Reply to thread

Message: [QUOTE="auror, post: 3223905, member: 1329"] A 3 point difference in a 1 game series that Duke led for much of makes it "clear" that UConn was the better team? On that night after the 40th minute, sure. You're actually making my point. 1 game trials don't determine who the best teams are. They never truly have. They crown a champion, not the best team. Teams that are favored by 6 points by Vegas win only ~80% of the time. Teams that are 9.5 point favorites still lose 1 out of 6 games. Those outcomes don't always make the spreads or AdjEm or whatever wrong and they don't mean that the best team was actually worse. They happen. And we watch sports for those 1 out of 6. The unpredictability is one of the things that makes sports great. AdjEM isn't perfect. It's not exact, because even 30-40 game trials aren't enough data. In addition, team quality isn't static (just ask 2011 UConn). But it's a hell of a lot more representative than 1 or even a handful of game results, and proven much more accurate than the polls. There is likely a margin of error between close adjEMs. Is a team with 37 definitively better than with 34? Usually, but probably not always. But what my argument comes down to is that the gulf between 37, let alone 34, and 43 adjEM is double (or triple) that. It's a legitimate separation beyond margin of error or inaccuracy after that many games. [/QUOTE]