Maya Chen: Nathan, okay, long week — but I have to start with something my nephew said to me, he's twelve, he's obsessed with March Madness, and he said, totally flat, 'UMBC beat Virginia once so Virginia wasn't actually that good.' And I — mm, I just froze.
Dr. Nathan Hayes: A twelve-year-old has located the entire philosophical problem.
Maya Chen: He really has — because here's what happened: March 16, 2018, UMBC, a No. 16 seed, beat Virginia, the No. 1 seed, in the NCAA Division I Men's Basketball Tournament. And not just beat them — ended their season, full stop, because the format is single-elimination. One loss and you're done.
Dr. Nathan Hayes: And that had never occurred before that night — a 16-seed over a 1-seed, in the modern 64-team bracket.
Maya Chen: First time ever. UMBC — University of Maryland, Baltimore County, a commuter school — walks into the building and the top team in the country walks out eliminated. Because of one game. So my nephew's logic is... not wrong, exactly? But it's also sort of — I mean, is it right?
Dr. Nathan Hayes: That's actually the sharpest way to frame what we're getting into today — whether a single-elimination format, where N teams need exactly N-minus-one games to produce a champion, is actually measuring what we think it's measuring.
Maya Chen: Which is the best team. Or — wait, is it? Is that even what March Madness is trying to find?
Dr. Nathan Hayes: Now that is the question. Because the format's answer and the audience's answer may not be the same thing at all.
Maya Chen: But that's what gets me — because if the format isn't trying to find the best team, what is a win even worth?
Dr. Nathan Hayes: Now, the clearest way I can put this: single-elimination is a spelling bee. You misspell one word — doesn't matter how many you got right before it — you're done. That's the whole mechanic.
Maya Chen: Oh. Yeah, that's — that actually lands.
Dr. Nathan Hayes: Best-of-seven is the opposite structure. You have to lose four times before you're eliminated. One bad night — a cold shooting game, a key player cramps up — it can't end you. The format absorbs it. Now here's the number that makes this concrete: sixty-eight teams in March Madness, which means sixty-seven games total. Every single one is a cliff edge. No averaging across contests, no second chance built into the structure.
Maya Chen: Wait — sixty-seven cliff edges, and UMBC only needed to walk up to one of them.
Dr. Nathan Hayes: Exactly that. And the NBA Playoffs — that's a deliberate structural rejection of this. Every round, best-of-seven. They looked at what single-elimination does and said, no, we're not building that.
Maya Chen: So the NBA knows something March Madness is sort of — mm, choosing not to use. What does the NBA know that March Madness doesn't, or won't?
Dr. Nathan Hayes: That's exactly the tension we need to pull on — because the answer isn't just about fairness.
Maya Chen: And that's the part I want to actually understand — like, the NBA isn't doing that on instinct. There's math behind it, right?
Dr. Nathan Hayes: There is, and it's pretty striking. Lopez, Matthews, and Baumer — 2018, Annals of Applied Statistics — built a unified framework measuring randomness across North American sports formats. What they found: the probability of the stronger team winning increases directly with series length. Not vaguely. Directly, as a measurable function of how many games you play.
Maya Chen: Give me the number.
Dr. Nathan Hayes: Single-elimination — one game — the stronger team wins roughly 75 percent of the time when there's a genuine skill gap. Best-of-seven? That climbs to around 97 percent. And the mechanism is — actually, it's almost elegant. Best-of-seven requires four wins. No single random outcome can determine the champion. Each game compounds the stronger team's structural edge, so the variance window that let UMBC walk through? In a five-game series against Virginia, that window closes.
Maya Chen: Wait — 75 to 97. That's not a small gap.
Dr. Nathan Hayes: It's not. And Ben-Naim, Hengartner, Redner, and Vazquez — separate work, arXiv 1209.4724 — put a label on this that I think is the most honest framing in the literature. They call single-elimination 'efficient but unfair.' Efficient because N teams, N-minus-one games, done. Unfair because the best team has a, quote, relatively low probability of winning the championship. Those two things are both true simultaneously.
Maya Chen: So we know this. The research exists. And March Madness — sixty-seven cliff edges, every year, on purpose.
Dr. Nathan Hayes: On purpose. Think about an 8-seed in the NBA Finals — it almost never happens. Not purely talent gap, though that's real. It's that seven games eliminate the variance window. A team would have to beat you four times, meaning they'd have to be genuinely better than you on most of those nights. The noise tolerance is — mm, it's built into the architecture.
Maya Chen: Which makes me wonder — and this is something we'll get to — whether some leagues have looked at exactly this math and decided, no, we actually want the unfair version, and then defended that choice out loud. That's a different kind of honesty entirely.
Dr. Nathan Hayes: Well — the EuroLeague actually did that. The Final Four, single-elimination semifinals and final, after a full season. And economists and analysts looked at it and said, plainly, this format is less fair. And the EuroLeague's response was essentially: yes. We know. We chose it anyway.
Maya Chen: Wait — they said it out loud?
Dr. Nathan Hayes: Defended it. Not apologized for it. The argument was that the structural unfairness is what produces the excitement. They weren't hiding the trade-off — they were naming it as the product.
Maya Chen: Okay, so — I mean, that's what makes this actually uncomfortable, right? Because March Madness does the same thing, just without saying it. Every year. And I keep thinking about — picture an assistant coach at some mid-major school, eleven p.m. on a Tuesday in February, watching film. Knowing that one great night, just one, could end a 1-seed's season. That's not a flaw he's working around. That is literally the system he's preparing for.
Dr. Nathan Hayes: Right — but is that a legitimate design choice, or is it capitalism wearing tradition's clothing?
Maya Chen: Both. I think it's — actually, no, I don't want to let that be a dodge. It's both, and the both-ness is the honest answer. Because what we're really choosing between isn't two statistical methods. It's — sustained excellence versus peak performance. Those are genuinely different ideas of what a champion should be. The NBA rewards the team that was better across eighty-two games plus four series. March Madness rewards the team that had one transcendent night at exactly the right moment.
Dr. Nathan Hayes: The competitive legitimacy question is settled by the math, though. Best-of-seven gets you to 97 percent. That assistant coach's entire preparation — it's optimizing for the 25 percent variance window. He's not trying to prove his team is better. He's trying to find the door the format left open.
Maya Chen: And we love him for it. That's — yeah, that's the thing that doesn't resolve.
Dr. Nathan Hayes: It doesn't. And maybe the most honest version of this — the EuroLeague version — is to just say: we have chosen drama over accuracy. Not as a failure. As the design.
Maya Chen: Someone made those choices. Like — the NCAA tournament started with eight teams in 1939. Eight. And now it's sixty-eight, and they voted to go to seventy-six in 2027. FIFA is going to forty-eight teams for the 2026 World Cup. None of that is tradition. Those are decisions someone made in a room, with specific pressures, economic and political and, yeah, philosophical. Every expansion is a negotiation about what the format is actually for.
Dr. Nathan Hayes: And the honest version of what they chose — if you say it plainly — is: we built a champion-producing machine that sometimes gets the wrong answer, because we believe peaking once at exactly the right moment is a real form of excellence. Not a consolation prize for the real thing. The thing itself.
Maya Chen: Which is — yeah. That's it. My nephew said Virginia wasn't actually that good. And I think the honest answer is: Virginia was extraordinarily good, and UMBC peaked at the right moment, and the format said that's enough. Not a mistake. The design.
Dr. Nathan Hayes: Most leagues do it. Almost none say it.
Maya Chen: Almost none say it. That's the whole thing.
Dr. Nathan Hayes: Good conversation. I mean that — you pushed the right places.
