Subscribe: Google Podcasts | RSS
Subscribe: Google Podcasts | RSS
Have you ever thought, “you know, it would be interesting to take my analytical knowledge and just totally run an organization based on what the data says?” Yeah. Us, either. That’s terrifying! But, that’s exactly what our guest on this episode did. Ben Lindbergh, along with his stathead-in-crime (aka, co-author) Sam Miller, took over the management of a minor league baseball team in 2015, and the result was The Only Rule Is It Has to Work: Our Wild Experiment Building a New Kind of Baseball Team. How does that apply to analytics in the business world? In a surprising number of ways, it turns out!
0:00:05.9 Announcer: Welcome to the Analytics Power Hour. Analytics topics covered conversationally and sometimes with explicit language. Here are your hosts, Moe, Michael and Tim.
0:00:22.0 Michael Helbling: Hi, everyone. It’s the Analytics Power Hour Episode 173. I know, I’m still testing the model. Most of us walk into a new job or a new project, we’ve got a lot of excitement and belief in our ability to use data to have a really dramatic impact, and we’re looking for the ability to drive that impact and transformation within decision-making, within our team or department, or across the entire organization. Unfortunately, way more common, the experience is something around starting off with some really brilliant ideas of where we think it’s gonna go and sometimes falling shorter of those that we really hoped. But in all of that stat and some success and a lot of failure, there’s a lot of learning that actually might be a really cool story. Moe, are you a sports fan at all?
0:01:17.3 Moe Kiss: I am a sports fan, except of course, down in Australia, I follow AFL, Australian Football League, as my primary sport.
0:01:23.7 MH: Oh yeah. The only thing I really know about that is the referees and how they do that thing with their fingers.
0:01:32.1 MK: Oh yeah, yeah, for the goal.
0:01:32.7 MH: Yeah, for the goal. That’s all I know about it. What about you, Tim? What’d you say is your sport?
0:01:42.1 Tim Wilson: See, my sport is probably college baseball, just to be weird and esoteric, and then I casually follow whatever else is on just enough to not look like a complete goofball in a bar.
0:01:54.3 MH: Yeah. And I’m Michael and I only watch the playoffs in most leagues and/or the Cleveland Browns.
0:02:02.5 TW: I was gonna say, that was gonna be the Venn Diagram of the Cleveland Browns and only watching the playoffs was gonna be a limited number of games.
0:02:12.1 MH: But we recently had the opportunity to talk to someone who had done a really good job of writing that story down, and he’s our guest today. Ben Lindbergh is a staff writer for The Ringer. He also hosts The Effectively Wild podcast for FanGraphs, and regularly appears on the MLB Network. He’s a former staff writer for FiveThirtyEight and Grantland, former Editor in Chief of the Baseball Prospectus and The New York Times bestselling co-author of ‘The Only Rule Is It Has to Work, Our Wild Experiment Building a New Kind of Baseball Team’, and ‘The MVP Machine, How Baseball’s New Nonconformists Are Using Data to Build Better Players’. He lives in New York City, and I took that bio from the MIT Sloan Sports Analytics Conference speakers page. So this guy is pretty top notch. Ben, welcome to the podcast.
0:03:01.7 Ben Lindbergh: Hi, thank you for having me, and sorry that baseball books have such long subtitles, I don’t know if that was unethical.
0:03:08.9 MH: Until you get to that Money Ball level of just the one word that gets made into a movie, right, it’s the holy grail for all of us out there.
0:03:17.2 BL: Yes, as long as Michael Lewis is on the cover, I think it probably sells best, Ben Lindbergh, you need a long subtitle.
0:03:23.3 MK: I did love hearing, though, that if this book gets made into a movie, there’s gonna be a montage of Excel spreadsheets, I think that really like sung to me.
0:03:33.3 BL: Yes, and that’s probably why there hasn’t been a movie so far.
0:03:37.5 MH: Let’s not sell it short. ‘The Only Rule Is It Has To Work’ was a New York Times best seller, it’s a phenomenal book, and that’s the one we’ll primarily talk about today. But maybe Ben, we should start closer to the beginning of how did this intersection of data nerd/writer/baseball fanatic kind of come into the world and what puts you in a position to start in on a project like the one you documented in the book?
0:04:05.7 BL: Well, it was an improbable path because I was an English major, and so I had to learn a lot as I went, but when you write about baseball, you find that it helps to have some statistical knowledge and hopefully some statistical skills, because there are just so many numbers in baseball dating back centuries, decades and more and more comprehensive as time has gone on. And so gradually I gravitated toward this more analytical way of looking at the sport, what’s known as Saber Metrics, which is essentially the objective search for knowledge about baseball or the search for objective knowledge about baseball, and that can take many forms, but often it takes statistical forms.
0:04:45.6 BL: And so I went to this website called Baseball Prospectus, as you mentioned, and it’s one of the premier stathead sites, the number crunchers, the people who are turning their analytical lens on the sport, and over time it became appealing to me to try to put that into a more practical application. It was me and a bunch of other numbers, nerds just sort of sitting behind our keyboards and passing judgment on transactions and saying the team should do this and team should do that, and granted as time has gone on, teams have hired from places like Baseball Prospectus and other similar sites and have kind of adopted that mindset into their operations for better or worse, but back in the summer of 2015, my co-editor and co-author on The Only Rule, Sam Miller and I got the opportunity to take over an actual professional baseball team and make the decisions for that team throughout that season, it’s a team called the Sonoma Stompers, which plays in a league called the Pacific Association out in Sonoma, California, and this is a lower level league, but it is a professional baseball league. These are players who are paid to play baseball and who have aspirations and dreams to climb the ladder. These teams in this league are not affiliated with major league organizations, but players do come and go between major league organizations and these independent leagues, this network of Indy Ball that exist throughout the United States and Canada.
0:06:18.1 BL: So we partnered with this team that said, “Hey, we don’t have a lot of resources, we’re always interested in a little publicity, why don’t you come in and take over our baseball operations and, in theory, at least, be responsible for signing players, deploying players, deciding on in-game strategy, the whole works. And we were, of course, terrified, but also intrigued by the idea of actually getting to test our merit in a real life situation and see whether these things that we thought would work would actually work when put into practice with real people and not just those numbers on the spreadsheet.
0:06:55.9 TW: It’s like the parallel of the analyst saying, “I’ve got all this data, if only, if only… ” And then some small company says, “Sure, come on over, you are now the CEO. Knock yourself out smart guy.”
0:07:08.9 BL: Yes.
0:07:09.0 TW: That would be terrifying.
0:07:10.1 BL: Yeah, it was really exciting for a day or so, and then it became quickly overwhelming. It’s like, “Hey, we’re gonna write a book about this. We got a book deal. This is great. Oh, now what do we do? Oh right, we have to actually put this team together and we don’t know what we’re talking about, or we hope we do, but we haven’t actually put it into practice yet.” So yes, we celebrated briefly and then panicked for a while and then gradually got our acts together.
0:07:38.2 MK: It’s funny you actually… You and Sam both kinda touch on the concept of impostor syndrome, and the nerves going in to basically convince these players to listen to the data nerds, which might be something that’s not familiar to them. What do you think finally made you, I guess, more confident? Do you think it is just about getting wins on the board, or do you think that this idea of imposter syndrome is ever present and there’s nothing we can do about it? How did you find your faith?
0:08:08.1 BL: Well, I think we took time, of course, to prove ourselves to others and also to ourselves, really, and we showed up at this team and the players had not… In all cases, there were a few hold-overs from the previous team, we also recruited and signed some players, but I think even the players that we recruited were not aware of exactly what was happening here and how unorthodox this would be, and so when we showed up and everyone else showed up and we sat them down in spring training and explained what we were trying to do here, I think there was understandably a lot of skepticism. And our names meant nothing to these players and Baseball Prospectus, which is a pretty respected name in professional baseball circles also meant nothing to these players who in many cases, were just out of college or recently out of college.
