Counterpoints /

The Sports Analytics Podcast from MIT SMR

Buyer Beware! MLB Free Agents Underperform Their Contracts

play Listen
« Prev Full Series Next »

Almost every team’s fans can point to a high-profile, free-agent signing that fell flat. Sometimes the reasons for the player’s underperformance are clear — injuries, for example. But often, it’s harder to pinpoint the cause. In those cases, we tend to point to lack of effort as the culprit. This is what researchers call shirking. According to Richard Paulsen, who presented his paper, “New Evidence in the Study of Shirking in Major League Baseball,” at the Sloan Sports Analytics Conference, “Shirking occurs when an employee exerts effort ... that is suboptimal in the eyes of the employer.” Paulsen speaks with us about his research on shirking and his belief that the Phillies are going to regret at least one free-agent signing from this off-season.


Ben Shields: After missing the playoffs for the third consecutive season, the Seattle Mariners needed to make a splash in the 2004 off-season. To bolster the American League’s worst offense, Seattle looked toward free-agent third-baseman Adrián Beltré. Beltré had been an average hitter over his first six seasons with the Dodgers, but he had the breakout of all breakouts in ’04 — smashing 48 homers with an OPS of over 1,000. The Mariners, dazzled by those gaudy stats, handed Beltré a hefty, five-year contract as the focal point of their lineup. And Beltré promptly went right back to being an average hitter, never achieving a .300 average or 30 home runs at any point of the deal. Though Beltré later went on to cement a Hall of Fame career elsewhere, his time in Seattle was a cautionary tale for the contract year player.

Paul Michelman: Which brings us to 2009: Having now missed the playoffs for eight straight years, the Mariners were in no position to proceed with caution. With an infield spot open after Beltré’s departure, Seattle handed a four-year, free-agent contract to Chone Figgins, who — stop me if you’ve heard this one before — was coming off the best year of his career, piling up a robust 7.7 wins above replacement for the Angels. And what happened when Figgins settled into his new home in the Pacific Northwest? Well, he delivered WARS of 1.2, -0.9, -1.2, and 0, which was not actually an improvement, because he was released just three years into the deal. The Mariners had once again been burned by an underperforming free agent, and their playoff drought continues to this day. At 17 years, no team in the big four professional sports has gone longer without a playoff appearance. You could say the Mariners lost the battle and the war. 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: Buyer beware — the evidence of shirking amongst MLB free agents is real.

Stories of disappointing free agents are as old as free agency itself. Almost every team’s fans can point to a signature signing that fell flat.

Paul Michelman: Sometimes the reasons for the player’s underperformance are apparent — injuries and age are the most obvious — but often it’s harder to pinpoint the cause. And in those cases, we tend to point to effort, or lack thereof, as the culprit.

Ben Shields: This is what researchers refer to as shirking. According to Richard Paulsen, who presented his paper, “New Evidence in the Study of Shirking in Major League Baseball,” at the recent Sloan Sports Analytics Conference, shirking occurs when an employee exerts effort that is suboptimal in the eyes of the employer. Yes, fans of (insert name of team here), it does sound a lot like (insert name of player here). Here’s Richard speaking with Paul about his research on shirking and his belief that the Phillies, in particular, are going to regret at least one free-agent signing from this off-season.

Paul Michelman: Richard, welcome to the podcast.

Richard Paulsen: Thank you.

Paul Michelman: So in the introduction we define shirking as you did in the paper: when an employee exerts effort that is suboptimal in the eyes of the employer. And so intuitively it’s easy to believe shirking exists — I think every fan believes shirking exists on his or her own team. But we’re interested in the evidence. So let’s begin with: What data did you look at?

Richard Paulsen: Sure. So I looked at a sample of Major League Baseball player data. I’m looking at player-year observations for the years 2010-2017 — looking at all Major League Baseball players: pitchers and hitters. The measure of performance that I’m using for the players is wins above replacement. Wins above replacement essentially looks to quantify player performance in his contribution to team winning. Now, as for the impact I find of these multiyear contracts on performance: On average, I find that being in the early years of a contract, or having two or more years left on the player’s contract, leads to a fall in wins above replacement of about 0.3. Now, to give a little bit of an idea of how large that impact is: The average player in my sample has wins above replacement of about 1.5, so it’s about 20% relative to the average.

Paul Michelman: And from that data, how did you make the case that shirking exists?

Richard Paulsen: When I first do those, I look to run a regression. I’m basically trying to use the data to answer the question: When a player has multiple years remaining on his contract, does that player perform below expectations? So I’m trying to control for everything I can about the player to try to isolate an impact of how many years are left on the contract on performance. In doing so, I control for things like the player’s age, to capture that natural decline as a player ages. I also control for things like how long the player has been in the League. I even try to control for players’ specific characteristics. And what I end up finding is that when players have multiple years remaining on their contract, they perform worse than if they’re in the final year of their contract.

