The Irrationalities of Baseball Fandom

Discussions of the modern game of baseball can appear sort of like alphabet soup with numbers mixed in. The delights and discoveries of advanced statistics have irreversibly revolutionized baseball, transforming the mentalities of baseball teams and the composition of their staffs. Still, the heart of the game beats for its players, and much of baseball’s lasting appeal lies in the relationship between fans and their favorite players and teams. With games virtually every day during the regular season, baseball cements itself into a permanent fixture of your life. Being a baseball fan – even a casual one – is a part of your lifestyle, and thus, the irrationalities that are innately human travel over into our perceptions of the sport.

Since the 1970s, there’s been substantive development of behavioral sciences, brought on by the ground-breaking research of Daniel Kahneman, who won a Nobel prize for his work, and his late colleague Amos Tversky. In his monumental book Thinking, Fast and Slow, Kahneman offers a comprehensive view of the cognitive fallacies and rules of thumb, known as heuristics, that affect our everyday lives. But, since baseball isn’t mentioned in the nearly 500 pages of his book, let’s look at how the heuristics apply to fans of the game.

Thinking, Fast and Slow is based on the notion that we all have two systems of thinking, aptly named System 1 and System 2. System 1 is fast and largely unconscious; it’s used for automatic processes such as reacting to a bad call by an umpire or leaning away from a wayward fastball. In contrast, System 2 is analytical and slow. It helps us dig through databases and navigate post-game traffic, and engaging System 2 requires a noticeable increase in mental effort. The brain continually tries to minimize its effort level, so when System 2’s logical capabilities aren’t needed, it defaults to System 1. Unfortunately, shortcuts can result in consequences: System 1 jumps to quick but sometimes dubious conclusions, giving rise to the heuristics that make all of us at least a bit irrational.

These irrationalities are what separate us from Econs, a term coined by Kahneman and Tversky to describe the completely logical rational agents assumed by classical economics. At risk of some confusion, let’s refer to Econs’ baseball counterparts as Players – a perfectly baseball-minded species that always knows when to swing, has total control over its pitches, and never commits those errors that find their way into YouTube highlight (lowlight?) reels.

Players

Wouldn’t it be fun if all players were Players?

Well, probably not. That would obliterate the point of the Hall of Fame and All-Star games. It also raises questions that I can’t start to wrap my head around: if both pitchers and batters are Players, would there ever be runs? If not, how would the winner of a game be decided? And, except for baseball’s die-hard fans, would anyone want to watch scoreless innings night after night? Perhaps too much food for thought.

Luckily, Players don’t exist. Instead, we are blessed with a motley assortment of players who, for all their otherworldly talent, are fully human – and their humanity doesn’t stop when they walk on the field.

This idea seems incredibly obvious, of course, when written out, but that doesn’t stop us from treating players more like their Players counterparts now and then. Per Kahneman, the mind is asymmetrical in its treatment of information that’s available and information that isn’t, heavily weighting the former. Because of how difficult it can be to consider all information, System 1 sticks to just what it knows, creating a narrative that doesn’t sufficiently account for the unknown. To quote:

“It is the consistency of the information that matters for a good story, not its completeness. Indeed, you will often find that knowing little makes it easier to fit everything you know into a coherent pattern.”

This concept of “jumping to conclusions on the basis of limited evidence” is known as What You See is All There Is, or WYSIATI. Baseball is a perfect breeding ground for WYSIATI: Outside of the information readily available on the screen or in the stadium, few of us have the opportunity to really get to know the players – therefore, we don’t get a full picture of the human element on the field. What we do have for each player is a comprehensive, data-driven view of his average ability. For the most part, these data fit together nicely; we understand that his performance that day might be a bit better or worse than indicated by his aggregated statistics. But why?

Clearly, day-to-day performance varies from average – here, WYSIATI strikes. It seems that this difference is virtually attributed entirely to statistical deviation (or whisperings of injury). Since baseball players’ off-field lives are largely hidden from public attention until tragedy or controversy arises, this sort of information is unavailable to System 1. Unruffled by what it doesn’t know, System 1 thinking converts three-dimensional players into just outward action – or inaction – during baseball games.

To a degree, this condensed narrative isn’t a problem, or at least not a serious one. It’s certainly none of our business what goes on in the private lives of the players, and elite baseball players have mental fortitude that most of us can scarcely imagine. Still, players are sometimes treated as if they’re sports cars, robots, or some other hot commodity. Day in and day out, they’re expected to strive toward being Players. It’s important to actively remember that baseball players are afflicted by a number of issues off-field – just like everyone else – that can occasionally affect their performance more than, say, variations in the field they’re playing on.

Coupled with incomplete context, System 1’s WYSIATI tendencies also give rise to another heuristic: the halo effect. Similar to WYSIATI, the halo effect refers to the extrapolation of one part to a description of the whole. This effect is particularly common in daily life: people often associate attractive people with more positive qualities, and they consider that a person who is outstanding in one field – say, a successful entrepreneur – should demonstrate increased aptitude in other areas as well – like golfing. The halo effect is typically applied to people we simultaneously admire and don’t know very well. Athletes and celebrities alike (though the argument can be made that some athletes are celebrities) fit into these categories. We like to believe that good players are also good people, and though these extrapolations sometimes hold true, a dive into baseball’s controversies reveals that our good intentions can be unfounded.

