April 8, 2015
Tools of Ignorance
Pitfalls of the Arb-Year Buyout
Wade Miley, Brian Dozier, Juan Lagares and Christian Yelich are among the most recent round of players to get extensions that cover their arbitration years and not much more. We think we mostly know why these deals happen: teams want to lock in players at below market cost and players want to lock in moneys. The discussion on the benefit to teams mainly centers on the fact that players—being people—are risk averse and overfocused on negative, small-probability outcomes, such as a career-ending injury or becoming terrible. As a theoretical consequence, players accept below-market deals in order to guarantee income.
However, the four extensions listed above did not receive the pro-team praise/anti-labor outrage that past extensions have received. Is this a coincidence? Maybe. Is this just agents and players getting smarter? Maybe. Is this a reaction to an overreaction to the Jon Singleton extension? Maybe (though the author notes that this would be a gross oversimplification of the Singleton situation). Another possibility is that such extensions lend themselves to decision-making errors for teams just as they do for players. More specifically, teams might be overweighting certainty, small-probability outcomes, and positive trends in handing out such extensions.
Management, Budget Planning, and Certainty
Not exactly groundbreaking stuff, but the desire for certainty may be even higher in management than it is in the general populous. Why? A professor once said something along the lines of: “Leadership drives innovation, while management drives compliance.” Now we have no way of knowing which owners allow management to be "leaders" or innovators and which require management to be "managers," but we do know that ownership does place restrictions on the spending of management. What this is all getting at is that hard goals tend to get certain results. Instead of trying to achieve as much is possible, these goals lead to results that could be place in the just-enough bucket. Put differently, if 90-100 earns an “A,” there are going to be far more 90’s and 91’s than 98’s and 99’s.
Put it all to together and nature of management, arbitration, budget planning, and our general behavior toward certainty makes these contracts attractive for front offices for reasons other than getting a player at a below-market rate or an even further below-market rate. Is this enough to believe that these contracts might occasionally be a trap for front offices? Maybe, but we also have some other behavioral evidence to support this notion too.
Prospect Theory for Teams
Moreover, as Zachary Levine points out in “Every Team’s Moneyball: Houston Astros: Luhnow Life Insurance,” teams do not need to be right on every one of these contracts in order for this strategy to be successful. In other words, the more good bets teams place, the higher the return on investment. There is certainly no contesting that. The question, however, is whether these bets are as good as the teams think. Rookie contracts already make teams similar to insurance companies or an investment portfolio, in that each team controls a diversified portfolio of risky assets (an asset simply being something that is useful or valuable). Therefore (and obviously), these contracts need to improve the value of holding these assets in order to be boon to teams.
The question then worth asking then is whether using an actuarial table or similar logic-driven decision-making tool/process precludes teams from being influenced by cognitive biases or common decision-making errors when agreeing to these extensions. There is evidence (given the lack of lopsidedness of these contracts the more data points come in) to suggest the answer is no. While “the kids” certainly have relatively more to lose, the teams still have plenty of negative, small-probability outcomes to overweight. What if the Player X who has shown improvements each of the past three seasons continues to improve to the point that by the end of arbitration we can no longer afford him? What if by not offering a respectable extension, Player X will most certainly bolt after his rookie contract? Also, is this more of a concern for smaller-market teams? Is that why they are more likely to hand out these extensions? It is possible that teams are not overweighting these small-probability outcomes, but given the nature of budget planning from above and how we view sequences and trends from the below, there might be something to this.
The Trends We Believe
Now it should be noted that everything that has ever been put on a slide in school is 100 percent true. But if you still have your doubts, this one is backed by evidence. When polled as to how people would like to be paid the next three years, most choose $40K then $50K then $60K as opposed $60K then $50K then $40K, despite knowing how interest works. In other words, we love seeing positive trends, even if they are only parts of random sequences disguised as trends. When we combine that with our tendency to frame negative outcomes as isolated, we get a relatively believable reason as to why these pre-arbitration extensions occasionally fall into the not-so-lopsided bucket despite all the advantages teams have over players. By mistaking sequence as trend (first bullet point) and the relevant as outlier (second bullet point), and when combined with the decision-making traps above, we can see how a team could conceivably convince itself that an extension is more beneficial than it really is.
Maybe every team properly filters out these biases when making decisions regarding these decisions. Maybe these biases are always outweighed by the risk imbalance that Miller discussed above. But given all of these maybes, it seems unlikely. Combine all of these maybes with some learning from players and their agents, and maybe these deals are not always as one-sided as they should be in theory.