Picture two versions of the same date. Someone threatens you both on the street.
In one version, your date steps between you and the threat. In the other, your date is built the same way, just as tall, just as strong, but doesn’t move.
Now imagine a third version: your date steps in anyway and the attempt fails completely. Before reading another word, it’s worth asking which of the three you’d rate as more attractive afterward, because a research team just ran a version of that exact question on more than four thousand people, and the answer wasn’t the one strength alone would predict.
What seven experiments actually measured
The study, led by Michael Barlev alongside Sakura Arai, John Tooby, and Leda Cosmides and published in Evolution and Human Behavior, ran seven experiments on 4,508 U.S. adults recruited online. Participants read short scenarios describing a date or friend during a threatening encounter, a mugging attempt, and rated how attractive that person was as a partner or friend afterward. Across the experiments, the researchers separately varied two things that are usually tangled together in real life: how physically strong the person was, and whether the person was willing to step in and protect them.
Separating those two variables was the entire point. In ordinary life, a stronger partner is also assumed to be a more capable protector, so it’s hard to tell whether attraction tracks the strength or the willingness behind it. By describing scenarios where a weak person acts willing and a strong person doesn’t, the researchers could finally ask which one was actually doing the work.
It’s worth being precise about what this kind of study can and can’t show. Participants weren’t observed reacting to a real threat. They were reading a written scenario and rating a hypothetical stranger, which is a common and useful method in this kind of research but a step removed from how someone might actually behave, or actually feel, in the moment danger shows up. What the ratings capture is a preference, a snapshot of what people say moves the needle on attraction when the two factors are cleanly separated. That’s still meaningful. It’s just not the same claim as tracking real couples through a real emergency.
The case for strength
Strength has an obvious logical claim on this question. A stronger partner should, in theory, be more capable of actually stopping an attacker, which should make them a safer choice regardless of intention. Prior research on mate preferences has repeatedly found that people do rate physical strength as desirable, and it would be reasonable to expect strength to carry most of the weight in a scenario built entirely around physical danger.
That’s not what the data showed. In the researchers’ own words, “we found only weak evidence that differences in strength, independent of willingness, increased the attractiveness of a partner.” Women did place some value on a date’s strength, but the study found that this was “mostly due to inferences about his willingness to protect them,” meaning strength mattered mainly as a clue to intention, not as a trait people wanted for its own sake.
The case for willingness
Willingness told a completely different story. “Discovering that a person is willing to protect greatly increased their attractiveness as a romantic partner or friend, regardless of their strength,” the researchers reported, and the effect held for both women and men rating both opposite-sex and same-sex dates and friends. The reverse mattered just as much: discovering a date was unwilling to protect decreased their attractiveness, and the researchers describe it as “a deal-breaker for women evaluating a male date.”
That framing, deal-breaker versus nice-to-have, is a meaningfully different shape than a simple preference. A trait people merely prefer nudges a rating up a little. A trait that functions as a deal-breaker changes the category a person gets sorted into entirely, and unwillingness to protect was landing in that second, more consequential bucket.
The pattern wasn’t limited to romantic partners, either. The same experiments asked participants to rate platonic friends under identical scenarios, and willingness to protect moved friend attractiveness in the same direction it moved partner attractiveness, for both women and men doing the rating. That consistency across two very different kinds of relationship is part of why the researchers treat willingness as a general feature of how people evaluate the people around them, not a quirk specific to dating.
What happens when the attempt fails
The most counterintuitive piece of the study tested what happens when someone tries to protect their date and simply fails. An earlier design had left willingness and success tangled together the same way strength and willingness usually are, so a later experiment built a version where the date steps in, the attacker pushes past anyway, and nothing about the outcome changes. The researchers describe the result plainly: “we found no strong evidence that a date who tried but failed was less attractive than a date who tried and succeeded.” Trying and failing rated about the same as trying and succeeding. Only not trying at all dragged the rating down.
The researchers frame this as a sign that what’s being evaluated isn’t really a prediction about physical outcomes at all. It’s a read on character. Across the full set of experiments, the team concluded that “inferences about willingness were the primary driver of how attractive a person was as a mate or friend,” with the ability to actually win the fight running a distant second.
None of this means physical strength is irrelevant to how people size each other up, and the researchers aren’t claiming that. What the seven experiments add up to is narrower and, in its way, more useful: the instinct to step toward a threat reads as attractive because of what it signals about a person, not because of what their body can necessarily deliver once they get there. A failed attempt still sends the signal. Standing still does not.
