Episode 4

Beyond Moneyball

Transcript

Speaker1: From the Loyola Quinlan School of Business. You're listening to the Q Talks podcast.

Rick Sindt: Welcome to the Q Talks podcast. I'm your host, Rick Sindt. Today I'm joined by John Dewan, the founder and CEO of Sports Info Solutions, and Professor Zac Binkley of the Sports Management Program at Quinlan to discuss the history and future of the sports analytics industry. Thank you both for joining us today.

John Dewan: Thank you very much.

Zach Binkley: Thank you.

Rick Sindt: So, Zach, I'd like to start with you. You're currently writing a textbook on sports data analytics. And so I thought you'd be in a very good position to give our listeners a brief history of the industry and its evolution.

Zach Binkley: Yeah, absolutely. Everyone kind of thinks that this analytics movement started with Billy Beane and the Moneyball story from the early 2000, and that obviously did happen. And Michael Lewis's story brought light to this movement within that industry. But using numbers within sports has been around for quite some time, even from its conception. Baseball especially, has been rooted in numbers and adding those numbers as a language to the game. So this hasn't been necessarily a new thing for pro sports and sports in general. It has adapted and evolved over recent years. And the story that Michael Lewis put out on the A's and Billy Beane did shine a light on that. So we're fortunate for that. We probably wouldn't be where we are today without that story being told. But baseball, basketball, there's numbers behind the sports and there's always been numbers behind the sports. And how we track it, how we collect those data has changed drastically over the last 30 years, really. So Moneyball wasn't the start. That was kind of the start of us talking about analytics and showing how it's being used and the negative sides of analytics as well was kind of exposing that. So analytics really isn't a new concept. It is becoming newer in the way that we collect data in a way that we find trends in the way we predict performance. Those kind of areas that we're starting to develop and go more into are kind of making it feel new. But the idea of numbers being behind sports and crunching those numbers to find insights or trends isn't necessarily a new concept. Bill James Sabermetrics. That was in the late seventies, early eighties. Mr. Dewan here will talk to you about his work as well. I think Bill James mentioned that baseball and numbers, it goes hand in hand with the language of the sport. So we do speak a language through sport with these numbers. And one kind of historical context to kind of think about is I teach a sports media class, too, and we've seen that with sports media. The technology in the current era is often the driving force behind the application of our sports. So if you look at media, the newspaper innovation, the radio, even the steam engine, the TV, social media and digital media, those have all led to new developments or new applications in sport media. The same thing can be said with analytics as we have more wearables, as we have more artificial intelligence and smart cameras and AI systems, that's unlocking new data and new ways for us to analyze performance. Historically, because of those evolutions, in the 2000, we had kind of a new era of innovators who actually started to apply advanced math and statistics to further our applications in this field. People like Ben Alomar was one who kind of created models centered around the decision maker and how everything we should be talking about or collecting is there to enhance the decision making process. Whether that's a coach making a decision on substitutions or GM making a decision on who the sign or even a player making a decision on who to close out and how to close out a defender in basketball, we ultimately get these insights and this data in the hands of our decision makers. And our decision makers have evolved as well. So not only has the technology changed and evolved, but how decision makers are using data to drive their decisions has also evolved. We've had people like Daryl Morey and Theo Epstein who have actually put these models into action and had success with them. I think Daryl Morey would want to have a little bit more success in recent years with his model, but it still rings true. Three is greater than two in basketball, so a three pointer just using common math. I said that in class. That's like the first lecture is three is greater than two. So if you can focus on making more threes, that will give you a statistical advantage. We've also had other researchers, Stephen Shea, who have innovative ways of taking data and making them more applicable to driving decisions and analyzing our sports in real time. Another innovation technology wise, that has kind of spurred us ahead, has been smart cameras. Coach Popovich with the Spurs and his staff were one of the first ones that use Sports View data, which is a smart camera system that we see in the NBA. So historically, what I've seen is as the technology advances, so do our analytics, and then so does our usage of those insights. But historically, numbers are numbers, and baseball especially has been number driven, I feel, from day one, and definitely want to hear more from Mr. Dewan's experience on this.

