To answer the question in the title, I do not know how the Phillies are framing the decision to trade Cole Hamels. We do know, however, how people frame decisions when facing risk and uncertainty—both of which the Phillies face with Hamels. Our risk-averse nature often drives the way we make decisions, but the way we frame decisions and consequent risk can often lead us to very different decision outcomes. Depending on how the Phillies frame their decision, both trading him now (ensure they get something back) or waiting for an elite return (ensure that they “get this one right”) can be viewed as the risk-averse decision. The hope (for Phillies fans) is that the Phillies are framing the decision simply as trying to maximize their return while taking all risks into account. That said, we will take a look at the ways people err when facing similar decisions and whether the Phillies are (or appear to be) falling into similar decision-making traps.
The Problem with Fuzzy Probabilities
Before we go any further it should be noted that we are starting with a premise that Cole Hamels must be traded by the Phillies eventually. An analyst or two out there might disagree with this premise, but given the Phillies' place in the competitive landscape, Hamels is worth less to them than to almost any other team in major-league baseball. So, given this starting point, questions like: How valuable will Cole Hamels be over the next four years? How valuable will the prospects get in return be? are irrelevant. All that matters is whether the Phillies can get a better return for Hamels in the future (spring training, offseason, etc.) than they can get for him now; and whether that better return is worth the risk inherent in waiting.
The issue with figuring this out is that we do not know the answer and we certainly do not know the answer with certainty. In order to know, we would need to know the probability of the demand for Hamels increasing or decreasing. To know that, we would need to know the odds of Hamels sustaining a significant injury, the odds of Hamels performing in a way as to devalue himself as an asset to other teams, and the odds of another team or teams having increased demand for Hamels in the future. As we can tell, we now have further issues. Not only do we face an uncertain future in this decision, we do not even know the odds we face. In blackjack, we can know whether we made the right decision when facing a probabilistic future because we know the odds, because we are dealing with risk. In baseball, however, we can never know if we made the right decision when facing a probabilistic future because we do not know the odds, because, rather than dealing with risk, we are dealing with uncertainty. Put differently, in determining when to trade Hamels, the Phillies are not facing known unknowns, they are facing unknown unknowns.
This is a pretty dramatic for a basic description of literally every baseball decision, but the point being made is important. When faced with uncertainty (unknown unknowns), we tend to let our best decision-making processes fall off, while allowing our biases to creep into our decisions (and our analysis of other people’s decisions). The Phillies should be trying to use the information at hand to take their best guess at starting pitcher injury rates and the seasonality of starting pitcher trade markets, as well as specific information about Hamels and specific information about the current starting pitcher trade market (as well as many other factors I am sure I am not thinking of). Should, however, is much easier to say than do, so let us take a look at some of the decision-making traps we frequently see to see if they are afflicting the Phillies.
Both Sides of the Disposition Effect
The disposition effect is a theory of behavioral finance that explains the tendency of investors to keep their losers too long and sell their winners too quickly. The disposition effect is partially explained through loss aversion. Selling a winner is a way to ensure that we get paid for our good decision making and risk taking. In this sense, if the Phillies considered their contract with Hamels a winner, then they would be eager to sell him in order to make sure they turn a profit.
Clearly, this is not happening with the Phillies and Hamels. (Check one box for sound process for RAJ.) However, this where decision framing comes into play. The Phillies could also be looking at having to rebuild and “having” to trade Hamels as a “loser” and thus be holding onto him too long in the hope that he is able to provide some kind of unreal return that will make up for previous poor decision making. The interesting part within the disposition effect is that it shows our tendency to place value on being right and to take sunk costs into account as opposed to only caring about making the best decision. If the Phillies are weighing sunk costs or are looking at the decision to trade Hamels as way of justifying past decisions, then they have erred in the same way investors tend to err in the selling (or not selling) of stocks.
There is evidence, however, that suggests that the Phillies are not doing so. And this evidence has to do with Roberto Hernandez, Antonio Bastardo, Marlon Byrd, and, most of all, Jimmy Rollins. The fact that the Phillies have been willing to trade these players shows that they can see the writing on the wall, and while their process might have been plagued by the disposition effect in the past, they are at least no longer clinging to sunk costs. Could Hamels be the exception? It is possible, but given the rest of the decisions the Phillies have made, it appears unlikely. (Check another box for sound process (in pencil) for RAJ).