0:09:00.9 BL: And so we really had to justify our presence there and we felt like fish out of water, we felt out of depth, we were out of our depth, we were not the jerks in school in our own histories, and so we were kinda crossing over, crossing clicks between the geeks and the jerks here and so we felt uncomfortable at first. And I think initially, we hoped that some of the resources that we were able to bring to bear would get us accepted to some extent, because at this lower level of professional baseball, there just aren’t a lot of resources, there aren’t a lot of statistical tools and technology, and people who are available to run studies or provide analysis. And that’s one reason why the Stompers were interested in bringing a support because their general manager also doubled as the person who would go around and collect the concessions money in the hot dog stands.
0:09:56.6 BL: So we’re not talking a big organization here. And so we were able to come in and provide these players with some tools that really they had not expected, they’d never been able to use before, things like swing tracking devices that you attach to the end of your bat and you’re able to get some read outs about your swing or the camera system called PITCHf/x, which at that time was in place in major league parks and had been for some years, but was unheard of in the Pacific Association. So things like that or even the human scouting network we set up with video cameras and people in the stands charting various events on laptops and putting these things together and being able to show the players, “Hey, here was your bat yesterday. You can look back at it.” This was not something that they had been able to avail themselves of before. So I think at first we said, “Hey, look at all our cool toys and tricks and these things that you can use now, because we’re here and we hoped that that would kinda get us in the door, but ultimately, I think it helped to get to know everyone so that we were not just the strangers, the statheads who were parachuting into the situation where typically we would not be.
0:11:12.6 BL: But we were people, we were Ben and Sam and hopefully over time, the players came to accept that we knew something about baseball too, even if it wasn’t quite from the same perspective. And we hoped that it would be sort of an exchange program that we may have some things to teach them, but that they might have some things to teach us too.
0:11:32.1 TW: That’s interesting on the highlights that whole… The communication or the relationship building piece, and I’m curious, as you’re kind of tapped into the broader world of MLB and other minor league systems, are there cases where there are the brilliant statheads who come in who can’t… I think it was in The MVP Machine that you talk about the idea of a conduit where you need somebody who can actually work with the data person and has credibility with both parties to bring that together like do you know…
0:12:05.7 MK: Some kind of data translator, Tim? [laughter]
0:12:05.9 MH: Oh, Interesting.
0:12:10.3 TW: No. I quit this podcast.
0:12:17.0 TW: But have there been cases where you’ve seen the data person just can’t actually manage to… They come in with the, “I’ve got the spreadsheets or I’ve got the R script, and therefore you must listen to me” and just never can get traction and they flame out, or do they eventually generally figure that out, or organizations, like the Major League level or getting to where they’re embracing the data so much that there’s kinda built-in credibility for those?
0:12:46.4 BL: Yeah, at this point, it’s pretty pervasive in 2021 because we’re now almost 20 years past Money Ball and this of players has grown up just steeped in the stats and the technology, and often they’re exposed to it, not just when they get to the big league team, but throughout the minors or even in college or high school, there are cameras that are tracking their pitches or other things that can tell them how fast they’re throwing and what the spin rate on their pitches is and all of this other data. So now when they’re getting to that highest level, it’s no longer that kind of culture shock, but certainly, early on when it was just a few nerds in the wilderness there, they tended to be ignored, overlooked, shunned, there was even some territorial nature, some kind of squabbling over, well, it’s stats versus scouts and the statheads are coming in and they’re threatening what used to be the scout territory, so there was kind of this conflict at that time.
0:13:47.0 BL: And I think it’s largely subsided now, if only because I think that battle is over, the philosophical conflict has largely been resolved and the stats won to a great extent, I think, although really, they’ve learned to live together, I think. And the smartest and most progressive organizations have realized that you can use both of these perspectives and that in fact, these perspectives are converging now because now you have this technology that tracks players’ motions and tracks the ball and tracks the bat, and it’s providing data and stats and numbers, but really it’s the same sort of information that scouts used to give you or still give you with their eyes. And so it’s not really two different schools of thought, it’s just two different ways of measuring essentially the same thing, so I think that is largely resolved. But at least initially, it was really helpful, as you said, to have the conduits as we call them, which in many cases were former professional players who could cross these boundaries because they were immediately accepted in the clubhouse in a way that I wasn’t and that Sam wasn’t.
0:14:58.2 BL: They just fit in, they had walked in those players shoes, they had worn the uniform, they knew what that life was like, and so they had that instant credibility, but they were certain players who had an interest in the statistical side of things too, and so they could speak the front office language fluently as well as getting along with the players, so that was pretty important. And now you even see a lot of teams will employ statheads who will travel with the team and they’ll just be in the clubhouse disseminating that information directly, and that’s a much more accepted part of the game, even then it was, say, six years ago when Sam and I were trying to be the pioneers in that area in the Pacific Association.
0:15:41.8 MK: And it’s funny, you talk a lot about, I guess, the realization that data is not some silver bullet that gives you all the answers, that it’s part of the information that you have, I guess, when you’re making these decisions, and ultimately as analysts, we’re pretty aware, you can tell any story you want, if you cherry-pick the right data and put it in the right format, so to speak, looking back now, is there anything you would do differently? Do you feel like you realize that straight away, or did that take some time, I suppose, to come to that conclusion that data is just one of these things, and there’s all these various inputs that we can use to make decisions and that it’s part of the story not the whole story?
0:16:29.6 BL: I think we knew that on some level coming in, but I’m not sure that we had totally internalized it until we were living it, because one thing that was jarring to us is that when we showed up in Sonoma, there was really no history of anyone in equivalent positions who was trying to capture this kind of information, and so we were almost starting from scratch, whereas Sam and I had been spoiled as analysts of Major League Baseball, where all of this information is captured automatically and provided to the public, and if you have the requisite skills, you can look it up, and it’s all right there, and you just have to figure out which questions to ask and what’s the best way to answer them is. When we showed up in Sonoma, there was nothing, essentially. There were some pieces of paper that people from previous years had tracked things on by hand, but it wasn’t entered into any database and wasn’t really all that rigorous.
0:17:27.2 BL: And so before we had to make these decisions that we had hoped to make, we had to actually gather the data in a really hands-on way, whether it was just going out to the outfield and making sure that the video camera was pointing at the right place or just making sure that the systems were calibrated, making sure that everything that could go wrong was not going wrong, and it so often did until we kind of got our process and our routine down.
0:17:52.0 BL: So there was definitely an adjustment period where we realized that we know what we wanna do with the information, but the information doesn’t actually exist yet, so we have to wait a while before we have a sufficient sample to be able to say anything, and I think that was something that held us back at first because our inclination was, “Well, we have to wait to have enough information where we can make these informed decisions.” And also we wanted to be diplomatic about everything, and we didn’t want to be tyrants and suggest that we knew everything about everything already and come in and dictate terms and say, “This is what you’re doing and that’s what you’re doing. We thought, “No, we’ll come in and we’ll use a lighter touch and we will build a consensus and we’ll all be on the same page,” and in theory, that sounded great, but what we found is that by not seizing control, by not acting quickly and sort of establishing our authority, we essentially… You’ve ceded some of that authority to others, to the people who would traditionally have it.
0:18:56.1 BL: So the manager, the coaches, the veteran people on the team, they just sort of stepped up and said, “Well, this is what we’re doing,” and I think because we were cautious and because we were not entirely comfortable and because we felt like we didn’t have enough information to really provide these people with better decisions or better information to make decisions, we deferred at first, and I think that was one of our great regrets because that sort of set the tone and that sort of set the pattern. And it was something that Sam and I sort of squabbled about too, how do we wanna handle this, do we wanna come in on day one and march up to the white board in the clubhouse and say, “Here’s the lineup for today,” which traditionally in baseball would be the manager’s job? But we thought, “Well, do we wanna come in and just assert our authority by saying, ‘No, this is our job now,’” and we figured, no, we won’t go full-tilt from the start, we’ll ease into it. And then what we found is that, of course, everyone took that authority that we had not seized, and so we had to seize it later on in the season, and then it became a greater source of conflict and dissension.
