Counterpoints /

The Sports Analytics Podcast from MIT SMR

“Out-Getting” Is the Next Great Baseball Strategy. Or Not.

play Listen
« Prev Full Series Next »

Most Counterpoints listeners are well familiar with the sweeping changes that Moneyball brought to offensive strategy in baseball. Are we now on the cusp of a similarly profound transformation of pitching? Some believe the practice of out-getting is just that. Out-getting focuses on getting outs as efficiently as possible, regardless of who gets them, when, and how. It upends traditional convention about how long or frequently a pitcher should pitch, in which role, or in which circumstances. Will out-getting prove to be a Moneyball-level transformation that ushers in fundamental change — or just an occasional strategy deployed as much by desperation as by tactical brilliance? Ben and Paul disagree.

Transcript

Ben Shields: The A’s opened this year’s A.L. wild-card game by trotting out a middle reliever to pitch to the big bats of the Yankees. The Brewers pitched their N.L.C.S. Game 5 starter for exactly one batter — by design. And the Red Sox got the final three outs of the World Series not from their big-time closer — but from their ace starting pitcher. Three separate events — but all connected by the simple concept that baseball teams need to get 27 outs. And it doesn’t matter who, when, or how they show up. Now for the A’s and Brewers, their strategies didn’t entirely work out. But in the long run, the 2018 postseason will go down in history as the year where managers threw pitching convention out the window and embraced the art of out-getting.

Paul Michelman: Except Bucky Harris and his 1924 Washington Senators would like you to take a closer look at those history books. You see, a full 94 years before these so-called revolutionary ideas about pitching emerged, Harris needed to gain an advantage in Game 7 of the World Series, facing the vaunted lineup of the New York Giants. So, he did what any forward-thinking manager would do: He started right-hander Curly Ogden to goad New York into fielding a lefty-heavy lineup. Then he pulled him after just two batters. And in the ninth inning with the game tied at three, who did Harris turn to? Why his Hall of Fame starting pitcher Walter Johnson, who went four scoreless innings and picked up the win. So-called openers in playoff games? Starting pitchers closing? Sounds pretty 2018 to me. Like many things in baseball, what’s old has become new again. So, why focus on out-getting now? Because there are reasons to believe that its time has finally come as a major part of pitching strategy. I’m Paul Michelman.

Ben Shields: I’m Ben Shields and this is Counterpoints, the sports analytics podcast from MIT Sloan Management Review. In this episode: Will the practice of out-getting do for pitching strategy what Moneyball did for offense?

Paul Michelman: So, we’re all familiar with the changes that Billy Beane’s Moneyball brought to offensive strategy. Get on base. It doesn’t matter how. Don’t steal. Power matters more than hits. These things transformed the game.

Ben Shields: So, will out-getting be the new big thing? Just as Moneyball was about producing runs as efficiently as possible, out-getting is about getting outs as efficiently as possible. It doesn’t matter who gets them or how — out-getting rethinks the roles of starters, middle relievers, and closers. It eliminates traditional convention about how long a pitcher should pitch or in which circumstances. As Moneyball brought a new systemic approach to getting runs, out-getting promises the same for getting outs. The focus moves from the player to the outcome.

Paul Michelman: But here’s where the sides get drawn. Some observers, including me, believe out-getting will indeed prove to be a Moneyball-level transformation that pervades the league and ushers in fundamental change.

Ben Shields: Others, including me, see it as a very useful strategy, but one that cannot rise to the level of Moneyball in terms of impact.

Paul Michelman: So, in this very special episode of Counterpoints, Ben and I go head-to-head in the debate on the significance of out-getting. Ben, I think Oxford debate rules dictate that the case is first made in favor of the motion. So, allow me to present my data-rich, airtight case for why out-getting will be as influential a pitching strategy as Moneyball has been on offense.

Ben Shields: Let’s do it.

Paul Michelman: Out-getting is a smart and sensible reaction to what the numbers say. The analytics clearly indicate that hitters adjust to pitchers better than pitchers adjust to hitters. You need 27 outs to win a game. If the data says there’s a more efficient way to get those outs, one that is superior to the outdated practice of trying to squeeze 18 or more outs from one player while 10 others sit on their butts and spit tobacco (don’t chew kids!), who wouldn’t follow it, especially when it leads to better business outcomes as well? So, there are four big arguments in favor of the motion: (1) The legacy approach to pitching management is woefully inefficient, and I have the data to prove it. (2) Out-getting promises not just a more efficient approach as an on-field strategy but as a business strategy as well. (3) The model for out-getting has already been shown to be successful both within baseball and without. (4) Out-getting is already an established practice with major market teams — it took years for Moneyball to establish itself in this way.

