). Thus, for the four major U.S. sports leagues, direct revenue from the sale of sports data is projected to account for less than 5 percent of the revenue generated by legal sports betting. An important implication is that, to the extent that official league data are a necessary input to the production of legal sports betting, the leagues have a strong incentive to sell the data since the indirect revenue generated by the sale is projected to far exceed the direct revenue from the sale of data itself.

 

III. DATA-RELATED THEORIES OF COMPETITIVE HARM AS APPLIED TO SPORTS LEAGUES

Given their approach to sports data, some sports leagues may find themselves accused of violating either Section 1 or 2 of the Sherman Act. Potential plaintiffs include sports data distributors and sports betting operators.

A. Potential Violations of Section 1 of the Sherman Act

Sports leagues that have collectivized the ownership and sale of sports data at the league level, such as the NBA and MLB, may be accused of a conspiracy in the restraint of trade of sports data in violation of Section 1 of the Sherman Act. The plaintiff may be a third-party such as STATS which finds that it must compete with a league’s exclusive distributor of sports data. Alternatively, the plaintiff may be one of the league’s clubs seeking to sell its sports data so as to avoid having to share the revenue with the league’s other clubs. It is even possible that the plaintiffs could be both a large-market team and the third-party to which it wants to sell its sports data.

In some respects, the former scenario is similar to that of the American Needle case. In December 2000, the NFL decided to offer an exclusive license for NFL team trademarks to use on headwear and apparel, rather than granting multiple licenses as it had done previously. Reebok was chosen for the exclusive license and American Needle, which previously had a license, sued the NFL alleging that the exclusive license constituted an unreasonable restraint of trade in violation of Section 1 of the Sherman Act. The case ultimately went to the U.S. Supreme Court, which in 2010 ruled that “the NFL’s licensing activities constitute concerted action that is not categorically beyond the coverage of §1” and “[t]he legality of that concerted action must be judged under the Rule of Reason.”²⁰ Now substitute “sports data” for “trademarks,” “Sportradar” for “Reebok,” and “STATS” for “American Needle,” and one has a hint of the arguments and issues that would arise in such an antitrust action.

But does the Supreme Court’s American Needle decision mean that the outcome of such an antitrust challenge is clear? STATS wins, the NFL loses? Not necessarily. A rule of reason analysis must be conducted. Possible anticompetitive effects of the collective sale of sports data must be investigated, as well as the potential procompetitive effects.

In a Section 1 case brought against Major League Baseball, the outcome may be simplified (or made more complex) by MLB’s antitrust exemption with respect to the “business of baseball.” If sports data fall within the scope of the business of baseball, MLB’s antitrust exemption applies and MLB prevails. However, the Curt Flood Act passed by Congress in 1998 removed employment-related agreements from the exemption and in 2014 a court ruled that MLB’s antitrust exemption does not extend to territorial broadcast restrictions (e.g. blackouts).²¹ Thus, MLB cannot be sure that courts will not find sports data to be outside its antitrust exemption as well.

As for a Section 1 challenge brought by a club, there is certainly some precedence. For example, in 1997, the New York Yankees accused the MLB and its member clubs of engaging in a concerted action “to combine and conspire together to restrain competition in the business and licensing of Club trademarks and of retail and wholesale baseball merchandise sales, and to misappropriate rights and revenues belonging to the Yankees and adidas.”²² In 1990, the Chicago Bulls challenged the NBA’s restrictions on the number of games that could be broadcast on superstations such as WGN, characterizing the NBA as a cartel whose television restriction limited the output of broadcast games in violation of Section 1 of the Sherman Act.²³

A Section 1 challenge by the U.S. Department of Justice is another (albeit seemingly remote) possibility. However, there is a precedent. In 1953, the DOJ challenged the NFL’s restrictions on telecasts in clubs’ home territories. In 1961, Congress passed the Sports Broadcasting Act exempting agreements concerning the sponsored telecasting of NFL, NBA, NHL, and MLB games from the antitrust laws. Congress could similarly pass a “Sports Data Act” antitrust exemption covering agreements concerning official league data of the four sports leagues.

There are other possible Section 1 theories of harm related to sports data. For example, one might put Sportradar at the center of a huband-spoke conspiracy. The company has about 90 percent of the U.S. sportsbook operator market, is partially owned by the NFL and three NBA owners (Michael Jordan, Mark Cuban, Ted Leonsis), and according to its own website has exclusive distribution rights agreements with the NFL, NHL, MLB, and Nascar, and has a non-exclusive betting data distribution rights agreement with the NBA.²⁴ Whether such a theory even makes economic sense may depend, in part, on the extent to which data from different leagues are substitutes. That is an empirical question.

