Archive for Monday, March 12, 2007

Long shot

Even with basketball knowledge, chances of bracket-betting perfection remain slim

March 12, 2007


Here's a number you probably don't want to think about today as you're filling out your NCAA tournament bracket for your office pool.

According to Kansas University math professor Ben Cobb, the probability of picking all 32 first-round winners randomly is 1 in 4.3 billion.

You'd better hope your basketball knowledge will give you more insight than that quarter you're about to flip.

And if you're looking for something more than your own wisdom from watching ESPN or your gut instinct, Cobb says not to hold out much hope.

"I really don't know how many mathematicians or statisticians are interested in this," he says. "There are occasional papers in journals about sports in general, but not many."

So rule out any surefire statistical way of knowing who will advance to the Final Four and help you secure your co-workers' money.

(Disclaimer: Office pools are illegal in Kansas, even if nearly every company seems to get caught up in March Madness).

So what should you look at?


We consulted Randy Rosetta, a KU alumnus who covers Louisiana State University basketball for The Advocate, the daily newspaper in Baton Rouge. On the side, he writes about basketball for

"I try to be an intelligent sportswriter," he says. "I look at where they're playing, and who's more likely to have a home crowd, who has a tougher schedule, who won the tough games.

"I do this to the point of overanalyzation, and it's never done me much good. I can't remember the last time I won an office pool. We're supposed to be experts and know what we're talking about, but we can't end up picking those things."

Added to that equation are the sentimental tugs on fans' heartstrings. What percentage of Lawrence brackets do you suppose have the Jayhawks advancing at least to the Final Four, if not winning it all?

"I'm not going to lie," Rosetta says. "I generally find a way for Kansas to make it to the Final Four, at least."


He figures picking a team with a No. 1 or 2 seed is usually a safe bet for the Final Four. Studying the expert brackets online at sites such as probably wouldn't hurt, either.

"It's definitely a crapshoot," Rosetta says. "If I get a couple of Final Four teams right, I feel good. It's become such a competitive tournament."

So does that mean you'd be just as well off choosing your tournament winners based on some non-basketball metric - say, how fun the town is, or how much you like the mascot?

"You don't have to be a college basketball fan to do well in these," Rosetta says. "I've heard the urban legends of someone choosing by team colors or different combinations, and they end up winning the thing."

Bracket science

Jeremy Chrysler is trying to turn his bracket selection process into a science. Chrysler, a KU alumnus who lives in Kansas City, Kan., runs, a site devoted to KU basketball.

He and his buddies also launched last week. The site offers statistics to help fans fill out their tournament brackets.

His method this year: Take a 10-game average of teams' recent performance using a variety of statistics.

"We think we can get a better idea of how a team is playing right now instead of how they've played over the course of the season," Chrysler explains.

His other suggestions:

¢ Always pick a 12 seed over a 5 seed.

¢ Experienced teams from mid-major conferences often give teams from bigger conferences trouble.

¢ "Overall," he says, "makes your picks different in a way that you can justify."

Holy technique

Or, if you prefer, go with the method his friend's mother tried one year.

"She used to make picks according to the teams that sounded 'holiest,'" Chrysler says. "She advanced every 'saint' team and summarily banished any team with 'demons' or 'devils' in it.

"Duke (the Blue Devils) won that year, so she didn't do that great, but you'll hear about people with oddball picking strategies doing well because they're picking away from the mean. As they say, even a broken clock's right twice a day."

Bottom line, Chrysler says: Making the logical choices probably won't win the pool - especially if everybody else is making the same logical choices.

"The more you do that, and the larger the pool, the more likely you are to be average," he says. "Looking at the stats helps you make educated risks."

But if your goal is to do well in the pool - and not necessarily win it all - KU mathematician Cobb offers this paraphrase from writer Damon Runyon: "The race is not always to the swift, nor victory to the strong, but that's the way to bet."

Looking for some guidance for filling out your NCAA tournament bracket? Try these sites:

¢Hack the Bracket is a site designed by Kansas University fans. It analyzes the average statistics for the last 10 games among tournament teams, with a goal of determining how the teams have been playing lately - not how they did for the overall season.

¢ ESPN's Bracketology site offers a variety of information about tournament teams and matchups.

¢ To see a list of strategies for tournament selections, to go It walks you through various statistical and non-statistical methods.

¢ Two other sites that provide broad information for brackets are Bracketology101 and the Bracket Project.


mom_of_three 10 years, 10 months ago

Two years ago, I picked every first round winner, except for one - Kansas!!

Mike Blur 10 years, 10 months ago

p> is crashed--is it dead?

I tend to make educated guesses while filling out my bracket, but usually wind up in the middle of the pack of whatever pool I am in.

Commenting has been disabled for this item.