0:20:02.3 BL: And that was probably good for the book because drama creates conflict and it led to a better story, but in the moment, it was more difficult. So I think in retrospect, even if we hadn’t had the information that we wanted, we almost should have faked it, ’til we could make it, I think.
0:20:25.8 MH: Let’s step aside for a second and talk about something that happens to everybody working with data analysis within their jobs, and that’s poor data quality impacting the results and outcomes of your data. Tim, have you ever experienced this?
0:20:42.6 TW: Never when I’m dreaming, but almost every single time when I’m awake and sentient.
0:20:51.0 MH: Yeah, that’s absolutely right. So we partnered with Observe Point because that’s what they do. They give data professionals confidence in their data and insights, and they do it through a number of different things that they offer with their products, so basically automatically auditing data collection for errors across your entire website.
0:21:12.3 MK: And don’t forget that they alert you immediately when something goes wrong, which I’m obviously a big fan of.
0:21:17.9 TW: That it could wake me up from one of those dreams where I’m dreaming of pristine data and then I get alerted and I wake up and then there’s a problem that needs to be addressed, and so then it’s just a minor blip. I think that’s actually one of the big benefits, is that rather than finding out that you are missing two weeks of data, you’re getting alerted and you can correct it, and you have a very small window with an issue.
0:21:40.1 MH: But actually, if you’ve heard of Observe Point before but you’ve never actually taken a closer look at their tool, you should go request a demo. You can actually go to observepoint.com/analyticspowerhour to learn more about Observe Point’s full data governance capabilities. We think it’s probably well worth your time. Alright, let’s get back to our conversation.
0:22:04.4 MH: There are so many parallels in terms of the work that most of our listeners do in the field of analytics, of you come into a new organization or a new place and it doesn’t have the right tools or doesn’t have the same kind of data.
0:22:18.5 BL: Yeah.
0:22:19.2 MH: Now you used the example in the book of sort of like, okay, what’s the average lead-off length, which is just a metric you can grab out of any major league park, easy peasy, and like, no, we’ve never heard of this metric at this level, and you’re just like, “Well, how am I supposed to work?” And I think that’s something you handled it so creatively and opportunistically, you’re like, well, let’s go find a way to get some of this data in here, but then using that data and having the confidence to walk in the room, and… I think you’re right, it did help the book because I started reading the book and I was like, “Oh, that’s cool,” about stats and whatever, and I was like, “This is not about statistics at all.”
0:22:55.6 BL: No, absolutely. Yeah.
0:22:57.0 MH: This book is about humans trying to find a way together, and it went different ways, and I think the drama of that actually really transcended the math and the numbers in a lot of ways, and I think probably made it a much better… I really enjoyed reading it. And…
0:23:15.4 BL: Thank you.
0:23:17.5 MH: Again, of the three of us, I’m the team, like dynamic, so all my notes are like, oh cool, the corporate ice breakers work at this level, two truths and a lie. That’s perfect, I’m writing that down. Never try to win an argument. I love that quote.
0:23:34.0 MH: All of those are the ones I was taking notes about, but it was… That was what was really sticking out to me, and I thought it was really interesting, ’cause all of us walk into corporate meetings where we’ve gotta go talk to a marketer or an executive. We don’t feel that confidence, and we don’t know what to tell them to do and they want… Sometimes they really want you to, and sometimes they’re like, “Why don’t you shut up and give me the data I ask for. Leave me alone.” So it was just so many parallels were flying out of this book, and so it was really delightful.
0:24:03.2 BL: Yeah, I’m glad you said that because the book turned out to be a bit different from our proposal, and the proposal was somewhat speculative because, of course, it hadn’t happened yet, and so we said, “Well, we’re gonna try to take over this team and run it, and we’ll see how it goes.” And we had some ideas of what the themes might be and what the conclusions and takeaways would be, but of course, we couldn’t predict all of that, and we knew that part of it was going to be those interpersonal dynamics and how do we persuade people to listen to us, but I think as you said, that became a bigger part of the book that we had anticipated, whereas we had envisioned it as, “Okay, stat nerds come in and they run a baseball team in the mathematically optimized way, and then we find out what works in baseball and what actually are the most effective strategies.”
0:24:50.1 BL: And there’s some of that in the book to be sure, but I think if that had been it, then it probably would have appealed to a smaller audience, I probably would not be talking to you right now because it would not have crossed over from the niche hardcore baseball audience. And also, I don’t think it would have held up as well in the long term, because some of the specific tactics and strategies that we were talking about, some of them have been adopted in subsequent years now to the point where if you read the book, it might not seem so ground-breaking, what we were trying to do, but I think those challenges that you encounter in any new situation like that where you’re trying to create some sort of change and there’s a data element to it, I think a lot of that is universal and hopefully, we’ll be closer to timeless, so I’m actually glad that it turned out that way, although it led to a lot of stressful days and nights in the moment.
0:25:48.4 TW: Well, that’s the other… The walking in from accustomed to MLB and thinking, “Well, yeah, it’s gonna be a little bit less”, and finding how much less it was. I feel like that’s another thing that analysts, we sometimes run into, where the pieces that get published or written or taught kind of assume this vast amount of data, one which you can build models and split into test and training, and you have this rich data set, and the reality is the small and medium-sized businesses are sometimes saying, “Can I count how much traffic came to my website? You want me to do more? I don’t have a full-time analyst”, just like the DM collecting the concessions, you don’t have an army of analysts who are doing this full-time…
0:26:36.7 BL: Right.
0:26:36.8 TW: You guys were wearing multiple hats as well, and I feel like the analytics world can fall into that trap as well. Like, “We’ll just do it the same way eBay did it.” And it’s like, “Well, yeah, but nobody’s eBay.” You define solutions that work for the equivalent of the major league ball clubs, and that is a tiny fraction of the organizations that are out there…
0:26:58.5 BL: Yes.
0:27:00.4 TW: And realizing how can you apply what you can? How can you get by?
0:27:04.9 BL: Yeah, and also it was a single season that we had here and a shorter season than the major league baseball season, and so essentially, the whole thing was a smaller sample than we would have liked. If we could have run these trials over, multiple seasons over a six-month season, then we would have been able to feel more assured about certain decisions, but we worked with what we had and we concluded what we could. And if we had done it over the next year and the next year, then we probably would have done it a lot better, although again, it probably wouldn’t have been as good a book, I think.
0:27:38.9 BL: And as you said, you are confronted with the realities of the situation that when you’re sitting at home and you’re just staring at your computer screen, it all seems so clean and orderly and efficient. And then you emerge into the harsh light of this actual situation, and you find that even if you have some smart idea about a strategy that you could put into practice, maybe you position your defenders in this spot instead of that spot because you’ve looked at the numbers, and you know that the opposing batters tend to hit the ball here and not there, that doesn’t work as well as it should on paper if you can’t convince the players that it’s actually a good idea. And if the players are dragging their feet and they’re resisting and they don’t wanna do this and they think it’s weird, and they think you don’t know what you’re doing, maybe they outright revolt and refuse to put it into practice, in which case you’re not getting any value from your brilliant idea because it’s not going any further than in your head.
0:28:38.2 BL: Or maybe they implement it half-heartedly, they say, “Alright, we’ll go along with this, but we’re not gonna really put our full effort in, ’cause we don’t believe in this, we don’t think it’s a great idea to begin with,” and then maybe the numbers you get are sort of skewed and distorted by that. Maybe it looks like actually this was a bad idea, but really, it’s just that it wasn’t implemented properly and you didn’t have everyone buying into it. So that became a big part of it for us too, where we could look at the numbers and we could craft these strategies, but we also had to figure out how to get the players on our side and hopefully even excited about these things so that they would be our partners in this, as opposed to just our unwilling lab rats, essentially, which is not what we wanted them to be.