One could argue that if a sports league sells data to a distributor, which in turn sells it to a sports betting operator, the latter is an “indirect purchaser” and thus denied antitrust standing by Illinois Brick.²⁵ However, the sports betting operator may be considered a direct purchaser under Apple v. Pepper.²⁶

B. Potential Violations of Section 2 of the Sherman Act

A Section 2 challenge to a league’s approach to sports data would likely allege that the league is leveraging its monopoly over league games to create a monopoly over league-related data. The exact mechanisms by which the league is alleged to do so may vary. For example, the league may be accused of preventing third-parties from collecting data, degrading the usefulness of the third-party’s data by requiring use of official league data, or denying access to an “essential facility” (e.g. next-generation sports data) for offering in-game betting. Some data practices of sports leagues may be covered by the state action immunity doctrine, such as the NBA’s and MLB’s successful lobbying that resulted in Tennessee, Illinois, and Michigan including a requirement for the use of official league data in their sports betting laws.²⁷

The plaintiff may be a third-party attempting to collect and sell sports data in competition with a league. One potential plaintiff is STATS, which was the NFL’s partner for collecting and disseminating sports statistics to major media outlets prior to the 2015 season but irked the NFL by licensing NFL statistics in 2010 to New York Life Insurance to build the “New York Life Protection Index” measuring offensive line play around the league even though the insurer was not an NFL official league partner.²⁸ In Europe, Sportradar has recently legally challenged as anticompetitive a five-year deal between Genius Sports and Football DataCo (“FDC”), the data rights holder for the English Premier League, English Football League, and Scottish Football League, that gave Genius Sports exclusive rights to collect, license, and distribute live data from those three leagues to sportsbook operators.²⁹

A rule of reason analysis of a Section 2 claim would entail definition of the relevant product and geographic markets, as well as economic rationales (e.g. efficiencies) for the challenged practices. Product market definition would require an inquiry into the extent to which substitutes exist for the official league data. In the case of next-generation data, there may be no good substitutes. In the case of official and unofficial data for in-game betting, opinions differ, with some reports that unofficial data tend to have a latency of several seconds relative to official data, making the latter much preferable, but with others disputing that data from third-party sources are inferior to official league data.³⁰ Ultimately, the relevant product market will depend on the empirical evidence. To what extent do bookmakers buy only the official league data feed? To what extent do they rely only on unofficial data? A natural experiment may help answer such questions. Tennessee, Illinois, and Michigan mandate the use of official league data for betting purposes, other states do not. How, if at all, does the usage of official and unofficial data differ across bookmakers in different states? As sports betting grows in popularity, the data needed to answer such questions will become available.

The relevant geographic market may arguably be an individual state, the entire U.S., or the world, depending on the Section 2 allegations. For example, in the case of sports betting, different states have different regulations. On the other hand, betting on games of U.S. sports leagues occurs around the world and those bookmakers need sports data, either official or unofficial.

Efficiency rationales for the challenged practices would likely be an important part of a rule of reason analysis. As noted above, the prevention of free-riding in the collection of sports data has been recognized by courts as a sound economic reason for restricting data collection by third-parties. Also, there could be a negative externality on consumers (e.g. fans) if betting-related events raise questions about the legitimacy of game outcomes. There may be transaction cost savings from bundling next-generation data, over which the league has a monopoly, with other types of league data that compete with data collected by third-parties. Once again, efficiency rationales may differ depending on the specific allegations.

 

IV. CONCLUSION

Data-related theories of competitive harm are being widely discussed as they apply to digital platforms such as Google and Facebook, and some other industries, including agriculture.³¹ Much less attention has been paid to possible competitive harm related to sports data, but that has begun to change as states legalize sports betting and some make the use of official sports data a legal requirement. A data-related antitrust challenge to a sports league sometime in the next few years should not come as a surprise.

 

---

1 Vice President, Compass Lexecon.

2 See Marc Edelman, Sports Data Policies Could Represent Next Big Antitrust Challenge for Pro Sports Leagues, FORBES (June 10, 2019), https://www.forbes.com/sites/marcedelman/2019/06/10/sports-data-policies-could-provide-next-big-antitrust-challenge-for-pro-sports-leagues/#6b1fa24b3284; John Holden, Integrity Fee Issues for NBA and MLB Run Deeper than They Appear, LEGAL SPORTS REPORT (May 10, 2019), https://www.legalsportsreport.com/32378/holden-nba-mlb-integrity-fee/; Ryan M. Rodenberg, Antitrust Standing after Apple v. Pepper: Application to the Sports Betting Data Market, 64 ANTITRUST BULLETIN 584-93 (2019); and Brett Smiley, Antitrust Tripwires: Legal Expert Explains Sports Betting Data Issues, SPORTSHANDLE (June 4, 2019), https://sportshandle.com/sports-betting-data-antitrust/.

3 Christian Frodl, Commercialisation of Sports Data: Rights of Event Owners over Information and Statistics Generated about Their Sports Events, 26 MARQUETTE SPORTS LAW REVIEW 55 (2015), at 56.

4 Lara Grow & Nathaniel Grow, Protecting Big Data in the Big Leagues: Trade Secrets in Professional Sports, 74 WASH. & LEE L. REV. 1567 (2017).