All right, let me dive into the first point. As soon as the first pitch is thrown in a game, hitters start gaining on pitchers. [There’s] a really telling statistic that demonstrates how through each cycle through the lineup, a pitcher’s effectiveness decreases — and decreases significantly. We’re going to look at a stat called wOBA, that’s weighted on-base average. It’s a version of on-base percentage that accounts for how a player reaches base, instead of simply considering whether he reached base. The value for each method of reaching base is determined by how much that event is worth in relation to projected runs scored. I know that’s a mouthful, but basically what we’re saying is a double is worth more than a single, a triple’s worth more than a double, a home run is worth most of all.

So, we’re going to look at the weighted on-base percentage for the average starting pitcher in 2018. The first time through the lineup, in the American League, the average batter had a 311 wOBA. The second time through the lineup, it rose to 323. And the third time, it jumped to 338. Same pattern in the National League: first time through the lineup, 298; second time, 310; third time, 333. Those are remarkable jumps in hitter effectiveness against the same pitcher. It wasn’t a one-year aberration; 2017 saw similar numbers. And the pattern holds across pitcher quality. If you look at the top Cy Young finishers in both leagues in 2018, the first time through the lineup their ERA was 1.86; second time, 2.47; third time, 3.32. If you look at the difference between a pitcher’s first time through the lineup and their third time through the lineup, you’re talking about a jump of 1.5 earned run average, and it’s close to double from the first inning to the third.

Look, we already know that managers don’t trust their starters. And every year they trust them less and less. Pitchers are getting pulled earlier in the game. And so, the role of starter was already evolving from this kind of legacy model of “go seven, eight, even nine innings” to “five innings is considered success.” The strategy works. Why wouldn’t we just continue on that path? If we see that a pitcher’s effectiveness materially decreases after three innings, pull them after three. Out-getting is driving us towards a much more efficient approach. The second part of the argument is that out-getting is every bit as good a business strategy as it is an on-field strategy. The highest-paid members of your pitching staff are your starters, followed usually by your big-name closer. Your middle relievers are often paid a fraction of what front-line starters and big-time closers get. Well, in out-getting everyone’s a middle reliever. There’s no need to pay Chris Sale $30 million a year. His role isn’t that valuable anymore…. As out-getting transforms on-field strategy, it’s going to return big bucks to the bottom line.

Third part of the argument is that the model is already proven. Inside baseball, we’ve seen clubs like the Rays, and maybe more significantly, the Brewers, use this strategy not just in the postseason but throughout the year. The Brewers got to the postseason on the back of this strategy. Outside baseball, there are models that demonstrate how this type of thinking can be particularly effective. Consider the notion of running back by committee in the NFL, which is a very similar approach. The traditional model in the NFL was you have a star running back you’d feed the ball [to] over and over, up to 30 times a game. The problem? They not only get tired and break down, it shortens their careers, and you’re paying your star running back a huge salary. Teams have demonstrated that by having two or three people share the load, they get the same, if not better, in-game performance. And similarly to the second argument, they’re saving money on the payroll and extending players’ careers.

The fourth part of the argument returns us to Moneyball. Moneyball didn’t really take off until the big-market teams started following the Oakland A’s example — and that took several years. And while the Brewers might be the biggest out-getters this year, that doesn’t mean that other teams aren’t already using that strategy successfully. The Red Sox pitched both Nathan Eovaldi and Rick Porcello out of the bullpen this postseason. The Dodgers closed out Game 7 with Clayton Kershaw. In recent years, you’ve seen Justin Verlander, Charlie Morton with the Astros; Jon Lester and Kyle Hendricks with the Cubs; Madison Bumgarner with the Giants — all pitching in relief. Where Moneyball took half a decade to go from small-market teams to big market, out-getting is happening almost immediately. All right, Ben, knock me down!