0:29:25.6 TW: And you talked about how, to a person, every single player on the team, their number one goal was for them to personally move up, to ultimately get into an affiliated league, so their motivation was they’re looking out for themselves, which they want to be a career and presumably they kind of knew… You had a book deal and so part of that, you wanted the team to do well, but it’s like in a sense I could see there being sort of a tension between looking at you and Sam and saying, “These guys are coming in, they’re playing their little things, they’re trying this stuff out, but this is my career. So if you put on the shift and that means I’m not there… I maybe don’t get an error, but I don’t get a chance to make a play, ’cause your crazy idea… ”
0:30:12.7 TW: “No harm on you, but my stats aren’t gonna be quite as good.” So it was kind of conflicting motivations that they really had to trust that what you guys were coaching and telling them to do was actually gonna make them more likely to move up. Right?
0:30:34.3 BL: Right. Yes, and these players were not impressed by our book deal.
0:30:41.7 BL: If you’re talking to another journalist and you say, “Hey, I got a book deal.” “Oh hey, congrats. That’s pretty impressive.” If you’re talking to a 20-something athlete and you say, “Hey, I got a book deal… ”
0:30:51.1 TW: A book?
0:30:52.7 BL: Right. Yeah, I don’t know whether they actually believed that this would be real or that anyone would read it. I think it was very abstract to them, even though we obviously warned them coming in that this was happening and got everyone on board with it, but I think it was very remote to them, they were worried about their own performances, they should have been. And one reason why we wanted to implement it at this level… Well, a, it was really our only option because some major league team was not gonna hand over the reins to me and Sam, but also at this level, it’s far enough away from the majors that we thought, well, these players will be more receptive to us than higher level players would be.
0:31:30.4 BL: Because if you’re at the doorstep of the major leagues, let’s say, and then suddenly someone comes in and says, “Do everything differently, change everything you’ve been doing to get you to this point,” then you’re gonna be resistant to that because you feel like what you’re doing is working, whereas these players were far enough away, and I think even they realized that the odds were long enough that they would be at least somewhat more receptive to changing things, not that they didn’t take this seriously and want it to be their career and believe in themselves to some extent, but they knew that there were many, many other players vying for spots who were ahead of them, so I think they were more likely to take a risk. And also, this is not a major media market where you have columnists and TV commentators who were gonna be criticizing what I’m doing or what Sam is doing or questioning these strange strategies, we’re sort of in the baseball backwater here and not in the spotlight. So I think that gave us a lot of freedom, which is something that Billy Beane and the GM of the Oakland A’s in Moneyball, he talks about too, that it might be difficult to do something if you’re running the Yankees that you can do if you’re running the Oakland A’s, and there’s a little less pressure and scrutiny on you. But you’re absolutely right that we had to show that these things worked and that was kind of the dual meaning of the title, ‘The only rule is it has to work.’
0:32:50.5 BL: We took that as our principal that, “Hey, we’re gonna think out of the box here and will be creative and we’ll try anything. All that matters is that it works, so we’re not gonna be bound by tradition or anything, just by what we think would work.” But also it has to work pretty quickly or else we’re not gonna get another chance to do it. And so when we would put the defensive shifts on, which are now very standard at the Major League level, but had just never been seen at that level in 2015, we knew that over the long run, that should work, that should be beneficial to the team, that if you position your defenders in this spot instead of that spot on the whole, you’re going to save more hits than you allow as a result of that, but it might take some time to show that, and you might need the numbers to tally up all of the hits that you prevented and rate them against the hits that you allowed.
0:33:42.2 BL: And if you’re the pitcher on the mound that day, who gives up a big hit because the defenders are positioned in this spot instead of the standard spot, then you’re probably not gonna be inclined to take the long view and say, “You know what, it’ll work out.” You’ll say, “These statheads came in and they screwed up my stats because they didn’t have the defender standing there.” Like the first time that we put these defensive alignments on we thought, “Oh, this has to work this first time”, because if it doesn’t work, the players just won’t let us do it again. And so there was a lot of pressure on these extremely small samples that we knew would not be meaningful, would not be predictive, but when half the battle or more than half of the battle is just persuading people that this is a good idea, then it better work right away, because if it doesn’t, then you’re gonna be lucky to get another shot.
0:34:33.4 MK: And there was even a discussion, I remember very vividly Sam talking about the first time he was in the dugout and how you guys represented the statheads, and if you got it wrong, then you might be taking… Like everyone on the team is stepped back from believing data and believing the numbers for a period of time, so it’s not just representing your ideas… And I think this is something that analysts actually face as well. If you’re saying something didn’t go well or you make the wrong choice, the wrong recommendation, whatever it is, the belief in data itself can kinda take a step back, and I feel like that’s a lot to have on your shoulders.
0:35:15.1 BL: Yeah. Yeah, and we sort of screwed up immediately because Sam and I showed up wearing essentially the same outfit, just by accident…
0:35:23.7 MH: That’s right. The corduroy.
0:35:26.1 BL: Yeah, right. We were both wearing corduroys, this is not something we had planned, and we had our hoodies on and it was almost like our uniform or something. And it was just an accident, we had not planned this ahead of time, but suddenly, we were the corduroy crew and everyone was sort of making fun of us for dressing in the same way and in a way that ballplayers typically don’t dress around a baseball team, and so…
0:35:48.2 TW: We didn’t have matching pocket protectors, though.
0:35:51.1 BL: No, we didn’t do that.
0:35:52.8 TW: So it could have been worse.
0:35:52.9 BL: But everything short of that. And so it was little things like that where we had our heads buried in these spreadsheets to know which players are we gonna sign, and we hadn’t thought, “And what are we gonna wear?” Which turned out to be pretty important ultimately. So these little things that you might not even realize and in some sort of coldly analytical way of looking at things wouldn’t actually matter, it doesn’t matter what someone is wearing, it matters what they’re saying and whether that makes sense, but when it comes to persuading people to listen to someone they don’t know, then yeah, you better look the part as well as be able to act the part.
0:36:29.5 BL: So that was an unanticipated difficulty and we were just trying to fit in, but also trying to do something different, and that’s a difficult balance to strike. And we had some authority because we were entitled to be there, the owner had given us permission to be there and the GN had given us permission to be there, and so we weren’t out of place, technically speaking, but we felt out of place and really the sort of approval… The stamp of approval of the higher-ups only goes so far, I think. You also have to justify your presence through the rank and file, really, just to say, “Well, the owner gave us a note and said we could be there.” That just… It doesn’t go that far when it comes to actually feeling like you belong. So that got us in the door, but then we really had to prove ourselves in order to stick around and that was a constant source of stress, because if we made the wrong move or said the wrong thing and everyone turned against us, or if we took too heavy a hand and we came in and said, “Do this and do that” and alienated everyone, then well… What if they kicked us out or if they couldn’t kick us out, what if they just tuned us out?
0:37:46.0 BL: And then what happens to our experiments and what happens to our story, and what happens to our book? We’ve got nothing. So…
0:37:53.2 MK: I think this is why it helps and I enjoyed the so much, is because it really… It is about people. And as we’ve discussed in this particular case, it’s actually about people’s careers, which is possibly even heavier stakes. And with us, we’re dealing with marketers and we’re dealing with their ideas or their strategies, but we’re not necessarily reporting on things… Well, I suppose, if you take a long-term look, you could be impacting their careers, but not in the same directly tangible way that you guys were. There’s this point where you mentioned cognitive biases, and that’s the topic I’m really obsessed with. Is that something that you aware of, wanting to be liked by the player and trusted? It can sway what you tell them or how you tell… Was that something consciously on your mind about, how do I protect against my biases if I like this guy and I wanna see him do well versus this other player that is kind of a jerk, or is that something you only realized in the hindsight?
