04 April 2011

Confounding the "experts" - updated

The unlikely finalists in this year's NCAA basketball tournament show why this event garners so much interest.  As the Wall Street Journal explains:
"To say that this Final Four may be one of the unlikeliest of the modern era would, for once, not be hyperbole. There is no No. 1 seed in the Final Four. There isn’t even a No. 2. VCU is an 11th seed. Butler is seeded eighth. Connecticut finished ninth in its conference."
President Obama has been chided by pundits for selecting four No. 1 seeds, none of whom are in the final four.  I've embedded at left a screencap of some of the predictions by the experts at ESPN.  Of the 56 choices by 14 participants, only 2 made the final four.

Now, here's what's really amazing.  The odds of someone picking this year's correct Final Four are terrible, but...
"Out of 5.9 million entries in ESPN’s Bracket Challenge, two people correctly called this group... ESPN had a phone interview with... Joe Pearlman of East Brunswick, N.J., this morning. He claims that his picks was based mostly on gut instinct..."
Truly incredible.

Update, with Butler/Connecticut pending tonight.  Again from the Wall Street Journal:
"Between the approximately 8.9 million entries in ESPN and Yahoo’s bracket challenges, only three entries predicted this Final Four. That’s 0.0000337% of all entries. Yahoo’s only perfect Final Four came from Diana Inch, a high school librarian in Oregon. Her strategy included picking teams whose names included the letters V or X and were seeded close to 7 or 11, her favorite combination of letters and numbers. VCU, an 11 seed, fit the bill. She then brought mascots into the mix, and came to the conclusion that UConn will defeat Butler on Monday, based mostly on the fact that she loves dogs and that UConn’s husky bears a resemblance to one of her own pets. “[I] thought that picking teams based primarily based on their seeding numbers and mascots would potentially irritate some of my male sports-obsessed colleagues,” Inch tells Yahoo’s Brad Evans..."

4 comments:

  1. I disagree with the assessment that the odds are astronomical.

    The odds of picking the winner for one region (say the Southeast) is 1 in 16. There are 16 teams and you are allowed to get every match-up wrong except the games your winner is playing in.

    The odds that you pick the winner for two regions is 1 in 16*16, or 1 in 256. The odds that you do it for all four are 1 in 16*16*16*16, or 1 in 65,536.

    If 5.9 million people made brackets completely at random by flipping coins you would expect 5,900,000 / 65,536, or about 90 people to have picked these four.

    If you require a Perfect Bracket (none wrong) to this point the odds do indeed get astronomical VERY quickly. The probability of having a perfect bracket is 1 in 2^(number of games played). The odds of having a Perfect Bracket to this point (60 games) are 1 in 1,152,921,504,606,846,976.

    See [1]

    ReplyDelete
  2. However, said pundits then railed him as well for having wasted valuable Mr. President time doing the brackets in the first place. sigh.

    That being said, I probably spent way too long on mine and got cleaned out in the third round. sigh.

    ReplyDelete
  3. nolandda, you're quite correct that the odds of someone picking the final four are not (as I originally said) "astronomical." With four play-in games making a total of 68 teams in four brackets, there are 18x17x16x17 = 83,232 possible combinatins of teams in the final four.

    That's the theoretical odds generated by picking teams at random - not the same as the odds of these particular four teams making it, since teams do not typically a 50:50 chance of winning each game.

    ReplyDelete

Related Posts Plugin for WordPress, Blogger...