Rick Sindt: Yeah, John, as Zach alluded, you've been a part of this industry basically since its inception. Would you mind sharing with us a bit about your story, what it was like to be at the beginning and how things unfolded?

John Dewan: So you say at the beginning is if I'm really old and I'm a young man, I'm only 66 years old, so I'm a young man and I'm still enjoying this work that we do in analytics, in sports. So I started out as an actuary. I was insurance actuary for my first ten years of my career, and then I read this book called The Baseball Abstract by Bill James. And I was mesmerized because he was doing with baseball numbers and analytics what I was doing with insurance numbers and analyzing them and finding trends. And I was just blown away. So I literally was reading his book at a table in my house, and I got up from that kitchen table and called directory assistance and got a hold of Bill James. And he was working on a project called Project Scoresheet where we gathered our own score sheets on baseball because we couldn't get the numbers from the powers that be at that time. And we just started putting all the numbers together. So my dream was to create a complete database of baseball information and complete every play-by-play in a box score, put it into a computer database. And it just was so cool to get involved with Bill because he's the godfather of sports, analytics, baseball analytics especially, but so many people have learned from and gone on. So we're all kind of disciples of Bill James. Zack, you mentioned Daryl Morey. Daryl started his first sports job at my company, an offshoot of the work that Bill James and I started together. So that's Stats Inc., which is still doing really, really well around covering sports around the country and around the world. So Project Scoresheet, Bill and I did that for about three or four years back in the mid eighties. And then he was involved with the startup of stats, which literally started in the very bedroom office that I'm sitting in my house right now, and it's grown into a multimillion dollar entity that has done a great job. And we sold that in the late 1990s, literally around the year 2000 to Rupert Murdoch. And it became a foundation for their analytics coverage, for their broadcasts, football and baseball and other. So analytics has been something that, as you say, Zack has been around for a long, long time. And many of us were so fascinated. And so what we did in the eighties and nineties was we created products that we wanted to see, that we wanted to use ourselves. So we did a lot of publishing of books, whether that's the baseball scoreboard or the baseball profiles or the scouting report. And we tried to interest teams in analytics and there was limited interest, but it really wasn't until Moneyball came out when teams began to realize that analytics can help them. So it actually went overboard initially because teams--I think Roland Hemond, the former executive in baseball, just a really great guy, shared a story with me how over 100 scouts were let go after Moneyball came out and it was like teams started going in the wrong direction. That thinking that they can cover everything with analytics and that made their scouts. But you really need both. They need to be put together. And so that's really what has evolved in baseball. My company, Sports Info Solutions got started right before Moneyball. So it has turned into a situation where 95% of our business is involving teams, whereas in my previous company stats, 99% of our business was 19 business. So that's an interesting change in how we have worked with our clients. And we very much work together hand in hand with our team clients in multiple sports. We now not only cover baseball but also basketball and football. So yeah, analytics has evolved and you know, there's more and more in technology, whether it's radar tracking technology or camera technology, analyzing the spin out of baseball, which pitcher can impart the most spin on the baseball, and what effect does that have on the ball as it gets to the batter? So much is being done in analytics. And, you know, the interesting thing for me is I think it will never stop. I mean, there will always be more and more data. As Bill James once said, 'The amount of analytics that is available to us to review the amount of data is akin to the iceberg in the water where only a very small portion of the iceberg is above the water. And you can't see the rest of it. But there's a huge amount of that iceberg and that's the unknown future analytics that we still have to work on.' And what we're doing is we have our little ice pick and we're just looking at a very small amount of that iceberg and picking at it and doing some pretty cool analytics. I mean, there's so much that has been done over the last especially over the last 20 to 30 years, whether it's the our specialty, which is defensive analytics or some of the analytics in the team developed technology like Statcast. And so there's so much that can be done, but there will always be more to do that will always be better and better computer systems. So those fans who are interested, especially if you're young, think about and you love sports like I did when I was a kid, I loved sports, I loved computers and I love numbers. And I've had my dream come true and being able to play all three and it will never stop. So if you have any combination of those three things, just keep pursuing it. There's a way.

Rick Sindt: I love to dive a little deeper into what the applications of this data actually look like in a product sense or in how sports teams or scouts or other folks use it. Perhaps John, could you start by telling us a little bit about that beginning where you and your colleagues were making products that you wanted to see and then how it's transitioned since where perhaps the consumer is driving the development more?

