Bracketology from the NYTimes - I hope this helps some ppl.

[Pony_Fan]

http://home.nc.rr.com/wtadams/poologic/poologic.htm

or most of the year, Bradley P. Carlin seems like a nice enough guy to have around. He lives in a suburb of Minneapolis with his wife and three sons, sings in several bands and spends his workday trying to keep people healthy.

Come March, though, Carlin morphs into just about the last person you would want to have in your office. He is a University of Minnesota biostatistician who has published an academic article about how to win an N.C.A.A. basketball pool, and he is not shy about putting his ideas to work.

He won the pool at his wife's office three times in five years. He was then informed that the pool would no longer exist.

"It reminded me of when I was in a band and we wanted a new bass player," Carlin, 42, said. "We told the bass player we were breaking up the band, and then we reformed it with a new name and a new bass player."

The part of Carlin's story that matters to the millions of people who will fill out brackets in the next four days is that his ideas are not a secret. Nor are the theories of a handful of other professors who have done serious research on what might be called poolology.

Poologic.com, a Web site that explains these ideas, claims to have helped people win $250,000 over the years. In return for the advice, the site asks only that winners give some of their bounty to the V Foundation for Cancer Research, named for Jim Valvano - and that people realize they will still probably lose.

Some pool players want to spend as little time as possible filling in their brackets. They may pick friends' alma maters or, like Diane Chambers on "Cheers," the teams that play in places with the prettiest state flowers. Other entrants think they know so much about Kentucky's defensive scheme that they do not need help. Fair enough.

But for anyone in between, there is a system that can help and that offers lessons about basketball along the way. It starts with knowing your competition.

Most pools award so many more points for the later rounds that you must pick the right champion to have a shot at winning. Choosing the team with the best chance of winning it all - Illinois and North Carolina were the top-ranked teams entering this weekend - sounds like the smart thing to do.

The problem is that everybody thinks that way. If you choose a favorite, you leave yourself with long odds of winning a pool, because so many other people will make the same choice.

When Andrew Metrick, a finance professor at the University of Pennsylvania, examined pools one year, he found that almost four of every five entrants chose one of the four teams seeded No. 1. But top-seeded teams have won only a little more than half of the titles since the men's tournament expanded to 64 teams in 1985. (A play-in game featuring a 65th team was added in 2001.)

"I can tell you right now that Illinois and U.N.C. will be overbet," said Tom Adams, a systems analyst for the federal government who runs poologic.com. "What you really have to do to get an advantage is to go for a contrarian champion."

Carlin has a handy way of making sense of this. Imagine that Illinois has a 30 percent chance of winning, roughly what statisticians say the team's odds are. If 30 people in a 100-person pool pick that team as champion, they essentially have to share that 30 percent. So each has a 1 percent chance of winning on average.

Now say that Louisville has a 5 percent chance of winning the tournament, and only one person picks Louisville. That person's odds of victory become 5 percent, too.

The best choices tend to be teams that are seeded Nos. 1, 2 or 3 but that are not among the two or three biggest tournament favorites and that are not local teams for your pool. Wake Forest and Michigan State seem to fit the part this season.

This seems like a good time to consider another question: Isn't all this illegal? Not really. Most states outlaw any game that requires money to enter, involves chance and offers a prize.

But because ESPN.com's online game, which drew 1.3 million fans last year, costs nothing to enter, it is legal. And the police say they are generally happy to ignore pools unless they are intended to make a profit for somebody. "There is very little law enforcement can or will do about the friendly office pool," David Bayless, a spokesman for the Chicago Police Department, said. "That's not much of a concern of ours."

The second big theorem of poolology comes from the way teams are seeded. The formula that the National Collegiate Athletic Association uses does not consider margin of victory. It treats a victory as a victory, just as the tournament does.

But anybody who has studied past seasons - in basketball or other sports - knows that a team that has won a lot of squeakers typically does not have as bright a future as one that has blown out opponents. It is tempting to believe that teams that win many close games, as Arizona, Boston College and Gonzaga did this season, have the character to keep on doing it. Usually they do not.

On the other side, Florida and Louisville have won enough blowouts to be tempting choices to reach the final rounds. The seeding system can also give short shrift to schedule strength. The tough schedules of Texas and Georgia Tech suggest they may deserve more respect than they will receive, provided they get in.

These distinctions matter much more in pools that give a bonus for picking upsets, as the Sports Illustrated online game does. An 11th-seeded team that looks almost as good as its sixth-seeded opponent then becomes a smart first-round pick. These bonuses gave Carlin his winning margin in his old pool, he said.

The simplest way to find these games in the opening round is to look at the Las Vegas point spreads. Using those betting lines may make you feel sketchy, but economists see them as packed with information, just like stock prices. The spreads are more accurate predictors of games than any computer rating - in fact, than all the ratings combined, according to research by Ray C. Fair, a Yale economist, and John F. Oster. They studied college football, but the point almost certainly holds for basketball.

There is one more lesson to keep in mind, at least if you live in Chicago.

"We'll investigate anyone," Bayless, the police spokesman, said, "who isn't taking Illinois to win it all."