Bill Squadron:
Thank you, Harold, and thank you to Jim Gagliano and the officials of The Jockey Club for the opportunity to be here with you today to announce an exciting new partnership between STATS and Equibase.

This partnership had its origins in talks between Bloomberg Sports and Equibase that began over a year ago, but it really went back five years to a time that Bloomberg, which of course is a giant in the financial services industry, made a decision to launch a sports division. 

[Bloomberg] believed that the whole trend of data and analytics that was sweeping all kinds of industries would have tremendous value and opportunity in sports in the kinds of technology that Bloomberg had developed for the financial world.

I was asked to join Bloomberg to lead a team and help build that business, and it was very appealing to me because my career for the last 20 years had been at the intersection of sports, technology and media ever since I left News Corp to co-found a company called Sports Vision, where we introduced a lot of ground breaking technology such as the yellow first down line in football, the K zone in baseball, the race effects tracking system in NASCAR, things that really had an impact on the way that fans could enjoy and consume their favorite sports. 

So this appealed to me greatly.

We began the Bloomberg business about five years ago with a focus on two areas at the time.

Team analytics creating very in depth, powerful tools for professional teams to use data in a way that would improve their scouting, their professional development, their player assessment and so forth.

The other area was predictive analytics where we could use the vast group of Bloomberg mathematicians, data scientists and others to create predictive models that would have impact across the world of fantasy, soon to be daily fantasy, and of course betting. 

We built a very successful team, and in fact our fantasy projections were award winning in the areas of Major League Baseball, the National Football League; our predictive models for betting across soccer, which we'll get to in a minute, turned out to be very accurate compared to the London bookmakers. 

But the thing that our team needed to do was find the right home to grow.  Bloomberg had given us a great start, but we really needed to be part of a larger sports organization, and in September of 2014, STATS acquired Bloomberg Sports, and made it part of the industry‑leading company which 30 years ago had pioneered the modern use of data in sports media and sports information. 

Now STATS itself had been acquired just previously by a very successful private equity firm called Vista Equity Partners based in Austin, Texas.  And Vista's vision was to aggregate the leading companies around the world in data analytics with a particular technology focus.

STATS had a tremendous platform to bring all of that in.  In fact, STATS over the years had built a world‑wide sports data business that had it serving more than 700 clients, media, and others around the globe, and actually tracking and collecting data from more than 80,000 sports events every year.

What we've done since September 2014, after Bloomberg became the first part of that acquisition plan, was to acquire four more companies, all of which have extraordinary new technology to apply to sports in ways that can improve the operations both of the organizers of the sports, the teams, the players and so forth, as well as make it more engaging, compelling, and valuable for fans.

We've acquired a company called Prozone, which provides the most advanced soccer analytics for the major clubs worldwide, Manchester United, Bayern Munich, across the major leagues. 

Another company called Insights has extraordinary technology that can take very complex data, and extensive mass vines of data and turn it into normal narrative language in a matter of seconds.  So it can take a quarterly report and generate a natural language summary in seconds.  It can look at, for example, an enormous amount of data concerning a pedigree and summarize it in two paragraphs in seconds. 

We acquired a company called TBTI, which is one of the leading video and engineering companies in the sports world.  So with all of this and all this focus on technology, the next step is to apply it in new ways and do the kinds of ground breaking things that we were envisioning at Bloomberg and that we want to do now at STATS.

And we're positioned to do it in part because STATS worldwide is one of the most connected companies with governing bodies and leagues. 

In fact, STATS has a very strong relationship, or as the official data partner with almost every major governing body around the world.  One of the ones with which we work very closely is baseball.  What we do in baseball is really taking the use of data to a completely new level.  So what you're looking at now [slide] is the baseball system that virtually every baseball team in the major leagues and many in Japan are using for their scouting, player development, performance evaluation, trading decisions, drafting, et cetera. 

The system itself has all of major league data going back to before 1900, so about 120 years of data.  All minor league data, all college data, a vast amount of amateur data coming from high schools in the United States as well as from overseas.  Also a lot of data from the Caribbean and from other places where they play baseball around the world. 

It has pitch‑by‑pitch data going back 10 years, so full data on trajectory, velocity and location of every pitch.

What you're looking at here [slide] is a scatter plot of every pitch that Mike Trout has faced in the 2015 season.  The pitches are color coded for the type of pitch.  They're shape coded for the location of the pitch, and on the other side of the screen you see a scatter plot of every pitch he hit into play.

If I could mouse over it, there would be meta data associated with every one of those dots telling you not only who was pitching, but the type of pitch he hit, the result of the pitch was, where it ended up in terms of the field and of course even who the umpire was. 

So it's got every piece of data that you could possibly imagine, and what it allows the clubs to do, which you see across the top, is to filter this data.  So the thing that's important, I think that this is really at the center piece of the partnership we have with Equibase is that you have these vast volumes of data and they're growing. 

Because as we employ tracking and other technology, the core databases are just going to get bigger.  The question is, what do we do with that data?  How do we use it in ways to create value?  How do we visualize it in ways that are meaningful as opposed to just having a fire hose of data coming at you that doesn't do any good. 