Our relationship with baseball players largely consists of how we view their quantitative performance. At the moment, I’m going for breadth rather than depth, so we’ll skirt around specific statistics that we can get quagmired in. As with most concepts in behavioral economics, the Law of Small Numbers intuitively makes sense, but it’s also often buried under System 1’s impervious nature.

I watched the Astros play the Angels in the last game before the All-Star break with my dad and brother. A Little League coach and a Little Leaguer,  respectively, they epitomize the cross-section of average baseball fans that watch the game with their heart, not their brain. The pace of baseball is conducive to lengthy conversations about the players, intermittently interrupted by cues to get loud and plays that invoke noise without outside encouragement. At some point after Yuli Gurriel’s game-tying grand slam, either my brother or my dad said, “Yuli’s been playing really well.” A few days prior, after Alex Bregman struck out in two high-pressure situations during a game, the other said, “What’s up with Bregman?”

Okay, Yuli Gurriel has been having a tremendous season. But really – accusing Bregman of a slump after two strikeouts?

This is a prime case of the Law in Small Numbers in action, and once you notice it’s there, it’s everywhere. Think about how frequently players’ futures are extrapolated from one season, or a team’s postseason potential from a single winning or losing streak. It’s also the root of things like lucky socks and beliefs that we, as viewers, can influence the outcome of a big game – namely, a disconnect between limited data and results. Kahneman again:

“The exaggerated faith in small samples is only one example of a more general illusion – we pay more attention to the content of messages than to information about their reliability, and as a result end up with a view of the world around us that is simpler and more coherent than the data justify…[Furthermore,] statistics produce many observations that appear to beg for causal explanations but do not lend themselves to such explanations…Causal explanations of chance events are inevitably wrong.”

If you think that his summary of the Law of Small Numbers sounds like WYSIATI, then you’d be right. WYSIATI is essentially an umbrella that covers a multitude of more specific heuristics, a great number of which address the human weakness to intuitively interpret statistical data. A particularly memorable statement from the famous “helping experiment” conducted by Richard Nisbett and Eugene Borgida is that “subject’s unwillingness to deduce the particular from the general was matched only by their willingness to infer the general from the particular.”

In plain baseball English, fans tend to make a bigger deal of specific situations – such as two at-bats in high-pressure situations – than the bigger picture. Does one awe-inspiring, jump-to-your-feet-and-holler catch make a phenomenal outfielder? Not necessarily. But in the heat of the moment, with the sudden spike in stimuli, System 2’s voice of reason can all too easily be overshadowed by System 1’s automatic, emotional response.

Also, the idea that players who are doing abnormally well in one aspect or another, such as those who have the lowest ERA or the best wRC+, will continue to improve that aspect, ignores another statistical tenet: regression to the mean. Typically, players who are absolute superstars one season will play a bit worse the next season, at least relative to their previous season. Very seldom do single-season WAR leaders improve their WAR the following season – those who do are the exception, not the rule. Like with the other heuristics, we know this, at least if we think about it. But most fans don’t think about that too often; we like to think that players will improve ad infinitum. And really, it’s no fault of our own – our brains just aren’t built to run regression models. If regularly reminded to engage System 2, most of us would do so, but otherwise, System 1 can become complacent in its intuitive evaluations.

The inclination to generalize from the particular is fundamentally due to the neglect of base rates, which are non-situational probabilities. For example, the likelihood of a team with 100 wins making the playoffs is a base rate, whereas the likelihood of a 98-win Twins team winning the AL Central is situational and therefore isn’t a base rate. When we posit the probability of a specific event happening, we tend to be overconfident and overestimate the chances of a rare event (especially if that event is vivid or familiar to us – more on that later).

In other words, we like to apply all sorts of special circumstances to make our predictions more coherent; in turn, System 1 feels our predictions are more justified, even though in a substantial subset of cases, we’re better off answering the question using base rates (how often do 100-win teams make the playoffs?) rather than situational, likely over-adjusted statistics (how often do 98-win Twin teams make the playoffs?). Note that this applies to intuitive predictions – mentally adjusting for a certain team is a very different story than trawling through databases and using rigorous statistical methods to adjust accordingly.

Teams

Switching tracks, you probably like one particular team a little (or a lot) more than the other ones. Why that team?

If you’ve dropped into the nostalgic corner of baseball, you’ve probably read touching stories of how people allied themselves with certain teams because of their parents, a life-changing player, or simple geography. In another, slightly less sentimental corner, there are people who get into baseball or switch their favorite team after a World Series.

If you look down upon bandwagoning, then think about the previous postseasons. If you’re not a Red Sox, Astros, Cubs, or Royals fan, who did you root for? And what was your process of picking that team?

When it comes to the dilemma of choosing a team to support, I’ve narrowed the process down to three options.