0:38:55.6 BL: No, that was definitely something we were aware of, and it was easy before we showed up and when we were still sorting spreadsheets to find potential players and kind of cold calling them and saying, “Hey, we’ve looked you up on our spreadsheets and we liked your numbers and we think you can succeed here. Do you wanna come play for the Sonoma Stompers?” That was all simple. We didn’t have relationships, we didn’t know these people. Then we show up and immediately it becomes much more difficult, because now they’re not numbers and names on a spreadsheet, they are people sitting right in front of you. And there was a point where, you have a brief spring training where you invite more players than you can keep and you have to whittle down the number who can be on your opening day roster and that means that inevitably, you are going to have to send some people home. And that can be very painful for people, even at that level, where it’s very far from the Major Leagues.
0:39:54.5 BL: If you say, “Hey, you didn’t make the cut,” in many cases, that’s the end of your dream, and even if the dream was a long shot, and even if you knew that on some level, it’s still sobering, I think, to have someone say that to you, “You can’t make this team and therefore all of the other teams beyond that are off limits to you too, and now it’s time to go back to school or study this or that or become a real estate agent or whatever is your backup plan here, you better get on that.” And so, sitting in that room with players as we are, or our manager delivered that verdict, that was tough. And of course, you don’t wanna have those conversations. And then I think we also were somewhat attached to our players, “the spreadsheet guys”, as opposed to the hold overs that maybe the manager had recruited just from the network of people he knew, we felt like our reputation was tied up in the performance of these players we had personally signed. And I don’t know if that blinded us exactly, but it certainly made us root extra hard for those players just because they were the proof of concepts. If they completely failed, then that was our failure, whereas if these other guys failed, well, that was a poor reflection on the team, but not necessarily on us.
0:41:14.7 BL: And we developed relationships with them. And of course, when you’re with a team and you’re around a team, baseball players, they spend almost every waking moment with each other over the course of a month-long season. So, how you get along with people or don’t end up being pretty important and that was not a column on our spreadsheet. So that was something that we had to figure out as we went along, if this guy is good, but he’s really disruptive, or if this guy is bad, but he’s great in the clubhouse, and everyone wants to be around him and they think that he makes them better, then how do you weigh all of that? And of course, we tried to take our responsibility seriously here so that it wasn’t just like this culty-clinical experiment, but we wanted to give deference to the dreams of these players and their careers. Even though the odds were against all of them, we didn’t wanna come in and sabotage anyone or make that harder for them. We hoped we would help by being able to provide this information.
0:42:17.4 BL: And really the greatest accomplishment, I think, the thing we were proudest of was that a couple of the players we’ve signed, sight unseen, off of spreadsheets who just had been at home and passed over by every team, they actually ended up doing so well that they then got signed by major league organizations later. Now, they didn’t make the majors, they didn’t make it all the way, but they got that shot, which they had not gotten before we showed up, and they got to live that dream. And so, that was really vindicating for us and I think also for them and that was what we were going for, more so than we wanna win a championship or we wanna prove that the specific strategy works, we wanted to hopefully find some gems, some diamonds in the rough that had been passed over because they were too short or they didn’t throw hard enough or whatever the reason was that they were not attractive to teams or had not been to that point. But we looked at the numbers and thought, “These guys could succeed despite not having the typical profile.”
0:43:18.4 TW: Which, that was one of the more fun parts earlier when you were talking about going through and describing… It’s again, sometimes as analysts, we think, “Oh, if I just crunch the numbers right, it’s gonna truly be diamonds in the rough.” And as you went through and explained that it’s like, “Look, these guys have been reviewed and passed over many, many times by sophisticated organizations with a lot to crunch.” So, it was gonna be unlikely that you were gonna find a major league all-star, that would be incredibly improbable. You do have what you have, and the data is not gonna magically turn over… It’s only gonna be able to take you so far with what you’re starting with, but that was… I love that explanation as you…
0:44:09.2 BL: Yeah, teams had seen something or not seen something in all of these guys, there is a reason why they were available to the Sonoma Stompers, and in some cases, those reasons did not preclude their performance at that level. We were looking at players who had succeeded in college and we thought, “Well, this is not that far from a high level of college baseball, this is kind of an intermediate step and so, if they were good there, then they could probably be good here and in some cases, that turned out to be true, in other cases, not so much. There was some fatal flaw, at least at that time in that season where they were able to produce… When they pitched it at their old teams, but not for us. And there was one particular player who ranked really highly on our spreadsheets, he was a hitter who had just dominated his region in college baseball, and we thought, “Why is this guy available? How is this possible? Look at these numbers.”
0:45:05.2 BL: And we were able to sign him and then he showed up and his swing just did not look like a major league hitter swing. We weren’t scouts obviously, but even we could see something looks a little funky here, and that was not something that we had on our spreadsheet either. Now, if you have a comprehensive tracking systems and some of the information that major league teams have available to them at some levels today, well, you can track that and then that does become a cell on your spreadsheet, but for us in that place at that time, it was not, we knew the results, but not necessarily the process. And so this player showed up and some of the old school baseball types were saying, “Oh, this is not gonna work. Look at that swing, he’s not gonna be able to hit the pitching here.” And we said, “But look at these numbers. How could he not hit the pitching here.”
0:45:55.5 BL: And that was a case where, at least in that trial, they were right and we were wrong, and he did that really produce. And there’s still a part of me, going back to the biases that we were talking about, there’s still a part of me that thinks that he would have produced, given a little more time, given another season just because he had succeeded so well, but not necessarily. Maybe he just had the skill set to dominate at one level, but not at the next, and maybe he isn’t the type of player who I wrote about more in my second book, The MVP Machine, who just needs some change to be made, he needs a swing change, he needs a new pitch, something different where he has the raw ability, but he needs to be refined in some way. And with the Stompers in 2015, we just did not have the time or the resources to be able to do that. So, players were pretty much finished products when we got them, we were trying to find undervalued players who could succeed from day one, more so than moulding them into the players they could potentially be.
0:46:56.3 MK: One of the things that you talk about in the set-up phase is pulling some other data nerds to kinda help you do some of the analysis, and you talk a little bit about automation. And this is something that as data practitioners, we’re always kind of dealing with. It’s like you need the data to be readily available, it needs to be automated, and I guess accessible. What was that process like for you? ‘Cause you can get in that stage where you’re like, “If I have this one more number, it’s gonna help me get the answer, but then you’re also trying to madly scramble to pull this data together to make sure that you have what you need so that it’s done in time to be helpful for the decisions they’re gonna get made, where is that balance between automation and having it readily available versus actually just getting it done in time so that it can influence decisions?
0:47:55.0 BL: Yeah. So, as I mentioned, I’m an English major and therefore was not equipped with the full suite of analytical chops or data retrieval and dissection, and so I’ve always relied on the kindness of strangers or near strangers to some extent, where I might know what information is out there and I might have a good question and know how to frame it and know what answer I want, but I might need to rely on someone else to actually pull that for me. And so that was really one of our strengths in this experiment, we hoped, is that we knew no one… We didn’t know any players, we didn’t know any baseball people at this level who we could just call up and say, “Hey, I played with you there in that year, would you like to come play for the Sonoma Stompers?” We didn’t have that network, which some of the traditional figures on the team did. But what we had was a network of essentially statistical consultants, people who work in baseball or are interested in baseball and were aware of our work and were intrigued by this project and were willing to help us out.
0:49:00.7 BL: And so, we actually got the assistance of multiple people, one in particular who had done this sort of work for major league teams and we were able to secure his help just because, hey, we’re working on a book and he thought this was fun, and he helped us out. And we end up with this data set of college stats, which we tried to look for the most predictive metrics, what at that level would predict success at the next level and we worked with what we had. And again, this was several years ago and even in that fairly short time, things have changed dramatically, where now we would be looking at this tracking technology and how hard does this guy throw and what does he throw and how much do these pitches move and how hard does he hit the ball? At that time, we essentially had none of that, we knew the outcomes, and so we had to make do with that, which was imperfect, but really the best we had, and we didn’t have a ton of time either. So, if we could have, we would have done background checks, we would have done our homework on these players and said, “Hey, this guy has a good strike out rate, but is he a good guy, can get character references, what’s his life story?”