Paul Michelman: And I’m assuming this only applies to post-free-agent eligibility, is that correct?

Richard Paulsen: Yes, I’m looking at a sample of players that are generally post-free-agent eligibility.

Paul Michelman: OK, so your focus is on the player’s performance, and that makes sense — but, of course, it takes two to tango for shirking to exist as you defined it in the paper. So how did you account for what the employer’s expectations were?

Richard Paulsen: One approach I took — I run what’s called a fixed effects regression, where I’m controlling for the player. So, what I’m essentially saying is: If we keep the player fixed, how does variation in the number of years left on his contract impact his performance? Now, the second approach I take is to control for the player’s salary under the assumption that the salary is some estimate of what the employer is expecting out of the employee or what the team is expecting of the player.

Paul Michelman: So talk a little bit more about that. You’ve run some kind of analysis that says: If the player is at this position — the player is a shortstop and the annual salary is $14 million — then the expectations equal question mark.

Richard Paulsen: Yes, so I’m not necessarily estimating the expectations, but rather what my finding is saying is that if we hold constant these various players’ characteristics, if the player has more years left on a contract, he’s going to perform at a lower level.

Paul Michelman: You mentioned that you controlled for various factors, age being probably the most obvious one. But I’m also interested in how you controlled for environmental factors: changing home ballpark, changing league, being a part of a different lineup, a different manager, or even a different team philosophy. How do you account for that type of factor?

Richard Paulsen: So, there are other potential explanations here that could drive this inverse relationship between years remaining on a contract and performance. One of them, that you allude to there, is it’s possible that when a player joins a new team — when he’s signing a new contract, he might be joining a new team. To address that possibility, I control for whether or not the player’s in his first year with the new team. I’ve also run regressions where I control for how many seasons the player’s played with that specific team. And then I’ve also controlled for team-specific effects as well.

Paul Michelman: Richard, did you find any differences in the likelihood of shirking in, say, average players versus star players?

Richard Paulsen: In this paper, I don’t break the sample up based on whether or not a player’s average or a star player. I have done analyses that might suggest that shirking is more driven by average players than star players. But I don’t want to say definitively whether or not I think it’s one group or the other.

Paul Michelman: Fair enough. There’s some future research we’ll be interested in there. And how about age? We know that you’re accounting for age as a factor in your analysis of shirking, but is the likelihood of shirking impacted by a player’s age?

Richard Paulsen: There is a paper in the literature — I want to say it’s by Berri and Krautmann, I think 2009 — something that they do is they look to say: If a player’s nearing retirement, are they more likely to shirk? And they generally find that the answer to that is yes. Now, when we talk about player incentives, I like to think that there are a few things that are going on in the player’s mind. One thing is that players in Major League Baseball have long-term, guaranteed contracts. So, no matter how they perform in a given season, they’re going to make that money. And if they’ve got a contract that’s five, 10 years in length, they’re guaranteed that money essentially, no matter how they perform. Now, in terms of incentives for the player to perform at a high level, most players want to sign a new contract at the end of their current contracts. Players might also have a love for the game, they can be shooting for endorsement deals, maybe they’re looking to make the Hall of Fame afterwards — we’re talking about a star player. So how retirement might come in there is that if a player’s got a contract that’s going to bring him near retirement, that player really doesn’t have an incentive to play hard for the next contract. Right? So if we talk about this Bryce Harper contract — 13 years, $330 million — he’s going to be 39 years old at the end of this contract. He’s not playing for the next contract in his career at this point, which kind of disincentivizes his effort.

Paul Michelman: But in the case of Harper, might he be — still, at the tender age he is at, he is building a pretty impressive career. One would think, and I’m just going to allude back to a point you just made, that he has the Hall of Fame in mind. Is that sufficient incentive to keep him kind of going at full speed for the full 10 years of the deal?

Richard Paulsen: You know, it might be. And I also mentioned endorsements when we were talking before. While he’s not playing for higher current salaries, necessarily, if he is playing at the highest level, he might have sponsors lining up at his feet wanting him to advertise various products. Right? And if we talk about famous athletes, a guy like a Tiger Woods in his prime was making more money from endorsements alone than he was from his playing. And he was making $30-40 million from playing.

Paul Michelman: So we mentioned Bryce Harper. I think you have a point of view on another free-agent contract that Philly executed in the off-season. Correct?