Ben Shields: All right. Paul, thank you for that compelling opening statement about the importance of out-getting for pitching strategy. And I must say, it’s interesting to find myself on the other side of this debate. I’m often always the pro-analytics advocate. But for this particular debate, I figured I would take the perspective of the other side in order to understand this issue from a wide variety of perspectives. And I’m going to get into my argument here in a minute, but I want to offer one quick caveat, which is I don’t think we have enough information yet on the validity of out-getting as a transformational strategy within Major League Baseball. The examples that you cited, whether it be the Brewers or the A’s or the Rays, did experiments this past year. Those experiments were interesting. It yielded some results. But we’re still, I think, a little ways away until we can make a grand claim about the influence of out-getting on baseball at the same level as Moneyball for offense. With that caveat now to the side, I want to dig in here and take a look at this argument, ’cause let’s say it does become successful. Then we have to look at the prompt and say, is it going to be on the level of impact that Moneyball was for offense? And from my standpoint, the answer is categorically no. And I’m going to go through three main arguments in order to prove this.

The first is that Moneyball is a false comparison for out-getting. Second, out-getting exists as a strategy because of a team’s lack of strong starting pitching. If they had better pitchers, then maybe they wouldn’t be using out-getting as much. And finally, which could be the death knell for out-getting: It will only exacerbate the pace-of-play problems that Major League Baseball has. And I wouldn’t be surprised if over time there is actually a detrimental effect to the game and there is a cap on the amount of relief pitchers that teams can employ per game.

But first, I want to talk about this notion of Moneyball. Moneyball is a false comparison. Now, the easy way to make this argument is to suggest that we will never see a blockbuster film about out-getting; I just can’t see one of the major studios in Hollywood calling in their best and brightest stars to do an Academy Award-type production about out-getting. Moneyball was so transformational, and it wasn’t just about on-base percentage. Moneyball was transformational because it represented a different way of analyzing the game — bringing Bill James’ sabermetric concepts, finally, to the big leagues. And it was about answering a couple of key questions. One, what attributes are most associated with winning baseball games? And then, how does the Major League Baseball market value those attributes? It just so happened in that seminal season of Billy Beane’s Oakland A’s in 2003, that at that time, on-base percentage was one attribute that was not overly valued by Major League Baseball clubs, so they used that as a competitive advantage to win.

The point that I’m making is that Moneyball was a significant philosophical shift in the way that the game was played and managed. And I think out-getting, while an important strategy for all the data that you suggested, cannot rise to the level of Moneyball for the very fact that the same level of Moneyball thought process — that analytical thought process — is actually what’s driving out-getting here as well. How do you win ball games? And how do you marshal resources in order to win those ball games more efficiently? It actually stems from the Moneyball concept, so there’s no way that it can ever overtake it in terms of influence in the game.

The second point that I wanted to make is that out-getting exists because of a team’s lack of quality starting pitching. It is a solution to a problem that teams have. And let me cite the Tampa Bay Rays here yet again. And, in fact, Tampa put out Sergio Romo, relief pitcher, as an opener this season, as an experiment, got a lot of attention, lot of interest — but keep in mind that the Rays also are the same team that had Blake Snell this year, who won the Cy Young Award. And, in fact, Snell was worth more wins above replacement than the next six pitchers on the Tampa staff combined. And, in fact, if they could have started five Blake Snells, they would have never used an opener! A good starter, despite not pitching in high-leverage situations, will still have the highest context-dependent win probability added. And Snell, in fact, was 10th on the team in average leverage index, but first in WPA by a mile. The point that I’m making is that even for a team that used an opener and experimented with this philosophy, it’s not like they limited the amount of innings that Snell could pitch. They still pitched him — because he was a great starting pitcher.

The other point that I wanted to make about starting pitching, Paul, and it goes back to the point that you were making about the business side of this: The reason why some starting pitchers get paid more than relief pitchers is because they’re more valuable. They are more efficient, they throw more innings, and they earn the salary that they are paid. You can’t tell me that a team would say no to Max Scherzer when given the opportunity to sign him in his prime. He’s just too valuable.