0:50:09.4 BL: That’s what you would normally rely on scouts for, to do that sort of human information dive, and we couldn’t do that really. And so again, it was kind of flying by the seat of our pants to an extent, and if we had done a second summer, I think we would have been much more adept at all of that, but we used what we had and fortunately for us, at least it was more than our opponents at that league, at that level had. So that was one of the advantages, is that we weren’t coming into major league baseball and trying to go head-to-head with these full quantitative departments, we were going up against the intern, essentially, who had none of this information. And so, we kinda had an unfair advantage in that sense and that helped us out a bit. And also, we found that we had to provide this information in a timely way to the people who were going to use it, so we spent a lot of time kinda crunching the numbers behind the scenes, but then we had to get to the park at a certain time if we wanted to hand out the scouting reports, for instance. Or we even had to figure out, say, how much information do you put on the little white board that we have in the dugout with us, where you’re trying to tell hitters about the pictures they’re gonna face.
0:51:23.4 BL: Because the first time we did that, it was crammed. Every inch of the thing was, here’s what he throws on this count and here’s how hard he throws and then the players were like, “This is… I can’t possibly keep this in my mind as I’m going up to the plate and trying to hit this thing. Tell me one thing or two things, maybe. Does he throw a curve ball or what? What should I look for on this count, should I worry about his fastball or his off-speed stuff?” It was that simple. And so we really had to scale it back, whereas our initial thought was, let’s just bombard them with all of this information that we have, but it turned out not to be all that useful when we did that. And we also found that when we interacted with our manager… And I won’t spoil what happens to the first manager in the book for anyone who hasn’t read it.
0:52:11.4 TW: There’s a chapter about it.
0:52:14.6 BL: Yes. At least one, and we end up with another manager who was maybe more receptive to our way of thinking, and we came to work with him. And so we would help him construct the line-ups, and we had to build this interface, essentially, this web tool that he could access if you wanted to look up certain things and try to rely on this information. It was one thing for us to just dump all of this data on him, but we actually had to present it in kind of an intuitive way and put it there on the page so that he could see it, so that it was sort of real for him, that he wasn’t just taking our word for it, and then we had to figure out how to present that information. And just going back to when I was talking about the diplomacy aspect of it all, initially, we were in the dug out and we were saying, “Hey, maybe you should make a pitching change now or something.” And it turned out that, at least from his perspective, we were kind of making him look bad or making him look like we were just pulling the strings and he was just the puppet, we were kind of sapping some of his authority.
0:53:18.7 BL: And so he said, “You can’t just come to us and tell me what to do,” essentially in the moment, in the dug out, and he didn’t like that because he wasn’t used to that in his previous experience. And this is a manager who now is working in the Minnesota Twins system. So, he too got a chance and got to advance beyond that. But we had to figure out, “Okay, we can come to an understanding here where we won’t show you up in the middle of the game in front of all the players so that it’ll look like you’re just following orders, we’ll talk to you before the game.” And we’ll say, “Hey, if this situation arises, we think you should do this.” And then we’ll talk to you again after the game and we’ll say, “You know, what were you thinking in this situation? Why did you make this move? Why did you not make that move? We think maybe it would have made sense to do this if that situation arises again.” So he kinda gave us the feedback that, “You are giving me useful information, but you’re giving it to me in a way that I can’t make use of and I can’t get the full effect of.” And so we had to adjust our approach too to make sure that we were actually putting this out there in an actionable and acceptable way.
0:54:24.6 MH: And it’s time for another quizzical query of the Conductrics nature brought to you by Conductrics. If you’re wondering if there’s an AB testing platform that’s really gonna get in the trenches with you and make sure things are successful, that’s what Conductrics does. They’re helping some of the largest companies run AB testing and personalization, you should check them out at conductrics.com. Tim and Mo, are you ready?
0:54:49.1 TW: Still not ready.
0:54:52.3 MH: Yeah, well, this is our third one, and we’re learning a lot. But if we…
0:54:55.6 TW: Let me make sure my coin is polished and ready to be quietly flipped.
0:55:01.5 MH: Well, we’ve got a little wrinkle this time, because there’s five optional answers, so.
0:55:07.3 MK: Oh, jeez.
0:55:07.3 MH: We’re gonna get to a winner, but it might take a… Just in a longer. But let’s talk about who you’re representing tonight. So, Mo, you’re representing someone from Tim’s hometown in Columbus, which is Katie Bradley and Tim, you’re representing Jaden Sender. So, no pressure, this is all fun, but also it’s very important.
0:55:32.9 TW: It is fun for you.
0:55:33.6 MH: It is fun for me, I’m having a blast at this, I will admit it, I’m loving every minute of it. Okay, Mo and Tim, you’re both exceptional at analytics, really, really great. And so you’re brought in to a whiskey distillery as analytics consultants, and they would like to determine if of the following have an effect on the amount of the angel’s share. Now, the angel’s share is the amount of whiskey in the cask that evaporates before bottling happens. The treatments of interest are, one, effect of two years versus three years maturing in the cask, two, the percent of alcohol for the new spirit when it’s put into the cask, either 50% or 60%, and three, the location where the casks are kept while the whisky matures, Texas or Kentucky? I think he’s asking us a real question that he wants an answer for…
0:56:28.4 TW: I’m pretty sure this was inspired in that I was touring a distillery in Kentucky last week and the guy made a disparaging remark about Bourbons that were not distilled in Kentucky, so…
0:56:39.8 MH: So, we’re getting down to… This is some real-world actionable analysis that we’re conducting here. You collect the data using a factorial design, so eight combinations of the three treatments, what analysis do you recommend and why? And I’m gonna give you some options here. A, analyze each combination as an RCT with a Welch’s t-test.
0:57:02.7 MK: Ohh.
0:57:05.2 MK: He’s just not trying to kill me slowly.
0:57:10.6 MH: I’m sorry. I was supposed to retain a professional announcer demeanor. B, analyze the data using a regression model with an intercept term and encoding the treatments into seven dummy variables. C, analyze data using a regression model with an intercept term and encoding the treatments into eight dummy variables. D, analyze with fixed effects analysis of variance. Or lastly, E, analyze each of the eight combinations as a time series and then use a Dickey-Fuller test to test for a unit root. Alright. That’s a lot.
0:57:50.1 MK: I feel like I should have taken notes.
0:57:52.4 MH: Well, I think Tim was taking notes.
0:57:52.9 TW: I was… I was…
0:57:54.0 MK: I think Tim is.
0:57:54.1 MH: Yeah.
0:57:54.9 TW: Yeah.
0:57:55.5 MH: Tim… Okay. But as history has shown, Moe, a random guess is as good as an educated guess in this quiz so far.
0:58:03.3 TW: No. I’m gonna try to shorten this and I am going to…
0:58:06.7 MH: Uh-oh.
0:58:07.3 TW: I’m going to come out bold and I’m gonna eliminate C for eight dummy variables, ’cause if you’ve got… There’s no way you would need eight dummy variables, and mostly you would need seven, maybe that’s totally wrong. And I’m gonna eliminate the time series because you’ve got…
0:58:24.6 MH: Which one?
0:58:25.6 TW: That was E.
0:58:26.5 MH: E. Okay, yes, as a time series.
0:58:27.2 TW: Yes. So I am gonna declare with confidence that we are down to A, B, or D. And…
0:58:33.2 MH: Alright. Well, that should be exciting. I think you can… I think, Moe, you just gotta phone a friend from your competitor.
0:58:42.8 TW: Maybe, or maybe I’m playing mind games then, and that’s like…
0:58:45.8 MH: Oh. Yeah, let’s add that… Tim, great job. Let’s add that to the complexity of this one, just like, get some tricks.
0:58:52.8 TW: Wait a minute, you can… If you wanna move things along, you can go ahead and say, “You’re right, Tim, it’s not one of those two.”
0:58:58.9 MH: Actually, you’re spot on, it’s not one of those two.
