Sports Betting – A Golden Age By Andrew Tottenham, Managing Director, Tottenham & Co April 24, 2019 at 2:45 am Despite the efforts of the US Department of Justice, the United States is crossing the threshold into a world of legal sports betting. Admittedly this is on a state-by-state basis, and the first states that have legalised sports betting have tended to be ones with smaller populations. So, to date, revenues have not been something to write home about. However, as the number of people that can bet increases, so does the demand for sports data. And as demand for sports data increases, so does its value. One of the first people to devise a method of recording information about a game was Henry Chadwick, an Englishman brought up in Brooklyn in the early part of the twentieth century. He is credited with creating the display of information about a completed baseball game that is called a box score. He is also responsible for developing statistics such as batting averages and earned run averages. Today this information is used by both bookmakers and bettors. Bookmakers require historic data about games and races to help them set the odds. Bettors use the same data to inform their decision-making about whether the odds offered on a specific event might be advantageous, and if so, how much they should bet. The rise of betting exchanges, where customers can bet against each other, has led to a new breed of bettor, one who scours the matches and the odds being offered and accepts or creates multiple bets in the knowledge that the likelihood of them resulting in an overall profit is in their favour. Some have created algorithmic trading platforms, allowing them to place bets automatically. Unsurprisingly, bookmakers do not like customers who win consistently, and will limit the size of their bets or block them completely. In Europe, sports betting accounts for almost 40% of online gambling revenues, and much of the growth in recent years has been put down to an increase in “in running” or “in-play” betting – betting during a game. Bet365, one of the most successful online betting companies, has reported that almost 80% of all their sports betting revenue is derived from in-play betting. Clearly it has become a very important revenue line. Instead of there being only a few bets per game (for example, who will win? By how much?), each game now offers far more opportunities to bet. Examples of in running bets on football [soccer] could be, for example, which player will score the next goal, what will be the exact time of the next goal, and if one team is taking a corner, will they score? I note that some of the debate in the United States is whether or not to allow such betting. Naysayers believe the opportunities for corruption are too high and that when corruption happens, it will destroy a sport´s image. But if these types of bets are not allowed, it will choke off an important revenue stream, and could push newly-educated bettors to the unregulated market for wider opportunities to bet. Offering in-play bets requires timely data. That is fairly simple for big events that are televised, but for smaller, less publicised events, how is that data to be collected? Some might be surprised to learn how unsophisticated the process is. Companies such as Opta and Sportradar employ teams of data analysts who watch the games live on television and convert what they are seeing into data points. One game might be the equivalent of 2,000 data points. A data point could be a tackle, a shot on goal, a pass, or an interception. To ensure the information that recorded is not biased, one analyst might be responsible for one team and a second analyst for the other team, with a third analyst being called in to arbitrate any discrepancies. This reliance on people to convert information to data points is expensive and can only be justified by scale, amortising the cost over a number of clients. Increasingly, data providers that are approved by the official sports bodies to collect and disseminate sports data are contractually obliged to provide their services only to regulated bookmakers. I will not get into the debate about who owns the data from a game or race, or whether officially-sanctioned data companies are a good thing, leaving these topics for another article. But the enormous size of the unregulated betting market has led to a demand for data from unapproved sources. That demand is so great that it is not unusual at football, cricket, or tennis matches to see a spectator with hood or collar turned up, talking into a phone, usually via a microphone hidden behind the collar or hood. These are the unofficial data providers of modern sport. They comment on the match to the person on the other end of the phone, who converts their words into data points that are fed to betting customers around the world. If such spectators are spotted at the match by security, they will be ejected from the grounds, but there are usually more than one at each game, so the data feed continues. Another method of data capture is for players (or horses, for that matter) to wear transponders so that their movements can be automatically tracked. This method is not very good for contact or fast-moving sports because the transponders frequently fall off and it is not very accurate. It also requires people to fit the transponders and to check that they are working correctly, all of which is expensive. A promising approach being developed is the use of technology similar to that for autonomous vehicles to accurately track players or the ball in a match, or competitors in a race. If all objects can be fully digitised and tracked in all three directional dimensions and time, the possibilities for sports betting and media companies are endless. Early results have been extremely encouraging, and the technology is out on live test at the moment. This technology could open up a whole new area of betting possibilities, with a matching growth in betting revenue. In short, the next few years could prove to be the start of a global golden age of sports betting.