Richard Paulsen: Absolutely. The contract that Philly signed during the off-season that I feel pretty strongly that I don’t like is the Andrew McCutchen signing. Let me give you a little bit of background on the McCutchen signing. It’s a three-year deal, totaled at $50 million. He’ll be making $10 million in 2019, $17 million in 2020, and $20 million in 2021. And then this contract has a club option for the 2022 season that would pay him $15 million if the Phillies execute it, or it would pay him a $3 million buyout if they don’t. Now, as for what my work might say about the McCutchen contract: The paper that we’ve been discussing that I presented at the MIT Sloan conference essentially comes to the conclusion that multiyear contracts disincentivize performance. And in this case, you have a guy (Andrew McCutchen at 32) [where] we’ve seen the decline in his performance over recent seasons relative to his peak in his late 20s. You give him a multiyear contract [and] that’s going to disincentivize effort in his early years. And I have another paper, in the Atlantic Economic Journal, that looks at contract options. And essentially what that paper says is that if a player has a club option, that’s also going to disincentivize effort. From the team’s point of view, they’re going to exercise that option if he’s performing at a very high level. So he can’t reap all the benefits of that good performance, if they exercise it. If they don’t exercise the option, he gets paid a buyout anyway, and he can go sign with another team.

Paul Michelman: So are club options always kind of a lose-lose scenario?

Richard Paulsen: The benefit [of] club options and the reason that teams might want to include a club option is that if, for example, a player gets hurt during the duration of their contract, they might execute that club option at the end of the contract if the player’s doing well. And they could potentially get a benefit there. You know, if it’s the case that McCutchen performs very high over the next few years and he’s going to be worth $30 million in 2022, they’ll exercise that club option of $15 million. And they’ll get a good surplus from doing so. So the team is going to put the club option in place with the hopes that it protects them against, say, injury or falls in performance for the player. But they should worry about the potential for that disincentive effect. Because the player can’t reap the full benefits of that contract and they get a buyout even if they don’t get the contract, that kind of takes away some of the incentive to perform.

Paul Michelman: In this case, is the size of the club option really an issue? Not so much the buyout? If you’re looking at McCutchen, he’s going to be making $20 million the year before the club option. Right? And then $15 [million]. Depending on kind of where the player sees himself, McCutchen could use that $20-million year as a contract year and really go all out, because he wants the $15 million. Or he could do the inverse, right? So I guess some of this depends on where the market is?

Richard Paulsen: If he decides to go all out during that contract year when he’s making $20 million, then the Phillies are going to potentially execute the option, get him at $15 million for 2022 when the market might value him more highly if he performs really well in 2021.

Paul Michelman: Right. I guess you’d be ascribing a lot of strategic thinking to the player if they were really trying to manipulate this one way or the other. One has to wonder how often one is kind of really digging into that type of thinking. Although, you know, maybe agents do.

Richard Paulsen: Yeah, I mean it’s certainly a question, right? You know, one point I think that there is to make in this work, generally, is that players — Major League Baseball players — they’re playing a game that they love. If these guys are shirking — and my evidence might suggest that they are — we could imagine that in other employment settings, shirking by employees might be quite a bit worse if you don’t really love your job.

Paul Michelman: Sure, I think underperformance is a massive problem in the world of organizations, so it’s interesting that you make that parallel. But there is a huge difference: Very few employees can be set for life at the age of 35 as most professional athletes are (and probably a lot younger than 35). And very few employees work on long-term contracts, right? Most of us are employees at will. Nevertheless, I think it is certainly a valid observation. You’ve been alluding to the... practical implications of your research throughout the interview, but for the general managers in our audience today, let’s kind of put a fine point on what Major League [Baseball] front office people ought to be doing with your findings and other research on shirking.

Richard Paulsen: What I would really like to see is more use of incentive pay within contracts. Now, if we take a look at the Bryce Harper contract: If he wins League MVP, he gets $500,000; second place in the voting gives him $50,000; third place gives him $25,000. He’d get $100,000 for a World Series MVP; $50,000 for an NLCS MVP; $50,000 to make an All-Star Game, get a Gold Glove, or [get] a Silver Slugger. Now... that sounds like a lot of money to you and I — $500,000 for an MVP — but if we take a look at the value of this contract as a whole, in any given season he’s making about $26 million. If you’re Bryce Harper and you’re thinking: “I’m making $26 million a year, but wow, I could make $500,000 more to win MVP,” is that really going to be the incentive you need to win the MVP? And if we look at some of these smaller pieces of incentive pay, is he really going to manipulate his effort to perform at a higher level to make an All-Star team for an extra $50,000? No. So, what I might like to see in these big, long-term Major League Baseball contracts is an even greater use of incentive pay. Rather than give him $500,000 for a League MVP, why not give him $5 million?

Paul Michelman: Is there kind of a percent that you think will be the tipping point? So $500,000 on a $26 million contract is just under 2%. (And no, I did not do that in my head — I have my iPhone calculator open.) But I’m wondering if you think there is kind of a magic percent, like if I can make 25% more, do I need to make...? What is that number?

Richard Paulsen: I don’t know if there’s a magic number, but I think if teams start to play around with this, we’d have the data to try and answer that question.