John Dewan: Sure. So back then we were collecting box score data and then we started collecting play by play data, and then we started collecting pitch by pitch data. So our data was going deeper and deeper and now we're into the nineties and fantasy sports get started. And so many of us were playing fantasy baseball games and we wanted to figure out what's the best way to win? What's the best way to stock a team? And so Bill and I came up with projections. So we wanted to project what a player will do in the upcoming years based on what he's done in the past. And so we began and developed a projection system which are very common now around all the sports. But we first published that in what was at that time called the Major League Handbook. We're still publishing it as the Bill James Handbook. And in fact, the season just ended and we're in our big two week, really a lot of work to put that book out and get it out by November 1st. So we're still publishing that book after I think this will be our 31st year. But yeah, projections were really key, starting to understand what happens on batters and different counts, how important that first pitch is to get over for a strike for a pitcher. That was really, really key. We learned those things because we were tracking the data. And then as we went on with sports info solutions, we started tracking even deeper data and locations of batted balls and the velocity of every pitch and the pitch location. So two really important defensive applications came out as a result of our being able to collect that kind of in-depth data. So we developed a system called Defensive Run Saved, which gave us a way of analyzing player's defense by literally measuring, measuring where every ball was, where every player was, how far he had to range to get to the ball. How did that translate into an analytical model? So that was a really cool application of in depth batted ball data. But one thing that I like to say that I had a little hand in is today's shifting in baseball. So it was literally 15 years ago when we developed some software that allowed teams to begin to understand where to position players, and namely that there were at that time there were a handful, maybe five or six players that got shifted on a regular basis. Guys like David Ortiz or Jim Tommy. But we said, Hey, this software is showing that there should be dozens and dozens of more players that should be shifted if you look at their batted ball tendencies, especially grounders and short liners. Almost all hitters tend to pull over 70% of their batted balls. And when you're starting to get up to 75 to 80% of batted balls that are pulled, maybe you should consider. Positioning 75% of your infield to decide where the batted ball is going to be hit. So it was interesting. For about five years we provided analytics teams didn't really do much with it. And then in 2012, the Tampa Bay Rays started shifting more than any other team. They were the classic team with the low payroll and the small market disadvantage. And they showed that they could be successful, similar to Billy Beane by using analytics to help them position their players. So they started shifting more. And right at that time we provided more analytics, I made I did a presentation to about 25 of the teams in baseball and demonstrated to them that what Tampa Bay was doing does actually save runs. So my small part was nice to be a part of it, but I think the bottom line is, as in all sports, copy what the successful teams are doing. And that's what happened with shifting. So that's gives you an example of how the analytics can make a difference and demonstrate separate specks from impressions and and make a manager realize or front office realize that yes, it's okay for that one ball to go through where your infielder was vacated. If you can save nine of them, know the other nine that are going to the pull side and you can turn those nine into outs.

Rick Sindt: Zach in the classroom, how do you teach or prepare your students to go into this field or use this data?

