How racing is clearing the in-play hurdle

“Collecting in-running data from horse racing has always been more challenging than other sports but recent innovations mean a viable product will soon be on the market” Our Head of Trading Adam Conway recently wrote in G3 Magazine. Below you can read the entire piece.


The advent of in-play betting has led to increased revenues for bookmakers from many sports – but horse racing has not been among them yet. There are many reasons for this, but a principal one is the difficulty of creating an accurate, yet cost-effective, real-time data collection mechanism. When it comes to football or tennis, you can collect data from streamed pictures or with the help of in-stadium scouting solutions. Racing, however, is a lot more complicated. It takes place on a huge variety of courses, rather than a standard pitch, and there are too many things happening at once to be accurately assessed by a single pair of eyes.

Nevertheless, when you look at the wide range of in-play markets driven by football and tennis, it’s easy to see why bookmakers would be keen to entice new customers and increase revenue by broadening the scope of racing bets they offer their clientele. There have been many attempts to solve the problem over the years but they have fallen at the twin hurdles of prohibitive cost and inaccurate data.

SIS, with its heritage in horse racing content delivery and desire to innovate, has been working for the last two years on developing a solution. We’ve tried to balance the twin aims of collecting robust, accurate data without incurring huge capital and operational expense. A GPS tracker in a horse’s saddlecloth, in conjunction with on-course mapping, produces streams of accurate data showing a horse’s location and direction without the need for prohibitive levels of investment. This data, filtered through a complex algorithm and supervised by an expert trader, can be processed into accurate real-time fixed-odds prices.

Creating an algorithm for a horse race is a hugely complicated process, and it has taken a while. We’ve had the help of skilled mathematicians and a large historical data set to develop it to a high level of accuracy. This is a process which will never cease, given the need to hone algorithms constantly by adding more and more fresh data.

However, even in this era of automation and big data which helps us to account for factors such as each runner’s form and the history of races at a particular track, the importance of manual intervention cannot be overstated. There will always be a need for an expert eye which can take into account visual factors such as whether or not a horse has come off the bridle, which will not always be reflected in positional data, and adjust the odds accordingly. We expect the ratio between automation and manual input will be 90:10. The team of expert traders I lead at SIS has been shown to improve margin and reduce the impact of outlying information when using the technology.

We think there will be an appetite for in-play betting from existing horse racing customers, but we also expect in-play customers from other sports to move a share of their wallet to racing. Perhaps the major advantage of data-led pricing systems is the range of markets that can be provided to an operator. When there is sufficient data available, a punter should be able to back the winning distance or finishing position of a horse while the race is on, as opposed to the limited current offering of win and each-way markets. We have seen operators experiment with labour intensive, inaccurate live trading initiatives. Across the industry, there has been little progress with these products.

At SIS, we see the solution as an easily-managed, outsourced pricing feed which offers accurate prices in less than a second to our partners. This in turn will free up operator resources to concentrate their expertise where it is most needed.

For more information about our Horses In-Running Pricing click here