Option one: Meticulously compare and contrast your different options, make pros and cons lists, and/or run a modified cost-benefit analysis.
Option two: Stop watching the postseason games.
Option three: Pick whichever team you (or the people around you) feel more positively about, possibly on the spot.

I might be letting my System 1 take over, but without conducting extensive research, option three feels like the most popular outcome. It also embodies the affect heuristic, which is the substitution of a challenging question with another simpler, typically emotional-oriented question. For option three, the question “what team should I support?” is replaced by “which team do I feel most positively about right now?” If these questions seem very similar or identical, that’s an indication of how deeply entrenched the affect heuristic tends to be.

The affect heuristic’s older cousin is availability. Kahneman says that “the importance of an idea is often judged by the fluency (and emotional charge) with which that idea comes to mind.” The more available an idea (or player) is, the more important (or talented) it appears.

Think about All-Star fan voting, in which players who play on big market teams or are often in the news make it through disproportionately, demonstrating a discrepancy between perceived and actual skill. In some cases, this availability cascade – the expansion of the availability bias into policy (or starting lineups) – sparks outrage among fans who vote using their System 2. If it’s sufficiently loud, this buzz can also spread to fans relying on their System 1. The reason? The names of those snubbed have now become widely available.

Most of us are bad at making good intuitive predictions. Kahneman writes:

“The prediction of the future is not distinguished from an evaluation of current evidence – prediction matches evaluation… People are asked for a prediction but they substitute an evaluation of the evidence, without noticing that the question they answer is not the one they were asked.”

The key idea here is that people don’t account for future uncertainties. We can find evidence of this everywhere, from enthusiastic assertions that a team’s current momentum will continue to power rankings that ignore the inevitability of injuries. Worse, the people making these predictions – including you and me – tend to be overconfident in their optimism. Recall that System 1 bases the validity of judgments on the coherence of existing evidence, not wariness on what it doesn’t know. It gives way to the ever-present confirmation bias, and in the process, spins theories that sound fantastic but fall away once the starting pitcher gets injured or the third-place team in the division goes on an unlikely winning streak.

One more concept to note: System 1 gets us through our daily lives mostly unscathed, and it does make correct predictions now and then, even without the assistance of System 2 and Statcast data. If that’s the case, you might hear some version of “I told you so!”

When pressed, that person might respond with a list of suspiciously soft reasons or strategies that were also used by people who guessed otherwise; with incomplete data, multiple convincing narratives can be created. Though innately System 1, outcome bias (aka hindsight bias) can come across as unbearable gloating – better luck next time!

Should We Be More Rational?

Sure, baseball fans are irrational. So are baseball players and managers. And everyone else. But just within baseball, is irrationality a problem?

I’ve been thinking about this question for quite a while now, and I have an enormously satisfying answer (I kid): it depends.

On what does it depend? Primarily, I think, on your relationship to baseball. Some people are casual fans who tune into an occasional Sunday game, can name a couple of players, and enjoy sporting the hat of their favorite team. Others, like my brother and my dad, love playing the game first and watching it second. Still others are fascinated by the numerical aspect of baseball. Is there anything wrong with any of these approaches? Of course not. It just depends on why the game matters to you.

It’s worth acknowledging that System 1 and System 2 aren’t actual characters driving your brain; System 1 is shorthand for automatic, unconscious processes and System 2 represents your deliberate mental functions. The psychologist Walter Mischel prefers hot and cold systems of thinking instead – hot thinking is rash and impulse-driven, much like System 1, whereas cold thinking is thorough, as is System 2 – but the fundamental understanding is the same. Baseball can be viewed through the emotional System 1, the logical System 2, or some combination of the two. Part of what makes the sport so fantastic is how the action on-field can be interpreted in millions of different ways by as many fans. Is baseball just a form of easy entertainment? A dream? A statistical goldmine? That’s up to you.

If you ultimately decide that you want to become more rational, though, there are a few things you can keep in mind.

WYSIATI’s influence on us is that we often feel as if our predictions are more likely to be right than they actually are: this is the illusion of validity. A psychologist at the University of Pennsylvania, Phillip Tetlock, describes decision-makers in terms of hedgehogs and foxes. Hedgehogs, he says, have a theory about the world, are opinionated and clear, and are confident in their forecasts. Even when they’re wrong, the prediction is “off only on timing” or “very nearly right.” On the other hand, foxes attribute outcomes to the interactions of many agents and forces, including luck, and though foxes are still “very poor” at making predictions, they perform better than hedgehogs.

If you think of yourself as a hedgehog but would like to be more like a fox – what fun terminology! – Kahneman recommends the premortem: presuming that the outcome of your plan/prediction is a disaster and writing about what went wrong. By doing so, you legitimize your doubts and can account for some of the uncertainties that fly under the radar in some intuitive predictions.

Then again – our irrationalities define our experiences with baseball, and they certainly make the game a lot more exciting.

References & Resources

Thinking, Fast and Slow, by Daniel Kahneman
Nudge: Improving Decisions About Health, Wealth and Happiness, by Richard Thaler
Predictably Irrational: The Hidden Forces that Shape Our Decisions, by Dan Ariely
The Marshmallow Test: Mastering Self-Control, by Walter Mischel