Zach Binkley: In my teachings in this area has evolved over the last five or six years. My previous institution, I taught a class that was sports analytics, but it was for sports management majors, which is the business side of sports. But we also had coaching majors as well as exercise science and performance enhancement majors. So using data, the data kind of coming in look the same, but the output side of how you are actually going to use it was dependent upon, say, that decision maker and there were some charts that we would create that are only for the coach and some that are only for the player. And we would be cognizant and this is kind of a teaching point in my classes is no matter how strong your analysis is or how accurate it is, if you're not communicating it correctly to the people who need to hear it or need to see it, and if they can't translate that into action on the field, then your analysis was kind of worthless. Well not worthless, just not as valuable as it could have been. So we always in my classes, we always start with the decision maker who is ultimately going to see these graphs or these charts things like that, like who is actually going to use this and can they use it? I've worked and coached with players. I've trained athletes. Sometimes too much data is an issue. We have paralysis by analysis. We don't want to be throwing players all these different trends and numbers and kind of making their decision-making process on the court or on the field an issue for them because they're trying to think about too much. So that's one way that we look in the classroom is kind of not only collecting data and make sure that the data that we do collect is accurate and clean. But then what we do with that data has purpose. It has action behind it. It is specific to the decision maker, whether that's the front office coach or player. And that user focus of design is really working our players or our students this past spring, they always have to do an infographic project where they create a data visualization on something they're interested in that has some action behind it. And it wasn't just enough to break down, say, the spin rate of a pitch. That's great. That's good data to find. But how are we actually going to apply that on the field? How can we apply that in our bullpen sessions? If we're looking at swing mechanics, do we start matching up the mechanics of a certain swing and the launch angle with the pitcher's spin rate and their exit velocity or pitch velocity? So everything that we try to do in the class mirrors the industry. We want it to be action based, we want it to be communication-based. We do some analysis, of course, and a big focus in the past, when I was teaching more exercise, science students had been the data collection part. That was always the starting point. If your data collection isn't accurate, then your whole project's wrong. So we focused a lot on combine testing and students that volunteered multiple years at the NBA combine in Chicago. Students ran their own combine at Louis University, a previous institution I was at where we tested multiple teams every pre season to collect data and to go over those tests. But then when we got those data reports in, how we reported that to the player was obviously different than how we report it to the coach. So every kind of output was going in a different direction with specificity behind it. So it made it more applicable for their usage. And the more insights that we get, the more data that we're collecting, I think the more that we can create action off of that. Right now, going back to John's point on what we're seeing in analytics today, the NBA playoffs are a good example of this. It's almost like teams are trying to get out ahead of the other team's adjustments, and it's a game of adjustments now. And everyone has the same insights, everyone has the same data. They know this player is going to do X, Y and Z. And if you're coming into that with that knowledge, you're already going to have a game plan or adjustment to make to counteract those actions. If teams are doing that, then we're almost at the point now where we're going to be counteracting someone else's adjustment. So it's an adjustment off an adjustment off an adjustment. And now there's this kind of cat and mouse game of teams trying to find more insight so that they can stay ahead of the adjustments. Or you let a team come out and run their system. You take that information in and you're going to try to adjust the best you can. Some teams are adjusting better than others. Some players have a higher court IQ or field IQ that they can make adjustments in real-time. The shifts in baseball have been a tremendous illustration of these adjustments, and it seems like the players have been slower than maybe they should be on these on these shifts. I would like to see maybe some more time spent on spring training with bunts and drag bunting. Maybe that could be an adjustment off an adjustment.

John Dewan: And it's funny Zach, that you mentioned that because you know, you're absolutely right. When we look at the numbers, the hitters are not adjusting. And it's like, you know, there are situations where even a slugger, if he could bunt it down anywhere near third base, will easily get a hit. And there are situations where you don't need to hit a homer, you just want to get on base. But they haven't figured it out yet and it hasn't changed. It's been you know, bunting has been growing, going for about eight years now. Every year is increased and I mean, shifting has increased it for eight years. And bunting hasn't changed. People are just not they just want to try to hit the ball harder, try to hit it, hit a homer instead of worrying about the shift. So it's I think there are more adjustments that can be made. And I just wanted to comment on your communication. Your point about communicating, that is key. You know, as we get technical, as we start using techniques that people who are not specialists in our field don't know about, we have to find ways to communicate. And being able to communicate and explain things in a everyday fashion is vital. You know, for example, at our company, we write a team newsletter that we send out to all of our team clients and even our non team clients every month. That keeps them updated on the analytics that we're developing, what new things have we done and what analytics are we continuing to look at. And we need to write that in a way that a baseball player, you know, a former player who might be the GM or might be the assistant GM who doesn't necessarily understand how you do your numbers, but they really want to know the result. It needs to be communicated in a straightforward way. So that's something that we really, really stress. So I wanted to just double down on that because communication is vital.

Zach Binkley: Yeah, I think I think Coach Harbaugh over at the Ravens has hired just a statistical analysts to be on the sideline next to him and just kicking him probability numbers on fourth and three. And just he just wants those raw probability numbers that they're going to have a 60% chance of going for it here success rate, which has streamlined the analysis part and then made it made it digestible for him to make that decision in real time. If this coach is trying to make a decision and you got 40 seconds on the clock and you're going to start throwing numbers at them in graphs, the decision making time has already passed, so how do we communicate? The data and insights need to be streamlined and really, really focused on the end user.

