In the flat heat of the Rose Bowl in Pasadena, on July 17, 1994, Roberto Baggio stepped up to take a penalty and sent it over the bar. Brazil beat Italy 3–2 on the shootout. It was the first World Cup final ever decided that way, and after two hours of nothing it ended in the smallest possible margin: one kick, one man, the ball rising into empty air.
The men who mattered for this story were nowhere near Pasadena. They were in Atlanta, watching the final on television in a pair of bars — one where the Brazil fans gathered, one where the Italians did — and before and after the match they were spitting into plastic vials.
They were part of a study. A team of researchers led by Paul Bernhardt and James Dabbs wanted to know what a game does to the body of a person who only watches. Twenty-one men, split between the two teams, gave saliva samples before kickoff and again once it was over. The samples went off to be assayed for testosterone.
The result was clean. In the Brazil fans, testosterone climbed by around a fifth. In the Italy fans, it fell by about as much. None of these men had run a step or touched the ball; they had never set foot in the stadium. The players had no idea they existed. And yet their endocrine systems had responded to the outcome as though they themselves had won or lost a fight.
The hormone of the contest
Testosterone is not, despite its reputation, simply the fuel of aggression. It tracks status. In animals that compete for rank, testosterone rises in the winner of a contest and drops in the loser, and the shift isn’t cosmetic — it tunes the animal for what comes next. A winner, flush with it, is primed to press the advantage. A loser, depleted, is primed to withdraw and avoid a fight it would lose. Biologists call this the “winner effect,” and it has been documented across fish, birds, and monkeys. In humans it is fainter and still argued over — but it surfaces most reliably where Bernhardt and Dabbs went looking, outside the laboratory, at the venues themselves.
The strange part is how loosely the hormone is tethered to actually competing. In an earlier study, chess players showed the same pattern — testosterone up in the winner, down in the loser — from a game fought entirely with the mind, no muscle involved. Bernhardt and Dabbs pushed the tether further still. They ran the same measurement on people who did not play at all.
Their first study, before the World Cup one, had watched eight men at a Georgia–Georgia Tech basketball game, two old rivals meeting on a neutral floor. Same result: the winning fans’ testosterone climbed, the losing fans’ sank. Two settings, two sports, one finding. The bodies of spectators were keeping score as if they had skin in the game.
The paper, published in Physiology & Behavior in 1998, framed it against an idea that was already well established in psychology.
We won, they lost
Two decades earlier, the social psychologist Robert Cialdini had caught the same instinct without drawing a drop of blood. Through the 1973 football season, at seven large universities, his team quietly counted the school-branded clothing students wore to class every Monday — the sweatshirts, the caps, the jackets. On the Mondays after a win, the apparel bloomed. After a loss, it thinned out. The wardrobe tracked the result of a game the students had merely watched.
Then came the language. Cialdini’s group telephoned students, posing as survey takers, and asked them to describe how their team had done. The pronoun tracked the scoreboard: describing a victory, students reached for “we won” far more often than they did describing a defeat, which got “they lost” and the bare score. When the team succeeded, fans folded themselves into it; when it failed, they quietly stepped back out. And the flip was sharpest among students the callers had just told they’d scored badly on a quiz — the ones with something to repair. That told Cialdini what the pronoun was for.
He named it “basking in reflected glory.” The glory belonged to the players; the fans warmed themselves at it anyway. Its shadow got a name a decade later, from another team of psychologists: “cutting off reflected failure” — the instinct to peel away from a loser so its defeat doesn’t stick to you. The whole system runs on the same underlying assumption, held somewhere below conscious thought: that the team’s outcome is, in some real sense, your own.
Cialdini treated this as a cognitive move — a way of managing the image others hold of you. Bernhardt and Dabbs showed it reaches deeper than image. The identification is not only in what you say and wear. It is in your blood.
The machinery is older than the game
Football is roughly a century and a half old. The World Cup dates to 1930. Whatever in the human body responds to a rival’s goal by dumping or withholding a status hormone did not evolve to watch football, because for nearly all of human history there was no football to watch. The response is a piece of much older machinery, firing at a target it was never built for.
That older machinery, one influential argument holds, is coalitional psychology — the mental equipment for belonging to a group that fights other groups. Humans spent the overwhelming majority of their existence in small bands, and those bands were not at peace. Estimates from surviving small-scale societies put the share of male deaths caused by warfare somewhere between 13 and 30 percent. For a man in that world, the coalition was not a hobby. It was the thing that kept him alive, and the thing that could get him killed.
A mind shaped by that pressure comes equipped to do certain things fast and without instruction: sort the world into us and them, feel loyalty as a physical pull, read a rival group as a threat, and rally to your own side when the two collide. The psychologist Mark van Vugt has argued that this equipment is sharper in men, because coalitional violence was overwhelmingly a male enterprise — the “male warrior” pattern of banding tightly with allies and bristling at outsiders.
Team sport is a near-perfect key for that lock. Two groups in matching colors, a contested field, a clear us and them, a zero-sum outcome, a crowd roaring on each side. None of the reproductive stakes of real intergroup conflict are present, and none of the real danger. But the surface features are exactly the ones the ancient system was tuned to detect. In 2010, the evolutionary psychologists Benjamin Winegard and Robert Deaner made the case directly: fandom is not an adaptation in its own right — it earns you nothing, in the currency evolution counts — but a byproduct, the coalitional mind misfiring, harmlessly and pleasurably, at a synthetic war.
Their study of nearly 500 undergraduates found what the theory predicted. Fandom tracked with loyalty. Among men, it tracked with a reverence for authority and group purity. And men were the more intense fans — a gap that ran through loyalty only slightly, the authors were careful to say, with most of it left unexplained by anything they measured. The wiring van Vugt described, pointed at a stadium instead of a rival band, and still mostly untraced.
Where you were born
This is where the accident of birth comes in. A fan rarely chooses the team. It arrives with the city, the family, the neighborhood, before choice is even possible — a boy inherits his father’s club the way he inherits a surname. There is nothing rational in it, and no one pretends otherwise. In a world of real coalitions, though, that was exactly the point. You did not audition your tribe on the merits. You were born into it, and its enemies were your enemies from birth, and the whole system worked precisely because loyalty ran deeper than reason. The arbitrariness is not a flaw in fandom. It is the fingerprint of the older thing it descends from.
None of which makes the feeling less real. The machinery does not know the war is pretend. It delivers the same loyalty, the same dread before a big match, the same lift of belonging in a full stand — and, when the rival scores, the same drop, measurable now in a vial of spit. The strangest thing the research turns up is not that fans overreact. It is that, at the level of the body, they are not reacting to a game at all. They are answering a much older call, from a time when the group really was everything, and losing to the other side really could cost you your life.