5 Morris Communication Corp. v. PGA Tour, 364 F.3d 1288 (11^th Cir. 2004).

6 NBA v. Motorola, 105 F.3d 841 (2^nd Cir. 1997).

7 Frodl, supra note 3, at 57-59. For a different categorization of sports data, see Rodenberg, supra note 2, at 588-89.

8 Daniel Kaplan & Eric Fisher, NFL Buys Stake in Stats Firm: Europe’s Sportradar Will Replace Stats LLC, STREET & SMITH’S SPORTS BUSINESS JOURNAL (April 20, 2015), https://www.sportsbusinessdaily.com/Journal/Issues/2015/04/20/Leagues-and-Governing-Bodies/NFL-sportradar.aspx.

9 See, e.g. Ira Boudway, The NBA Will Now Track Every Player’s Movements, BLOOMBERG (Sept. 6, 2013), https://www.bloomberg.com/news/articles/2013-09-06/the-nbawill-now-track-every-players-movements.

10 Rian Watt, New Technologies Are Forcing Baseball to Balance Big Data with ‘Big Brother’, VICE (May 27, 2016), https://www.vice.com/en_us/article/8qygbp/new-technologiesare-forcing-baseball-to-balance-big-data-with-big-brother.

11 Kristy Gale, The Sports Industry’s New Power Play: Athlete Biometric Data Domination. Who Owns It and What May Be Done with It?, 6 ARIZ. ST. SPORTS & ENT. L. J. 7 (2016), at 11.

12 CBS Interactive v. NFLPA, 259 F.R.D. 398 (2009).

13 Frodl, supra note 3, at 59.

14 Marc Edelman, Lack of Integrity? Rebutting the Myth that U.S. Commercial Sports Leagues Have an Intellectual Property Right to Sports Gambling Proceeds, 15 NEW YORK UNIVERSITY JOURNAL OF LAW & BUSINESS 1 (2018).

15 Frodl, supra note 3, at 63-65.

16 Murphy v. NCAA, 138 S.Ct. 1461 (2018).

17 Keven Draper, Sports Betting Has Arrived to Transform the N.F.L. Or Not., NEW YORK TIMES (Sept. 4, 2019), https://www.nytimes.com/2019/09/04/sports/football/sports-betting-nfl.html; Matt Rybaltowski, Implications of Landmark Sportradar Lawsuit Unclear on US Sports Betting Market, SPORTSHANDLE (March 4, 2020), https://sportshandle.com/sportradar-2020-landmark-suit/.

18 Matt Rybaltowski, Shakedown Fees: NBA, MLB Demanding Nevada Sportsbooks Pay More or Get Cut Off, SPORTSHANDLE (May 2, 2019), https://sportshandle.com/nba-mlb-demands-data-fee-nv-sportsbooks/.

19 Nielsen Company & American Gaming Association, How Much Do Leagues Stand to Gain from Legal Sports Betting?, https://www.americangaming.org/wp-content/uploads/2018/10/Nielsen-Research-All-4-Leagues-FINAL.pdf.

20 American Needle v. NFL, 560 U.S. 183, 186 (2010).

21 Laumann v. NHL, 56 F.Supp.3d 280 (S.D.N.Y. 2014).

22 New York Yankees Partnership and Adidas America v. MLB Enterprises, complaint filed in Florida federal court (May 6, 1997), at par. 5.

23 Chicago Professional Sports Limited Partnership & WGN Continental Broadcasting Company v. NBA, 961 F.2d 667 (7^th Cir. 1992).

24 Rybaltowski, supra note 18; Eben Novy-Williams, NFL Takes First Major Gambling Step with Sportradar Data Deal, BLOOMBERG (Aug. 12, 2019), https://www.bloomberg.com/news/articles/2019-08-12/nfl-takes-first-major-gambling-step-with-sportradar-data-deal; NBA, NBA Announces First Betting-Data Partnerships in U.S. with Sportradar, Genius Sports (Nov. 28, 2018), https://www.nba.com/article/2018/11/28/nba-sportradar-genius-sports-partnership-official-release; and the Sportradar website, https://www.sportradar.com/about-us/our-partners/.

25 Illinois Brick v. Illinois, 431 U.S. 720 (1977).

26 Apple v. Pepper, 139 S.Ct. 1514 (2019). See Rodenberg, supra note 2.

27 See, e.g. Rodenberg, supra note 2, at 592, and Smiley, supra note 2

28 Danny Ecker, Stats Loses NFL Deal to Sportradar, CRAIN’S CHICAGO BUSINESS (April 20, 2015), https://www.chicagobusiness.com/article/20150420/BLOGS04/150429968/stats-loses-nfl-deal-to-sportradar.

29 See Rybaltowski, supra note 17.

30 See, e.g. Rybaltowski, supra note 18.

31 See, e.g. Jacques Crémer, Yves-Alexandre de Montjoye & Heike Schweitzer, Competition Policy for the Digital Era (2019), https://op.europa.eu/en/publication-detail/-/publication/21dc175c-7b76-11e9-9f05-01aa75ed71a1/language-en; and Thomas J. Horton & Dylan Kirchmeier, John Deere’s Attempted Monopolization of Equipment Repair, and the Digital Agricultural Data Market – Who Will Stand Up for American Farmers?, CPI ANTITRUST CHRONICLE, volume 1(1), 21-27 (January 2020).