Both Sides of Prospect Theory
Now working under the presumption that the Phillies are not afflicted by the disposition effect (which we cannot know with certainty (lol)), we can now also work under the presumption that the Phillies are holding onto Hamels because they think they can get more for Hamels later than they can now, and because they think that the better return in the future is worth the risk of not dealing him now. We can guess at what the reward might be (obviously, it is a return that they value more than the current offers they are receiving), but we cannot know how valuable the reward is relative to the current offers because we do not know the current offers. What we can take a look at, though, is what factors the Phillies might be underrating or overrating when taking this uncertainty into account. Specifically, we can look at prospect theory, a theory of behavioral economics that explains how expectations influence our decisions when risk or uncertainty is involved, in order to see if the Phillies are allowing expectations to get in the way of their best decisions.
For starters, when faced with multiple options that exceed our expectations, most people choose the safest option as opposed to the most valuable option. Put differently, if we expect a return on a $100 investment, then we are more likely to choose Option A, a 100 percent chance of a $110 return, over Option B, a 90% chance at a $120 return and a 5% chance at a $60 return, even though the latter has a slightly higher expected return. This is clearly not happening with the Phillies and Hamels. If the current and future options both exceeded their expectations for their return on Hamels and they were framing their decision through expectations, then they would have already accepted an offer, as it carries the least amount of risk.
Alternatively, if the Phillies view the current returns being offered for Hamels to be below their expectations, then this could cause more risk seeking in their behavior, which could be problematic. Why? Because this would mean that their decision is being driven by expectations rather than being driven by wanting to make the decision that will return the most value. Why is this problematic? Because when we frame decisions in this manner, we tend to choose the option with the best chance of beating our expectation rather than the option that is likely to return the most value. This is why bettors who have lost at the tracks all day make riskier bets on the last race of the day as they try to get back to even. It is not difficult to imagine that this is what the Phillies are doing; passing on what might be the best deal (once risk is factored in) and holding out hope that a better deal, one that at least matches their initial expected return for Hamels, comes along.
Analyzing the Phillies and the Focusing Illusion
Again, it is easy to imagine the Phillies having their decision-making process negatively affected by prospect theory, but there is no real way of knowing because we are not in the room. And again, like our analysis of the disposition effect, the Phillies should probably get the benefit of the doubt given their recent, seemingly rational actions during their rebuild. However, given the Phillies actions prior to this rebuild (the actions that resulted in this rebuild being necessary), it is easy to see why so many people are not giving the Phillies the benefit of the doubt.
Our minds are great at pattern making, but often they are too good, making patterns that do not exist. In this instance, it appears that we are suffering from the focusing illusion by focusing too much on the fact that it is Ruben Amaro Jr. making these decisions. Does RAJ deserve as much benefit of the doubt as other general managers? Probably not, but that is not the point. The point is that, by acknowledging that he doesn't deserve the benefit of the doubt, we are probably giving him even less benefit of the doubt than he deserves. The person who coined the term focusing illusion, Daniel Kahneman, explains it as follows in Thinking, Fast and Slow:
Nothing in life is as important as you think it is when you are thinking about it.
(Now I have entered a bit of logic-wormhole here as I am thinking about the focusing illusion and therefore overweighing its importance, but let us assume I have adjusted for this bias). Ultimately, the decision the Phillies face is filled with uncertainty and it is filled with even more uncertainty to us, the onlookers. Just as the Phillies need to caution against filling this uncertainty with decision-making biases (such as framing their decision through expectation), we (the analyzers) need to be wary of filling the uncertainty we face in analyzing this decision with our own biases.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.Subscribe now
-trade now (during 2015 offseason)
risk: market is saturated with sp options
reward: shiny new toys
-trade later (during season or future offseason)
risk: injury/performance lowers value, fewer buyers, market correction means less is being spent on SPs
reward: market is traditionally willing to spend more on SPs during season
risk: lots of money to spend on an unhappy guy to stay on a losing team
reward: if the phillies expect they can be competitive in 2017, they have a good LHP under contract
Love these articles. They are a big reason why I subscribe here. Absolutely fascinating and relevant
Risk aversion is simply preferring certainty to uncertainty, and being willing to take less in expected value in order to get more certainty. It's an entirely rational thing to do, in the economic sense of the term.