The final point that I wanted to make is that out-getting will lead to an exacerbation of the pace-of-play issues within Major League Baseball, and even if it did rise to this level of transformational impact within the sport, it’s also going to mean longer games. The pitching changes are just one variable in playoff baseball that makes the games interminably long. This is the first year that even Chicago Cubs manager, Joe Maddon, who we explored with Jayson Stark in a previous episode of this show, [was] saying that even from [his] perspective the games are getting long. I think that Major League Baseball is looking at a wide variety of tactics in order to improve the pace of play. Putting in a constraint about the number of relief pitchers, I think, is one easy way to help cut down and improve the pace of play. And constraints actually can add more interest to the game. Think about the constraints around how many soccer players are available as substitutions in a soccer game. There are constraints around pit stops in Formula One — Formula One teams can stop probably no more than two times or else they’ll be left in the pack behind the rest of the racers. The point that I’m making is that there are constraints that can actually add to the strategic interest of the game that would not only satisfy hardcore fans but also make the game more easily consumable for casual fans.

My final point, as I yield the floor to you, Paul, is that Moneyball is about regular-season optimization to hopefully get you to the postseason. I’m not sure if we’re going to see out-getting as a sustainable strategy throughout the regular season, especially when it’s important to keep pitchers’ arms healthy and ready to go on a more consistent basis. In the playoffs, we see out-getting quite a bit because, quite frankly, it’s an all-hands-on-deck situation. I’m not so sure, in a league with 162 games, that an all-hands-on-deck situation for each of the games is realistic. So, that’s my defense as to why out-getting is an important strategy. I can’t argue with the fact that hitters in the first inning are a dangerous lineup and that over time, especially when pitchers see hitters for a third time, they’re vulnerable. I can’t argue with that data. But what I can suggest is that within the context of this debate, Moneyball is a false comparison, out-getting exists because of teams’ lack of starting pitching, and then finally, out-getting’s life as a strategy may over time be curtailed because of its impact, or because of its potential impact, on pace-of-play issues within the game.

Paul Michelman: All right, then — well argued. Fraught, and I’ll explain why, but well argued. Let’s take your counter to my counter, one point at a time. (See what I did there, kids — Counterpoints?) The first is that Moneyball is a false comparison. We didn’t say that out-getting would be as impactful as Moneyball. The argument isn’t that out-getting will be as transformative as Moneyball, it’s that it will be as transformative on pitching strategy as Moneyball was on offensive strategy. And so, I’m going to affirm that that’s not at all a false comparison. It’s a very fair comparison. At the core of your argument that it was a false comparison was the basis of Moneyball, which was identifying the attributes that were key for success, right? And valuing those attributes in a new way, which is precisely what out-getting does for pitching strategy. Moneyball said runs are what matters, and the market is incredibly inefficient in valuing how runs are produced. Out-getting says getting outs are all that matters, and the market is really inefficient in how it goes about achieving those outs. False comparison? No — that comparison is right on.

So, second argument is that out-getting is a reaction to the dearth of starting pitching. Exactly! We can’t magically conjure more starting pitching. Things have changed. The effectiveness of the average starter has decreased because offense has surpassed it. It is because of Moneyball, and its focus on greater efficiency and offensive production, that we have this perceived lack of starting pitching. We can’t fight that — we need to accept it and react to it, and that is precisely what out-getting does. It says: Yup, we don’t have enough Justin Verlanders and Max Scherzers in the world; we need to adapt. We need to recognize that a team might have one or two guys who can reliably go into the sixth or seventh inning. That’s it. We need to implement a strategy that accepts the reality and compensates for the lack of efficiency that has entered into the game.

Third argument is about pace of play. I’m no big fan of four-hour games. A Red Sox-Yankees postseason game takes two days, it seems like. And yes, the more pitching changes that occur in a game, the longer that game is going to go. But here’s the thing: Teams don’t care. Teams want to win. If they are going to have two or three more pitching changes in the course of the game and increase their chances of coming out with a “W,” they’re going to do it. So, the only way that argument holds is if Major League Baseball steps in. Maybe they will. I don’t see that happening in the near term. I think there are other things that they’re focused on like pitch clocks, for instance, to focus on pace of play. And it would be really hard to regulate. If somebody gets hurt, what are you going to do — make them stay in the game? If they’ve already hit their allotted number? So, I think that’s unrealistic. It is an accurate statement by you that if Major League Baseball did limit the number of pitching changes, that it would throw a damper on out-getting. I’m just not believing that that’s realistic.

And then finally, there’s the argument about regular-season strategy versus postseason. And it is absolutely the case that we have seen out-getting used much more in the postseason than as a full-season strategy.

