March Madness mission control: Big comfy chair? Check. Big-screen TV? Check ... (Kat Snow/KQED)
March Madness mission control: Big comfy chair? Check. Big-screen TV? Check … (Kat Snow/KQED)

Wednesday Update: Why the eight “play-in” games, determining the last four teams to make the dance, really matter.

Tuesday Update: Silver computes his own odds of winning Warren Buffett’s challenge as one in 7,419,071,319.

In case you were thinking of taking on Warren Buffett’s billion-dollar challenge to fill out a perfect bracket for the 2014 NCAA Tournament, Nate Silver has some data that you might want to check out.

Silver is the statistician and writer who correctly predicted the presidential election outcomes in all 50 states in 2012; he has relaunched his website,, funded by ESPN.

It’s loaded with articles on everything from how statisticians could help refine the search for that missing Malaysian Airlines jet to which of the 68 teams have the best odds of winning the NCAA Tournament this month.

First to the latter.

According to Silver and computer programmer Matthew Conlen, only 14 teams have any appreciable statistical (better than 1 percent) chance to prevail to the very end of March Madness. (Stanford, it should be noted, is not among them.) The chance of success is expressed as a percentage and listed next to each team:

  • Louisville (15)
  • Florida (14)
  • Arizona (13)
  • Michigan State, Virginia, Kansas (6)
  • Wichita State, Duke (5)
  • Villanova (4)
  • Michigan, Creighton, Wisconsin (3)
  • Kentucky, Ohio State (2)

Silver’s model for predicting the brackets, which is in its fourth year, is principally based on a composite of five computer power ratings:

Good luck with that.

Finding a Lost Airplane

As for the mystery surrounding Malaysian Airlines Flight 370, which disappeared just off the Malaysian coast on March 8, Silver’s associate Carl Bialik writes:

“Bayesian statisticians are particularly helpful in a search operation. Their methods allow hunters to update their estimates of the probability of finding their target in any latitude-longitude combination — or even in three dimensions, accounting for depth in the water. Bayesians helped hunt U-boats in World War II, a U.S. submarine in the 1960s and an Air France jet in 2011.”

The problem in this case may be the vast distances that need to be searched since the plane disappeared. The Malaysian government seems to have no idea to which end of the earth the Boeing 777 proceeded before sending its last-known signal to a satellite.

I’m going to take a wild guess here and say someone, someday will locate that airplane. But as for the odds of winning Buffett’s challenge, with or without Silver’s analysis at your side, forget it.

  • Arthur Saget

    Naah, even Bayes couldn’t find the plane.


David Weir

David Weir is KQED's senior editor for digital news.  He previously worked at Rolling Stone, Salon, Wired Digital, Excite@Home, Mother Jones, and as a co-founder and executive director of the Center for Investigative Reporting.

Over the past 40 years, he and his teams have won dozens of awards, including a National Magazine Award, an IRE Award and a Webby. He has authored or co-authored four books, including (with Mark Schapiro) Circle of Poison.

He has a bachelor's degree in journalism from the University of Michigan, and has taught journalism at the University of California at Berkeley, Stanford and San Francisco State.

Sponsored by

Become a KQED sponsor