Use data, not hype
March Madness bracket advice just went quantitative — KenPom top‑20 teams, low turnover percentage, and free‑throw rate are the metrics the latest analytics videos say most predict tournament success, not momentum or name recognition video — Rice University also pushed a professor’s short explainer showing analytics beat random picks by a small but real margin thread.
Scott Powers, an assistant professor of sport analytics at Rice, built the simulator and explained the method in Rice’s campus explainer video and story. (profiles.rice.edu) Powers’ simulations show a concrete lift: in a 100‑person pool a uniform random pick has about a 1% chance to win, while his tool can raise that to roughly 2.5% in many scenarios. (news.rice.edu) Historical KenPom patterns back the approach: virtually every recent national champion has been inside Ken Pomeroy’s top‑20 efficiency group on at least one side of the ball, and most champions rank top‑20 in both adjusted offense and defense. (kenpom.com) The analytics emphasis on turnovers and free‑throw rates tracks established theory: Dean Oliver’s four‑factors framework lists turnover rate and free‑throw rate among the metrics most tied to winning, and recent analyses repeatedly flag turnover margin as a better predictor of deep runs than “momentum” narratives. (sportingnews.com) The short analytics video that sparked this thread was posted to YouTube this week and cites KenPom and season‑long correlations to justify prioritizing top‑20 efficiency, low turnover percentage and high free‑throw rate. (youtube.com) Bracket services and probabilistic projects have followed suit: recent data‑driven bracket models from SportsBrackets and PoolGenius feed adjusted‑efficiency inputs rather than reputation, and ESPN’s bracketology visuals similarly surface efficiency metrics. (sportsbrackets.net)