The road to wisdom begins with a single step. As teams recognize that this strategy works in a short series, they will think to themselves: “Well, gee, this won us the World Series. Why are we limiting it to this particular context?” I’ll go back to the data. The data is clear. Starting pitching doesn’t work. As teams pull back, recognize the value of deploying this type of thinking in the postseason, and they look at the data — they’ll recognize that there is huge untapped potential.

Ben Shields: All right, Paul, very interesting. Appreciate that rebuttal. And you show that you’re a great active listener, which I appreciate as well. I’m going to make a few final comments here. First and foremost, I stand firm that we still don’t have enough data on the experiments that were run this past season in order to make a broad claim about the enduring impact of out-getting as a pitching strategy on the level of Moneyball for offense. We are just not there yet. Perhaps if we come back in two or three years and have this conversation, I will be happily proved wrong if the data does, in fact, support your position. The other point that I’d like to make is that there’s no question that the data is trending toward this being a viable and, in fact, powerful strategy for Major League Baseball teams. But as we’ve seen over and over again, not only in sports but organizations writ large, just because data exists does not necessarily mean it will turn into lasting change. And, for instance, you’ve got a number of different pitchers in Major League Baseball who have been “starting pitchers” — that’s what they know. You’ve got a number of coaches that have “started pitchers” — that’s what they’ve done. And although I am a major baseball fan — love Major League Baseball — we can’t exactly say that Major League Baseball is the most fast-changing, quick to adopt new strategies across the board. And so just because the data exists, [it] doesn’t necessarily mean that it’s going to have significant impact in the way that you suggest. So, I would just put a note of caution on both the fact that we don’t have enough data as well as the fact that just because some of the data is trending in a certain direction, that doesn’t necessarily mean that people and organizations will change their behavior on a wide scale. I still believe that my reasons are salient about the false comparison with Moneyball, about the fact that if teams actually had quality starting pitchers, they wouldn’t necessarily be employing out-getting all the time, and the fact that the pace-of-game issues are a key driver here. I guess we’re going to have to agree to disagree on that point and see how this out-getting strategy evolves in the coming years.

Paul Michelman: Ben, thank you. Very well argued.

Ben Shields: Paul, same to you.

Paul Michelman: And this brings us to the exciting conclusion where producer Mary Dooe, who was only warned that she was going to be put into this position minutes ago, judges who won the debate. Will you dare to dream big Mary? Or will you throw cold water on the argument?

Mary Dooe: Yeah. So, I think, you know, in listening to both of you guys talk, I originally was very convinced by Paul’s arguments because, of course, there was a lot of data — threw a lot of stats at us in the beginning. I also think it’s the kind of argument that’s really hard, because you’re a little bit arguing about the future, and it’s an unknown. And so, of course, there’s not data about the future yet — so that’s kind of a difficult thing to argue. The thing that really stuck out to me is this idea of transformational. And the idea that Moneyball really was transformational in baseball and whether something else is going to be transformational in the same way. And I think about this a lot. In other mediums and other realms of the world, things that are truly transformational sometimes come out of nowhere, right? So, you think about the iPhone. I was thinking about in the world of podcasting, right? Serial came along and was this transformational podcast, got a lot of new people listening to podcasts, and then after that everyone kind of wanted to be the next Serial. And none of those people that tried to be the next transformational thing really ended up being the next transformational thing. So, I was thinking about that a little bit in terms of strategies that are sort of in the same realm of Moneyball in terms of being a little counterintuitive or thinking about baseball in a different way. And I think to that end, I slightly buy Ben’s argument a little bit more, only because I think it is too soon to tell. And usually when things make that big of a difference, they kind of sneak up on you, and you can’t kind of tell from one season of data. So sorry to say, Paul, your arguments were great. I loved your stats. I’m going to give the win to Ben.

Paul Michelman: Thanks, Mary. This has been Counterpoints, the sports analytics podcast from MIT Sloan Management Review.

Ben Shields: You can find us on iTunes, Google Play, and wherever fine podcasts are streamed. If you enjoy Counterpoints, please take a moment to rate and review the program on Apple Podcasts. And tell your friends while you’re at it.

Paul Michelman: Counterpoints is produced by Mary Dooe — for now. Our theme music was composed by Matt Reed. Our coordinating producer is Mackenzie Wise. Our crack researcher is Jake Manashi, and our maven of marketing is Desiree Barry.