What you describe is loss aversion. Unlike risk aversion, loss aversion is not rational (again, in the economic sense of the term) and reflects an individual making an error of economic logic, in particular failing to accept past decisions as sunk costs and failure to make decisions at the margin.
The classic example is the coin-flip bet in slightly different circumstances. In version A, the game is I offer you a bet in which you win $2 for heads, win $0 for tails, or offer you the chance to collect $0.99 rather than take the coin flip. In version B, the stakes are reversed- you pay me $2 for heads, pay $0 for tails, or just pay me $1.01 and avoid the flip. A risk-neutral actor will take the coin flip both times to maximize expected value. A risk-averse actor will take the fixed payout both times. A risk-averse but also loss-averse actor will behave inconsistently however- he will take the fixed payout when he is in line to receive money but will go with the coin flip when he is potentially on the hook for paying it. The loss aversion is that he is more interested in the 50% chance of "not losing" than he is in aligning his decisions with his preferences regarding risk level and expected value.
Risk aversion is considered economically rational because of diminishing returns. Add a few zeros to the numbers above and one will easily understand- If the choice is over $100k instead of $1, the first $100k will change your life a lot more than the 2nd $100k, and so giving up a little bit in expected value in order to get certainty makes sense. On the reverse side, (assuming for the sake of argument that wiping it with bankrupcty is not an option) the 2nd 100k of debt is way worse for you than the first 100k is. What is irrational is loss-aversion, because it treats situations identical at the margin as somehow different. Behavioral research shows people have a pretty strong tendency to be instinctually loss-averse rather than risk-averse (e.g. the investment behavior study mentioned), but (on the good side) that people can and will realize the irrationality when it is pointed out to them and adjust behavior accordingly
"Option B, a 90% chance at a $120 return and a 5% chance at a $60 return"
That only adds up to 95%. How about:
Option B, a 90% chance at a $120 return and a 10% chance at a $30 return
BTW, another way to describe loss aversion is that the Phillies are trying to postpone regret: they will be unsatisfied with whatever return they're likely to get so they postpone it as long as possible to maintain the thin hope that they get lucky.
I appreciate the economics primer, but, as is often the case with Economics analysis, is it possible that you are giving too much weight to assuming the Phils are 'rational' actors? How do the numbers account for other factors, such as the message to the fans they'd send by trading Hamels? The Yankees BP annual piece effectively quantified Jeter's value to the 2014 Yankees baseball operations and gate, and there may be similar cost to the Phillies. If we are analyzing Amaro's considerations, how much should we consider his personal philosophical prejudices vis-a-vis these terms you outline above? Or, would you argue that you've already addressed with the Rollins trade? One could argue that dealing Hamels would be a tipping point that the once-proud Phils can't stomach, no?
I don't think we have any way of knowing if they are 'rational actors'. My hope was to take a stab at different ways they could be looking at the decisions and what the consequences of them looking at the decision in those ways. Similarly, I don' think we know Amaro's "personal philosophical prejudices," so we can't know how they are affecting these decisions. We know poor decisions were made previously, but the recent decisions seem better. There is certainly a chance, maybe bc Hamels's contract is long enough that he should be around when they should hope to be competing again, that they are having more difficulty letting go of him. But that said, the simplest, most reasonable explanation seems to be that the offers they have been receiving are not very good, at least not very good compared to what they anticipate future offers to be. Could they miss? Sure, but that doesn't mean they aren't playing the odds correctly.
Come the trade deadline, another team might deal young talent or prospects, as the A's and Royals did last year, but even those teams were not looking to take on multiple years of a top tier salary. And a team looking to add an ace will likely have other options available at the trade deadline to compete with the Phillies' price for Hamels.