Paul Michelman: So in your research — we’ve talked about the couple of Philly contracts — but are there any kind of particularly spectacular cases of shirking that are worth calling out?

Richard Paulsen: I like to point out the contract given to Pablo Sandoval when he joined the Red Sox a few years back. (He’s since been traded to San Francisco Giants.) I also like to point to the Jacoby Ellsbury contract with the Yankees. Now, in both of those cases, these guys were given contracts that were at least three years in length. These guys were being paid at least $20 million per year. And in the first, second years of those contracts, those players performed at a very low level.

Paul Michelman: Richard, we’ve talked about player incentives. What about team-based incentives?

Richard Paulsen: Sure. When we think about firms in general and the economy, we often assume that the goal of the firm is to maximize profits. Now, it’s the case in sports that teams are looking to potentially win games, but these teams are firms, so profit should be one of their motives as well. So when we talk about performance of a player like Bryce Harper, we could think about how his performance might leave the team winning. But when we talk about whether or not his contract is a good contract for Philly, it’s also kind of important to think about [the fact that] they’re hiring him not just to help win games but also to draw in fans. In terms of drawing in fans, a star like Harper might be able to draw fans into games even if he’s not performing at the highest level.

And to expand on that a little bit, there was another paper at the MIT Sloan conference that was looking at the economic impact of NBA superstars. And what they were looking at was how when the superstar in the NBA doesn’t play a game — a couple of days before, a couple of hours before, it’s announced that the player’s not going to play — how does that affect ticket prices for that game on secondary ticket marketplaces. And they found that these guys, simply by not being there playing in that game, led to a $9-$25 fall in the average ticket price. So we could imagine with a guy like Bryce Harper, even if he’s not performing at a high level and contributing to wins for the team, he might draw fans to the game just through his star power alone.

Paul Michelman: And how do we then factor that into the kind of deal we’re going to offer him?

Richard Paulsen: So that’s a difficult question to answer. And this is a question that has been thought about for years in economics and even in baseball. The first paper that looked at this issue of estimating player contribution in dollar terms was a paper by Gerald Scully in 1974. Now, what he was looking to do is estimate, as I said, how does player performance translate to dollars for the team. And in doing so, what he first did was he asked the question: How does performance impact winning? And then how does winning impact revenues for the team? And now this concept has been expanded over time — a number of authors have looked at this sort of approach to doing so. But the challenge in doing that, of course, is that while we can figure out maybe the dollar value of a win, it’s kind of hard to quantify the dollar value that the team players bring to a team outside of winning. You know, we might be able to look at things like jersey sales, Twitter followings, all those sorts of things — but really where future work could go would be to look at, say: Can we quantify how Bryce Harper contributes to attendance in the absence of team winning?

Paul Michelman: You know, it’s really interesting. And it’s kind of tempting to think that teams could be overestimating that now as a way to justify making what could be seen as an irrational offer to a player. Right? You can fall in love with a Bryce Harper and desperately want him on your team and in order to justify the whatever (330? Is that what the number is?), whatever-it-is-million-dollar contract by just assuming or telling yourself that he is going to have this outsized impact on the bottom line. It’d be really interesting to see those numbers at some point. So, on that note, what’s next in your research? What data do you wish you had that you don’t yet have?

Richard Paulsen: So I’d really like to try and look into the mechanisms through which players are actually shirking. Now, we can imagine a few different ways through which the shirking manifests itself. One possibility is that the player might shirk by not performing during the off-season — maybe the player’s not spending as much time in the weight room as he should be; maybe he’s falling out of shape. Another possibility is that the player might shirk between games — maybe it’s the case that the player is not watching video [or] not getting scouting reports between games when he has an opportunity to learn about the opposing pitcher. And then, of course, the third way that shirking could manifest itself: The player might not be giving his best effort during games. Now, I think future work that I’d like to do, and that others should be working on, would look to kind of identify where that shirking’s taking place. You know, if I were to talk about kind of dream data that I might want — maybe good biometric data on players, or even good data looking at player height and weight, for example, in the off-season. Or body fat. Some sort of measures of — is there shirking taking place during the off-season? Or, you know, to reference one of the other podcasts you guys have had that looked at the impact of sleep, it’d be interesting to ask the question: Are these guys shirking by not getting enough sleep? Maybe it’s [that] these guys on multiyear deals are out partying the night before a game, they’re not getting that sleep. If we had that good sleep data, we might notice differences in sleep patterns for guys that are on contract years versus non-contract years.

Ben Shields: This has been Counterpoints, the sports analytics podcast from MIT Sloan Management Review.

Paul Michelman: You can find us on Apple Podcasts, Google Play, Stitcher, Spotify, and wherever fine podcasts are streamed. If you like the show, please rate and review us.

Ben Shields: Counterpoints is produced by Mary Dooe. 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.