0:59:00.7 TW: Okay. Okay.
0:59:01.6 MK: Great.
0:59:02.7 TW: Okay. Now I’ll let Moe take the first guess.
0:59:06.7 MK: No, I’m like…
0:59:08.2 MK: I’m not saying anything. I’m just like, you go for it and I’ll pick from what’s left.
0:59:12.7 MH: Well, there’s three choices left, A, B, and D. A is analyze each combination as an RCT with a Welch’s t-test. B, analyze data using a regression model with an intercept term and encoding the treatments into seven dummy variables. And, D, analyze with fixed effects analysis of variance.
0:59:35.3 TW: Okay. I’m gonna talk this out, make my guess and then, Moe, you can make a conflict for the other two. I’m gonna say, this seems a lot like it would be like an A/B test on a website that you’ve got these different treatments, you’re just applying the treatment. And I assume there is some objective measure of the outcome and that’s probably Heather’s palette saying which is better. So I am gonna say this seems very much like an A/B test, which is a form of a RCT, and I’ve heard of a Welch’s t-test, but I’m gonna go with A.
1:00:14.5 MK: Fucker. That’s…
1:00:14.8 TW: ‘Cause I could cut myself out of B and D.
1:00:17.1 MK: I feel like it is actually A, is what I was gonna guess, but that’s… I’m gonna…
1:00:20.7 MH: Alright. Well, Tim selected A, but… So, Moe, you’ve got a choice between…
1:00:23.3 MK: No, but then I also was like, I’m not gonna lie, I got lost in the detail of what we were trying to figure out, but there were three different things we were trying to figure out, and I was like, “Man, you’d have to run a lot of tests to figure out all those different variables.” But anyway. Okay. So then I’m just gonna guess B or D. B was regression. And what was D?
1:00:42.9 MH: Yep, regression model. D, analyze with fixed effects analysis of variance.
1:00:48.6 MK: I don’t know what that is, but I’m gonna go with D.
1:00:52.7 MH: D. Okay.
1:00:54.5 TW: It’s ANOVA. I mean, so it’s a…
1:00:55.5 MH: Yes. Are you ready? I’m not gonna hold you in suspense. The answer was actually either B or D, and so Tim…
1:01:03.9 TW: So Moe wins.
1:01:04.1 MH: You managed to pick the wrong one.
1:01:06.3 MH: Moe, you are the winner.
1:01:08.4 MH: Because Tim stole your wrong answer now.
1:01:11.9 TW: Awesome.
1:01:15.8 TW: How do I get this is for you?
1:01:17.4 MH: Katie Bradley…
1:01:18.2 TW: Oh, you’re welcome.
1:01:19.7 MH: Katie Bradley, you’re a winner.
1:01:20.9 MK: Oh, I’m so excited I won by like pure fluke.
1:01:25.8 MH: But you were right about the fact that B and D are both acceptable, both are members of the general linear model, ANOVA, formalized by Roland Fisher. I’m not gonna read the rest of that, but basically…
1:01:39.5 MH: A is less efficient than B and D and doesn’t allow for estimating marginal effects of each treatment setting, which I’m sure was what was on top of your mind, Moe, as you were selecting B or D in that answer.
1:01:50.1 MK: Totally. Yeah.
1:01:52.2 MH: And I’m reading from the page, so I can act like I know that. Anyway. Alright. That’s been the conductor’s quiz. So hey, send us a DM on Twitter or reach out to us on an email contact at analyticshour.io if you’d like to be included in the quiz. Moe and Tim going at it on your behalf and winning prizes, pretty exciting. Thank you, Conductrics. Check them out at conductrics.com. And let’s get back to the rest of our show. Alright. Last question, Jonah Hill played the role of Paul DePodesta in Moneyball, so when this is made into a movie, who’s gonna be representing Ben Lindbergh, who’s the actor?
1:02:39.6 BL: I feel like I’ve gotten that question before, and I…
1:02:42.4 MH: Oh, okay.
1:02:43.0 BL: Forget what my stock answer was. I would… I’d be fine with just about anyone as long as they’re true to my character, as long as they portray me as the geek that I was and not some suave, self-assured Brad Pitt type, like in Moneyball. I think Jonah Hill did a good job there. Yeah, anyone who looks out of their element and slightly uncomfortable and worried about how all of this is going to work out, I think would be the perfect choice to play me.
1:03:14.2 MH: Nice.
1:03:15.5 TW: We can have listeners. That can be their proof they made it to the end of the episode, they can tweet a recommendation.
1:03:21.6 MH: Tweet a recommendation. Go for it. Alright. Let’s roll into our last call, so this is something we do on the show where we go around the horn, just share something that we think might be of interest to our listeners. Ben, you’re our guest, do you have a last call you’d like to share?
1:03:38.9 BL: Sure. Well, I know this doesn’t have to be about baseball, but I figured I would make it about baseball just for the benefit of any baseball people who are tuning in or wanna find out more about baseball. One problem with baseball is that there’s a lot of it, there are 162 games per team, per season, and there are 26 players per team. So we’re talking games going on everyday, every night, simultaneously, and it’s kind of overwhelming. And so you might figure, “Well, how do I keep track of all of this, and what do I wanna watch?” Because if you subscribe to Major League Baseball streaming service, MLB.TV, you can access just about any game that goes on, and often there are up to 15 games going on at the same time. So one thing that I found that is very helpful, and I’ve written about this and recommended it before, is a tool called the Stream Finder, which is available at baseball-reference.com, and this is the stat site of record, the statistical compendium of Major League Baseball.
1:04:41.4 BL: But it has a little known utility called the Stream Finder, used to be called the Game Changer, where you can essentially pre-set your conditions and your priorities when it comes to games, and then the Stream Finder will select the games for you. You can just sit back and say, “I wanna watch this player, when this batter comes up I wanna see that, when this pitcher is on the mound, I wanna see that, if this runner gets to second base and maybe he’ll steal I wanna see that, if there’s a no-hitter going on I wanna see that,” or if certain preset conditions are satisfied, if the leverage index is high enough, that is the stakes, the suspense are ratcheted up enough, then I want my game to automatically switch to that. And it’s great because you can set up dozens of these conditions and rank them in terms of priority, so if you have a fantasy baseball team and you have a bunch of players on that team and you wanna keep track of how your players are doing, you can mark them in these drop-down boxes and say, “When this guy’s up I wanna see that guy, and if this guy comes up, oh, switch to that one, that has a higher priority, he is more interesting to me.”
1:05:49.6 BL: So I think it’s a very useful tool, and there’s a lot of dead air in baseball. Baseball games that go on for quite a long time, they’re not always action-packed, and so if you want to bring a little of that intensity, if you’re a football fan, American football, that is, and you watch the NFL RedZone channel, which is kind of, “Hey, when people are about to score, that’s what we’re watching,” and it’s just always ratcheted up and super exciting. That’s what this turns baseball into, where it’s not just this lazy activity that’s on in the background, but whenever you’re watching something that you wanna see is happening. So again, that’s the Baseball-Reference Stream Finder.
1:06:30.4 TW: Is that part of your secret then for Effectively Wild? ‘Cause when I listened to the episode, I’m like, “Oh… ”
1:06:35.9 TW: It literally seems like you watched seven games yesterday.
1:06:40.3 BL: Yes, yes, that is part of my secret. Many tips and tricks to try to keep track of these things because no one has enough time in the day to keep track of all of the baseball.
1:06:46.2 TW: To watch all the baseball. Wow.
1:06:51.5 MH: Yeah. Does it allow you to target games where there might be a weird booming noise coming from the dugout or pitches?
1:06:58.6 MH: No, I’m just kidding.
1:07:00.5 TW: I don’t think it can do that yet.
1:07:01.6 MH: Well, hopefully, that’s not happening much anymore.
1:07:03.6 BL: Yes.
1:07:03.8 MH: Okay. Moe, what about you? What’s your last call?
