Here's the arithmetic a new BMJ study did on Tuesday. 878 former NFL players died between 2016 and 2021. 235 of their brains were donated for research. 215 of those had CTE. Now do the thing no previous study could: assume every single brain that wasn't donated was perfectly healthy — all 643 of them, spotless. You still get 24.5%. At least one in four. * Those two words, 'at least,' are the entire invention. For a decade, the comeback to the famous brain-bank studies — the ones finding CTE in 99% of donated NFL brains — was selection bias: families donate a brain when they've already watched something go wrong, so of course the donated brains look bad. Fair objection. This study's response was to concede it completely, hand the other side every uncounted brain for free, and see what number survives. One in four survives. That's not an estimate. That's a floor with a building on it. * Now read CNN's headline: "1 in 4 former NFL players who died from 2016 to 2021 had CTE, study estimates." The cohort is scoped, the dates are right — genuinely careful work — and the two load-bearing words are gone. Better yet, CNN's own first sentence begins "At least 1 in 4 former National Football League players..." The operator survived the reporter. It didn't survive the headline desk. * Why fuss over two words? Direction. "1 in 4, study estimates" reads like a midpoint — a number that might get revised down next study. The actual paper's range runs from 24.5% to 97.7%, and the researchers say the truth is probably far above the floor, given that 215 of the 235 brains they did examine had the disease. 'At least' tells you which way the uncertainty leans. Deleting it flips a minimum into a maybe. * The neighbors handled it differently. The New York Times: "Study's Alarming Finding: At Least 1 in 4 N.F.L. Players Gets Brain Disease" — operator intact, though 'gets,' present tense, quietly promotes a cohort of dead players into every player who ever laced up, which is its own little inflation. The Washington Post kept it clean: "at least 1 in 4 NFL players suffered from disease." The BBC went with "'higher than previously shown,'" a phrase so cautious it would also cover a finding of six. * The stakes aren't academic. The NFL's statement, printed in CNN's piece, says the league "continuously strives to make the game of football safer." The league has spent ten years living rent-free in the gap between 'donated brains' and 'all brains.' This study bricked that gap shut — and a headline that trades 'at least' for 'estimates' quietly pries a brick back out. Meanwhile the study's lead author points out the bulk of these head impacts happened in college, high school and youth ball — meaning the people who most need the floor stated plainly are parents reading exactly one headline before practice sign-ups. * CNN's article, to be clear, is thorough — the history, the caveats, the CBA, the numbers, all there. But a floor study earns a floor headline. When scientists finally produce a number that can't be argued down, the least a headline can do is not argue it down.