John Dewan: That's exactly right. Zach. And, you know, I think what's happened is that the type of person that is now in the front office and more and more on the sidelines and in the dugouts, are people who not only understand from having played the game, but are open to including the analytics that are going on. And that will just that just gives those teams that use the analytics an advantage when your manager or your coach understands it and uses it as well as just your front office, that really gives them those teams an advantage.

Zach Binkley: And I can see the counter to that as well is we've done some projects in the past in my labs. We did a in basketball, we did a pass assist analysis chart. And one of my students, Phil Palm, who does analytics for the women's basketball team at Marquette, was the one who created this. We basically took a five game period during the season and tracked and highlighted where every pass on the court was starting and where it ended up. And then the result of it, and this was in 2015. So we really had to do some interesting visualization work then. But we created this report and we got it all done and we handed it to the coaches and the players. And the feedback was, We can't use this. We don't know how to read this. This is great research, it's a great analysis, but to us this is unusable because we don't see how it can be used. It was just giving us information, and that's not necessarily what's needed in sports analytics. We have a ton of new graphs that are coming out and all these charts that really have no end use in the player's hands. So in the classroom, that's kind of the first teaching point is working backwards off of what players and coaches need. The GMs are a little different level. They can go more in the business side of analytics, business intelligence. But if you're going to work with players and athletes as well as coaches, speaking that language, playing the game helps, being around the game helps. And the idea, I mean, the chart we made was, was fantastic. It looked great and we understood the breakdown of it. But its application use was just not there.

John Dewan: I just want to tell a personal story in that regard. You know, baseball players, football players, basketball players, they have raw talents. They have the best possible talent that there is. And sometimes why they are successful is not something that they can think about. It's not something that they can pinpoint. It's instinctual and it's part of their athletic ability. And from a personal standpoint, I haven't been the greatest of athletes. I totally love sports. And I play I played every one of them. And the only one I got pretty good at was bowling. Back in the day, I was on the Loyola Varsity bowling team four years in a row, and so I enjoyed that. But as I got into the analytics, I actually started collecting some bowling analytics and tried to apply it to my game. It totally did not work. I could not. I had to stop looking at the analytics because it was ruining my game as opposed to making it better. So that was a real lesson for me to realize that there are some situations where, depending on who the person is, the analytics may or may not help them personally. So it has to be all the teams have to work with their athletes individually and work with them in a way that works for them. And sometimes working with the analytics individually with a player might not work.

Zach Binkley: Sport performance, and I think this is kind of what you're talking about is we have this underlying science that we have a tough time measuring in real time, and that's motor control and motor learning and we're starting to get more technologies to do that. But how you acquire a skill, how you develop your jump shot, your swing mechanics, learning those processes and then kind of adjusting with the numbers. And in that could be a revolutionary way of training our young athletes because like what you're saying is absolutely true is the athletes are just so natural and the movements that they practice over and over again and decisions that they make on the court or on the field are natural to them. And that natural sense or that feeling comes from training. And that training is what more control, more learning is doing. So I always thought about teaching analytics in terms of a on core or on field skill set needs to be started at the earliest age as possible because it is cognitive, it is decision making, trying to adapt the brain of a pro athlete at age 30 in this regard could be a challenge.

Rick Sindt: I like how this conversation has encompassed like you were talking about before. John, how we're just at the tip of the iceberg. There's so many more possibilities ahead of us. And we're now we've taken into account just the constraints of application. And how do you work with people in different positions if you're an athlete or or front of house or a general manager. So I'm curious to to wrap up this conversation where you both see this industry going in the future. How do you see it developing? How do you believe it will adapt and respond to this wealth of knowledge that we're gaining access to and the just practical application needs of the user?