1:07:07.7 MK: I have two because they’re little bite-sized ones. The first one is from a guy called Mike Acton, it’s a Medium article, which I really enjoyed, it’s called “Stop Wasting My Time With Your Shitty One-on-One Meetings.”
1:07:23.1 MK: And the reason I really like it is because, as a coach, I end up being in lots of one-on-one meetings, lots of coaching sessions, and they can either be amazing or they can be total train wrecks, and I just thought this guy, Mike, did a really good job of summarizing a whole bunch of questions for people that do struggle to have a bit of direction in their coaching sessions with their team. I thought it was a really strong article about, how can you make this meeting really, really useful and productive? And he’s got a whole bunch of questions in there if you’re stuck for what to ask your team. So yeah, that’s a really good resource. The other one as I’m gearing to head back to work, which someone shared with me is “Texts To Send A Mom Who’s Going Back To Work.” And I just really loved it because I feel like when moms are going back to work, it can be a pretty big change, and I think sometimes people in the workplace don’t really know what to say to you or your friends and family, and there was just some really lovely suggestions in there that I was like, “Yeah, that would be a nice thing to hear as I’m gearing up for my first day back.” I’ll share the link to that one as well in our show notes.
1:08:33.4 MH: Good, then we can text you when you go back…
1:08:36.6 TW: If your baby could talk, they’d say, “Good luck at work, mom.”
1:08:40.4 MH: That’s awesome. Thank you, Moe.
1:08:41.3 TW: Your feelings right now are so valid. Oh, this is great.
1:08:44.5 MH: Yeah, I’m excited about both of those last calls. Thank you. Those are gonna be personally helpful. Okay. Tim, what about you? And by golly, I hope you don’t take mine.
1:08:56.4 TW: I’m pretty sure I won’t. I’m gonna do a twofer as well. One is a… It is my now absolute favorite former MLB player, and this is, thanks to Ben, is the book, ‘The MVP Machine’. Deep in the book it just has this passing reference to Nate… Is it Freiman? Freiman?
1:09:16.1 MH: Oh, I think it’s Freiman. Yeah.
1:09:18.4 TW: Freiman? Nate Freiman?
1:09:20.6 MH: Yes.
1:09:20.7 TW: Who… And I’ll read the one sentence, “Freiman, who minored in Math at Duke, started taking online courses in machine learning and the programming languages, SQL and R.” So, I shed a tear of joy that there was a former Major League player who had spent some time with R.
1:09:38.4 MH: Yes, so did [1:09:38.5] ____ that was very validating for all of us in the sabermetric community, when we had a former Major Leaguer who was doing this sort of stuff. And actually, he subsequently was hired by Cleveland and he is now working in their front office, so it paid off for him too.
1:09:51.8 TW: Oh.
1:09:52.7 BL: Wow.
1:09:54.7 TW: Well, now I have to admit that he was our first choice as a guest and he said no…
1:09:57.7 BL: I wouldn’t blame him.
1:10:01.8 TW: So, that’s a little bit of a point for The MVP Machine as well, ’cause that’s also a fascinating book where it flips it around to what players can do to personally optimize with more of the technology, so I really enjoyed The MVP Machine as well.
1:10:17.5 MH: Thank you.
1:10:18.2 TW: But my other last call was, I had a little bit of a rant on Twitter a while back and [1:10:24.7] ____ read through the whole thing and she actually shot me a paper called ‘Conducting Research With Quasi-Experiments, A Guide for Marketers.’ It looks like an academic paper and I think the guys are academics, but it’s actually digestible and consumable, it’s like actually being able to read and abstract and the font and the whole academic paper. But it just does a really, really good clear explanation of some metric methods and different sorts of examples related to research, and it was a really interesting read to actually get to read an academic paper and follow along with it, but good information in it as well. What about you, Michael?
1:11:07.3 MH: Well, interestingly enough, in this world of analytics-related podcasts, we’ve actually got a new… I don’t know, is it a sister podcast, a brother podcast? I don’t know, but it’s another podcast that was launched earlier this summer by none other than Simo Ahava, which is the Technical Marketing Handbook. And if you’re a listener to this podcast and you probably know about this podcast already, but you should probably give it a listen. In fact, it’s probably already more popular in our community than ours is ’cause…
1:11:40.6 MH: Given who Simo is and his impact on the community. But well worth a listen, there’s a few episodes already out and very specific information but it’s really high quality as is anything Simo usually does.
1:11:56.5 TW: He was a guest with us a few years back and he got the bug.
1:12:00.6 MH: Yeah. Well, we paved the way for so many. What can I say? No…
1:12:04.7 MH: Alright. The book is called ‘The Only Rule Is It Has to Work.’ Ben Lindbergh is our guest. Ben, phenomenal book. It took me by surprise actually, like we’ve talked about this last hour, it just… It was a really enjoyable read. And I think if you’re in analytics, you’ll enjoy it, even if you’re not a baseball fan, I’m not a huge baseball fan, but actually, I was pulling so many things out of the book that I felt like I could use. Well worth it. Thanks so much for being on the show, Ben, really appreciate it.
1:12:33.7 BL: My pleasure, yeah. Thanks to all of you for having me.
1:12:36.5 MH: Alright. So you know that we would love to hear from you, and the best ways to do that is our home away from home on the web, which is the Measure Slack channel and also on our Twitter account or our LinkedIn group. We’d love to hear from you, give us your comments, let us know who should be playing Ben in the movie. I’m thinking Timothée Chalamet maybe. But let us know on Twitter, we’ll get that out to the appropriate Hollywood folks, ’cause I’m sure… Moe, you’re probably the one with… Probably the most likely to have Hollywood connections, you’re cool. So, I figure…
1:13:13.1 MK: Really?
1:13:14.4 MH: Well, it’s not gonna be Tim or me.
1:13:19.2 MH: Anyway.
1:13:19.3 TW: It’s all relative.
1:13:20.8 MH: No show would be complete without a shout-out to Josh Crowhurst, our producer. And if you’re vibing on that new intro and outro theme, you know what, Josh wrote that too, and it slaps as the kids say. So anyways, yeah, we really appreciate everything you’re doing, Josh, thank you so much for all your help with the show. And remember, whether your on-base percentage is hovering in the low 400s or way over 950, if you’re an analyst out there, I know I speak for my two co-hosts, Moe and Tim, when I say, keep analysing.
1:14:00.5 Announcer: Thanks for listening. Let’s keep the conversation going with your comments, suggestions and questions on Twitter at @analyticshour, on the web at analyticshour.io, our LinkedIn group, and the Measure chat Slack group. Music for the podcast by Josh Crowhurst.
1:14:18.5 Charles Barkley: So smart guys want to fit in, so they made up a term called “analytics.” Analytics don’t work.
1:14:25.1 Thom Hammerschmidt: Analytics. Oh my, God. What the fuck does that even mean?
1:14:34.1 BL: Yeah, I grew up a Yankees fan too, as I guess you know just because I happen to live close to Yankee Stadium and they were winning the World Series every year at the time, so it was a natural fit for me, but I did not have warring factions within my family, that’s an interesting dynamic.
1:14:51.2 MH: As someone who grew up in Cleveland, I didn’t like people like you growing up, Ben. No one does.
1:15:05.5 TW: If the frog right outside my window starts back up, I don’t know what I’m gonna do.
1:15:09.9 MK: Intriguing.
1:15:15.8 MK: Sorry, one second, my fucking dog is drinking.
1:15:18.9 TW: Sure.
1:15:24.7 TW: Rock flag and play ball.
Subscribe: Google Podcasts | RSS
This site uses Akismet to reduce spam. Learn how your comment data is processed.
https://media.blubrry.com/the_digital_analytics_power/traffic.libsyn.com/analyticshour/APH_-_Episode_220_-_Product_Management_for_Data_Products_and_Data_Platforms_with_Austin_Byrne.mp3Podcast: Download | EmbedSubscribe: Google Podcasts | RSSTweetShareShareEmail0 Shares
Subscribe: Google Podcasts | RSS