# Is there a science to bracketology?

The Selection Process
Mark Jones of the University of Dayton Flyers rebounds the ball amidst the defense of the University of Tulsa Golden Hurricanes during the first round of the NCAA Tournament on March 20, 2003, in Spokane, Wash.
Jonathan Ferrey/Getty Images

Each NCAA basketball season starts with as many as 345 Division I men's teams and 343 women's teams. Don't let that number panic you, though. Before the start of the tournament, a selection committee of athletic directors and conference commissioners narrows those 300-plus teams down to 68 (64 for women's basketball). From there, it's up to you to place the teams in the brackets.

There are 32 conferences in NCAA Division I men's basketball. Each conference, with the exc­eption of the Ivy League, has a championship game. The selection committee automatically extends a tournament invitation to the winner of each championship. The committee then selects the other 37 teams, known as at-large teams, by Selection Sunday, when it reveals its choices.

The committee is sequestered in a hotel beginning the Thursday before Selection Sunday to determine the at-large teams. Each member has watched specific teams during the season and speaks about those teams about during deliberations. Committee members can't, however, speak about their own school or conference and have to leave the room when those teams come up. All decisions made by the committee are final, and there's no appeal process.

The selection committee uses several criteria to choose the teams:

• Ranking in the national polls
• Conference record
• Wins versus ranked opponents
• The way a team finishes its regular season
• Rating Percentage Index (RPI)

Most of these are self-explanatory, but the RPI is a little more complicated. It's a mathematical equation applied to a team to determine the strength of its schedule and how well it played that schedule. The committee takes into account how many games the team won, how many games its opponents won and how many games the opponents of its opponents won. The formula is:

• 25 percent winning percentage
• 50 percent opponents' average winning percentage
• 25 percent opponents' opponents' average winning percentage

In 2004, the formula changed to give more weight to teams that win on the road.

After choosing 68 teams, the selection committee seeds, or ranks, them. It divides the teams into four regions -- North, South, East and West -- each with 16 seeds. Four regions with 16 slots equal 64 teams, so the last eight teams selected play one game for a chance to get the final no. 16 spot in each region. The women's tournament skips this step, since only 64 teams are chosen.

The committee also has to keep the location of the schools in mind. Whenever possible, the committee must put each team, especially the top seeds, in the region closest to its school. This is so more fans can attend the games, which will raise more money. There­'s no limit to how many teams from one conference can play in the tournament, but the first three teams from a conference have to be in different regions.

Once the tournament begins, the committee members still have work to do. Each member follows a particular region, and all of the committee members attend the Final Four games and the championship game. About a month after the tournament is over, the committee comes back together to evaluate its work. The next year, the process begins all over again with a new committee.

Read on to learn about what happens when the tournament starts and how to boost your odds as you fill out your bracket.