When Statcast(TM) is everywhere, for everyone , prejudice will be more costly
Not so long ago, baseball managers only ordered dramatic infield shifts because they noticed a pronounced tendency of certain hitters (most notably Ted Williams) to pull the ball to one side of the diamond. Shifts were very rare. The purpose of a shift is to raise the chance that a fielder will be able to handle a batted ball and gain an out that would not otherwise be made. The decision by a manager to shift against a batter can yield benefits over the cost of a season, but every batted ball that would have been caught by a traditionally placed fielder exposes the manager to criticism. So it was rare.
Today, shifts are common — many times in a game infielders will rearrange themselves for the next batter, and sometimes they’ll rearrange themselves again in reaction to the count. Managers rarely order shifts — they are determined by data analysts long before the game, and so players position themselves based on a predetermined policy.
What’s changed? The availability of Statcast, Major League Baseball’s proprietary system to record each action on the diamond to an extraordinary level of detail. Every pitch, every batted ball, every fielder movement is preserved in huge databases that analysts can sift through to detect patterns. This mining process is far more powerful, and far more discriminating than the choppy recollections of manager about players they remember seeing.
What does this have to do with privacy, or discrimination?
Consider the act of shifting your fielders as if it were profiling, or discrimination. You are treating people differently, and you need to justify that. I remember in Little League, we used to pull the outfielders in for small batters, and order them back for the big boys.
Major League players look for, and exploit even the tiniest edges, and sometimes they used to position themselves by profiling the batter. In the days before Statcast,the Mets introduced a new shortstop — Reuben Tejada — who was a short, thin Latino. He looked like someone who can really fly. For a surprisingly long time, infielders used to shift in a step or two when he came to the plate, so they could pick up a ground ball just a fraction of a second earlier, and get the ball to first in time to beat a guy who they thought could really burn his way down the first base line. Quite often they rushed the play, thinking they had no time to spare, and bobbled it.
But Reuben wasn’t fast — he was as slow a runner as the Mets had.
When we stereotype people, when we “go with our gut”, when we decide whether to cross the street when a group of young men approach, when we choose a stranger to approach when we need help, we’re doing the best we can with incomplete information. It’s a roll of the dice, sometimes we’re right and sometimes — like the infielders positioning themselves for Reuben Tejada — we’re laughably wrong.
Panhandlers profile as much as anybody — like cops, their success depends on their ability to focus attention on the people who yield the most rewards for the expected risks, and so when you and a panhandler approach each other, you are being sized up. You are being profiled based on who you look like. He chooses based on his experience, his stereotypes, his prejudices, or his bigotry — the basis is the same, a word for it is chosen because the author wants to highlight, or obscure, the sinful nature of his choice.
While you’re considering the morality of the panhandlers choice, and hence the correct word to use, you’re missing the mote in your own eye, because you probably didn’t pause in your reading to question the use of “his” to designate the panhandler. Because in our experiences, our stereotypes, and our prejudices, a panhandler will always be imaged as a man. Woman just don’t panhandle in our minds.
But there are female panhandlers, some here, far more in Third World countries.
Consider that not too far in the future, Statcast will be everywhere, recording everything, and available for use by everyone. Just 15 years ago, no one predicted MLB Statcast — the technological leap in this century has been that dramatic. And it’s accelerating. In many cities, there is already video monitoring of all public, and many private, spaces. Facial recognition is becoming better, cheaper, and, for those who seek anonymity by wearing masks, being augmented with gait recognition. Speech recognition makes it possible to create text logs, and databases of people’s public conversations. Many police cars have camera systems that routinely scan and record every licence plate seen along their patrol route. Every movement, every act, every word, every gesture and every interaction will be digitized, logged , mined for patterns, and sold.
Statcast will be everywhere, for everyone. Even for panhandlers — after all poor people now have, in their phones, more computing and storage power than NASA used for the moon launch. So sometime in the future, panhandlers will have simple devices that scan approaching crowds, instantly identifying individuals, and characterizing them based on their actions that day, that week, or in the past years. Which ones have ever given to panhandlers? Which ones are panhandlers themselves?
So while a panhandler now chooses a mark based on clumsy stereotypes, or avoids encounters with people that look — to his prejudiced mind — like escaped felons, in the future his app will scan the faces before him, and highlight candidates to approach, and even recommend some introductory lines that have worked with that individual in the past. Stereotypes and generalization and bigotry are strategies to deal with incomplete information about people. Now, with Statcast everywhere, panhandlers will circle someone solely because of their individual past behavior.
Of course, the beseeched will not be left without tools of their own. Now, people avoid panhandlers by shifting their route in order to avoid certain approaching people, because a quick judgement in their mind — again from experience, from stereotyping, from prejudice — resolve on a characterization even before a supplicant has spoken. With Statcast everywhere, simple pedestrians will no longer be confused with panhandlers.
Now, prejudice has hardened many of our hearts, and so we often ignore the pleas of panhandlers that have approached despite our best efforts to avoid them. Sometimes we say no to the very first request a stranded, or hungry, person has ever made, because our ignorance forces us to assume the worst: that we’re being played, and that our natural generosity is being exploited for frivolous gain. But when Statcast for everyone arrives, someone who has never panhandled before will find a more fertile field for their pleas. No app alerts will be raised when they approach pedestrians. Those approached will have their generosity fostered by the certainty they’re alleviating a true need. They won’t even have to insult the person by checking their background — the Statcast app would have discretely cleared them for approach when they first became visible.
So we see that with better knowledge, characterizations based on race, gender or ethnicity lose their power. They’re too coarse to compete with finely grained information about an individual persons history.
We think of privacy as an unquestioned good, fearing that close inspection of our actions will give others power over us. But keeping our actions, and our history private forces others to defensively frame us with blunt, and often unkind generalizations. Keeping actions private free us to do wrong with less consequences, because others will not know what we’re capable of.
I’m not really arguing against privacy — this is just a thought experiment to clarify the link between prejudice and privacy, and prepare us for the consequences of a world where our public actions will be known by everyone, forever.