So what we've done at Bloomberg and now working with STATS, we've combined our capabilities with theirs in an ability to really make sense of that data.

What the filters allow you to do is filter every single pitch to exactly where you want it as well as connect it to the video.  What you're looking at here [slide] is the fact that the system allows you to click on any one of the pitches that we just saw there, any one of those dots, and go directly to the video and create a play list showing you exactly the video associated with that data. 

Now the other area that we talked about was predictive analytics.  One of the first areas we devoted our attention to and had our mathematicians and data scientists dig into historical data on was soccer.  Of course, soccer is the largest sport worldwide and particularly European soccer.

So the system that you're looking at now [slide] depicts a system where we project the outcome and the probabilities of every major European soccer match:, Italy, Germany, France, Spain, the premier league in England, the champions league and so forth. Our system that was built by looking at three years worth of historical data, building models, back testing and so forth and it now generates projections across every match. 

The bar chart you see at the top is actually a dynamic bar chart that changes live during the match.  So we'll project, let's say a match between Manchester United and Chelsea, that before I even see Manchester United is a 38% favorite, and Chelsea has a 32% chance of winning and the chance of a draw is 30%.

As that match begins, we'll be getting data from the stadium on crosses, corner kicks, time of possession, shots on goal and so forth, and that will inform the predictive analytics system that we built and update the probabilities live as to matches going on, so that bar chart when the match begins will be blank.

It will be white, and then it will fill in as it goes along, reflecting the change in probabilities.  You can see that big drop just after the second half begins because that's where a goal happens, and that, of course, has the biggest impact on the probable outcome. 

But what this allows our customers to do is to compare our projections with their bookmakers online, which in London is a very popular pastime, and in running betting in Europe has become far and away the largest growth area. 

They're able to look at our probabilities as they change, compare them to the bookmaker odds, spot differences in value.  I can tell you, when we did our analysis by comparing our projections against the bookmakers and simulating bets in those situations where there was a material difference, you would have done very well last year. 

So we're quite confident.  We know we're not right all the time, but we're quite confident that over the scope of time the projections models our data scientists and mathematicians have been building are very good, which leads us to the main topic of the day. 

We began talks with Jim Gagliano and Hank Zeitlin some time ago to apply these capabilities in interactive applications, data visualization, predictive analytics to horse racing, and it touched a real passion of mine because going back to when I was growing up in New York I spent a lot of my high school years at Aqueduct and Belmont and Roosevelt and Yonkers. 

In fact, when I went to law school at Berkeley, I had the dream that Golden Gate would put me through law school.  I can say it got me through law school, but I can't say it put me through law school. 

But we began to talk about how we can create handicapping tools that would be more powerful, more interesting, more valuable, more compelling, more engaging, both for veteran horse players as well as for newcomers and casual fans who are all trying to draw into the sport.

I'm very pleased to say that our teams have come up with some excellent ideas for how we're going to do that.  The real focus has been customizing the tools in a way that allows our data scientists and mathematicians to build projections, to project how a race would be run, to weight different values and so forth, but to give the horse player control, because veteran horse players anywhere, they all believe they have certain theories and views that they built up over time that they believe in. 

If you create a tool that allows them to apply their views to a base that's very strong and manipulate it that way, we believe that's a key part of a tool that take things to the next level. 

So our team has been working closely with the teams at Equibase and TrackMaster, and we've begun working on prototyping the service that we envision for horse players.  We foresee the product including custom formatting, and print settings with the ability to do deep data dives that allow real-time research.

So to take vast vines of data and really dive into it, do it quickly, and do it effectively so that you're not having to click eight times to find the one thing that may or may not be relevant, but really create an interface that allows horseplayers, whether they're veteran or whether they're newcomers and whether they're looking for a simple answer or whether they really want to dive deep and compare various factors and theories, to be able to use a tool that is really a next‑generation tool. 

The thought is to allow the horseplayer to back test handicapping theories, so once those theories are applied to the tool, it will then be back tested across a whole range of races that allow the horseplayer to see how his or her theory really plays through.

They can apply those outcomes utilizing advanced search capabilities to identify the most playable races all over the country, because of course the database will be such it can search throughout the country for that day's races and apply those theories in places where they're going to be most relevant and most valuable for that horse player.

And leveraging the strength of our mathematicians and data scientists, the better will have a tool that enabled him or her to assign their own values, and across these many factors, see how the STATS projections change. 

So you can see here, as the better applies their own values and change it, it changes.  So we'll have a projection and predictive algorithm, which is based on what we believe the proper weights should be, but it will allow the better to change those weights and therefore run through our predictive algorithms and show how that changes the probabilities of the race. 

So our team at STATS is working very closely with the teams at Equibase and TrackMaster and transitioning our work from the work bench to the market, and we're excited about working with everybody in the industry.

I look forward to working with many of you in the coming months as we begin to bring out these advanced tools. 

Thank you very much.