John Dewan: So I'll take the first stab at it. Technology is just vital. Back in the eighties and nineties and even in the early 2000s, you know, we were investing hundreds of thousands of dollars on technology in order to provide cutting-edge research and have data that never existed before. And back then, that's a lot of money. But now leagues are investing literally billions of dollars in collecting information, creating tools to collect information. So the technology and the information that it provides will continue to grow and it will grow faster than we can keep up with it. It has, in fact, I mean, in my entire career of 30 plus years, the data has grown faster than our ability to keep up with it, and that will continue. So as I said, the tip of the iceberg. The iceberg is actually getting bigger and our tools are not getting much, you know, are getting a little better to work with it, but it's getting bigger and there's more and more data. And it just will I think it will continue. I think that performance measuring performance using these tools will allow athletes to do more and more things and get better and better.

Zach Binkley: Yeah, I think with the way technology is moving, I make a point in my class I just made it yesterday in sport media is things like Moore's Law that's not slowing down and we're not going to be able to slow that down. What we need to do is think critically on the new data that we're finding and realize if that's usable or not. And too much data is an issue. Too little data is an issue. Data that you collect that isn't usable or actionable is an issue. There's some data that it's just extra static noise. So really being specific in our data collection is kind of one way I think the future will go, and I think we'll do that with more advances in technology. I was, and still am to a point, in favor of wearable technology. And wearable technology has hit a bump the last couple of years. I think we have the concepts down, the technology's there. We're working on the validity and reliability of it. But that's hard to do because every human body is different and you're making a wearable tech that is a one size fits all. I know we do have some gender issues related to wearable tech and how there isn't a lot of wearable tech being created specifically for women athletes. So wearable tech, the issue there is we could be doing more with that. We should be doing more with it in the future. But you also have smart cameras that are going to start to kind of chip away at that market. You also have things like force plates that can collect data and give insights to the force production of athletes. So I think we have an idea of what high performance is in sports, and we know kind of how we get to those high-performance areas. We know we know LeBron is successful because he's LeBron, and that's a good model to look at. But how can we create more Lebrons? How can we train the athletes to maximize their potential? Not everyone's going to be LeBron, obviously, but with these new insights, I'm really focused and really looking at the long term health of athletes, long term athletic development, their injury prevention, their maximizing, their peak and their bell curve, extending their bell curve in terms of performance, I think when we start looking at that and using numbers to enhance the human, I think that's where we're going to have another kind of boom or explosion within sports analytics. Sports is a human-based activity, right? We do have E Sports. I've seen there's like robotic sports leagues out there. We want to keep the human-centered around or at the center of sports. So I think we're going to start seeing the shift of analytics, taking a more specific approach to the individual to enhance their human ability to perform at a higher level.

John Dewan: Zach, you mentioned injuries, and I think that's that is a very fertile ground for more and more analytics because as extreme performance takes place, you're causing more stress on the human body. And so we've actually expanded our databases to collect the most in-depth injury data that we can possibly come up with. So I think that's a very fertile ground.

Zach Binkley: And that's kind of my motivation and driving force, if I could. I grew up as a coach's son, played two sports. If I could go back and collect some workload data on myself in my youth career, I would love to see what that data looks like because.

John Dewan: Maybe you'd still be playing then maybe you'd be on the Lakers with LeBron.

Zach Binkley: That's exactly what I was thinking. My arm was done, I think, by junior year, but I had that nasty curve, you know, in third grade. So and I think I was on a two-man rotation during our middle school baseball year. So I'm sure that caught up to me. And now there's organizations like the NCAA National Strength Training and Conditioning Association. I put on a clinic two years ago on long-term athletic development and how we can start to track and collect data on these athletes in hopes of prolonging their careers or eliminating injuries or at least limiting them. So that's good for everyone. That's good for the player, that's good for business, that's good for the sport itself. We want people playing sports into their later years. So yeah.

Rick Sindt: John, Zach, this was a wonderful conversation with a lot of insight into the industry. Thank you for joining us today.

John Dewan: It was a real pleasure. I enjoyed meeting you both, Rick and Zach. Thank you for having me. Thank you.

Rick Sindt: This has been an episode of the Q Talks podcast where we seek to marry the wisdom of the Quinlan community with the issues of today. Special thanks to our guests for their conversation and insight with additional thanks to Dean Kevin Stevens for his generous support of this project. Matt Shelley, our student producer for editing this episode, Loyola School of Communication and WLUW for their continued collaboration. Please take a minute to support us by rating and reviewing our episodes to help expand our reach. Thank you for listening and we hope you join us next time.