How hard can it be to select the NCAA Tournament field?

Over the course of two weeks, The Pitt… How hard can it be to select the NCAA Tournament field?

Over the course of two weeks, The Pitt News followed the selection committee’s principles and procedures to mimic the process of forming the tournament bracket. Using the same rules that the real committee follows, we hoped to achieve a valid, fair bracket and also learn about the challenges of the process along the way.

The 10-person selection committee has three tasks: select 37 at-large teams, seed the at-large and automatic bids No. 1 through No. 68 and, in the final step, arrange the teams in the bracket. The committee spends by far the most time on the first step, just as we did.

Before the committee gathers, members receive a list of all eligible Division I programs in alphabetical order (to eliminate bias). Each member submits an initial ballot in the form of a two-column list: the first column for “at-large selections” and the second for “at-large considerations.”

The first column, in short, is for locks, or, as the guidelines state, “teams that, in that member’s opinion, should be at-large selections in the tournament based upon play to date, regardless of whether the team could eventually represent its conference as an automatic qualifier.”

The second column is for teams with something left to prove. They aren’t in, but they aren’t out. “There is no minimum or maximum limit in the second column; however, only teams meriting serious consideration should receive votes.”

When The Pitt Newssubmitted our ballots in the final week of the regular season, 70 different schools appeared in the second column on at least one ballot.

If a school receives eight first-column votes from the 10-person committee, it is in the tournament. In our six-person committee, we required five votes. Twenty-two teams earned a bid through this step, including Kansas, Duke, North Carolina, Syracuse, Murray State, Temple and New Mexico.

If a school receives more than one vote in either column but not enough first-column votes to receive a bid, it moves to an “under-consideration board.” Call this the bubble. Schools that won or shared their regular-season conference championships are also automatically included on the board, as are schools that a committee member verbally recommends.

Our bubble started with 63 programs. They included Saint Mary’s, UNLV and Vanderbilt, which fell one vote short of receiving an at-large bid through the initial ballot. It also included Long Island-Brooklyn, Stony Brook and five other mid-majors that won regular-season titles but did not receive enough member votes. Other programs — Butler, Buffalo, Old Dominion, Wyoming and 12 others — received a single vote and were left off.

At any point, though, the real committee can add late-rising schools to the under-consideration board with three votes. Marshall received just one mention in our initial ballot, but we added the Thundering Herd after it beat Southern Miss twice in one week and appeared in the Conference USA championship game.

It works the other way, too. At any point, the committee’s commissioner (former Connecticut Athletic Director Jeff Hathaway this year) can request that it review schools to eliminate from the under-consideration board. We had several fat-trimming votes, the first of which purged many mid-major regular-season champions and the last of which purged Oregon, St. Joseph’s and other schools that stumbled in conference tournaments.

As the under-consideration board shifts with conference tournament play, the committee slowly awards at-large bids to schools. The process is tedious and repetitive.

First, each member lists, in his or her opinion, the best eight teams on the under-consideration board. Those votes are tallied, and the eight teams that receive the most are placed on the at-large ballot. Members then rank those eight teams No. 1 through No. 8. The committee totals those ballots, and the four schools with the lowest totals receive at-large bids.

Gradually, members select just six and then four schools at a time based on the number of teams left under consideration. The process cycles, again and again, until the committee has 37 at-large teams.

When and how a team receives its bid does not influence its seed later on.

With each ballot, decisions grow tougher. Gonzaga, UNLV, Vanderbilt, Iowa State, Memphis, Purdue, Creighton and Virginia joined the field easily enough to bring our at-large field to 30. But when ranking South Florida, Brigham Young, California and Texas, all four schools received both first-place and last-place votes, and the difference between California earning a bid and Texas returning to the bubble was a single point.

Some schools lingered on the bubble vote after vote. Alabama and Saint Louis repeatedly received enough votes to appear on the at-large ballot, only for members to rank them in the bottom half. Eventually, though, they made it.

As Selection Sunday neared, schools started to earn automatic bids. Some early automatic bids, such as those to Loyola and UNC-Asheville, didn’t impact our committee, as those schools were not under at-large consideration. Conference tournament wins for Murray State, Creighton and Saint Mary’s, though, moved them from at-large to automatic berths, freeing up bids for other bubble teams.

As Saturday’s games went final and Sunday’s title matchups formed, we knew there would be five to six additional at-large bids available for 14 remaining teams on the under-consideration board: Arizona, Colorado State, Drexel, Iona, Marshall, Miami, Mississippi State, North Carolina State, Northwestern, Seton Hall, Tennessee, Texas, Washington and Xavier. We had these schools ranked No. 1 through No. 14 to allocate bids based on the number available.

Texas, Xavier, North Carolina State, Drexel and Colorado State earned at-large bids from us Sunday. With Xavier in the Atlantic 10 championship game, Seton Hall waited to see if the Musketeers would win and shift to an automatic bid that would free up a berth for the Pirates, the next team in our rankings. St. Bonaventure, though, won in an upset, and Seton Hall did not make our bracket.

To evaluate tournament resumés, the guidelines note that “each committee member independently evaluates a vast amount of information during the process to make individual decisions. It is these qualitative, quantitative and subjective opinions … that each individual ultimately will determine their vote on all issues related to selections, seeding and bracketing.”

Statistical information, including the Rating Percentage Index, can be used by committee members in any manner they choose, if they choose to use it at all.

Discussions with fellow members, athletic directors, coaches and commissioners also influence decisions. But there are strict rules about who can talk about what. Any committee member associated with a school cannot lobby or vote on ballots containing that school. Committee member Jamie Zaninovich, the commissioner of the West Coast Conference, is excused from discussions and votes regarding Gonzaga, Saint Mary’s and Brigham Young.

The committee seeds teams much like it selects at-large teams. Of all the automatic and at-large teams, each member selects the subjective “top eight,” and then the committee settles on four using the same process as governs the at-large ballots. Seed and repeat, No. 1 through No. 68.

The top 16 teams are kept separated as No. 1 through No. 16, as one goal of the bracket is to balance each region’s top teams. No. 1, No. 8, No. 9 and No. 16 are grouped in one region, No. 2, No. 7, No. 10 and No. 15 in another, and so on, as best as possible.

After the top 16 teams, the rest are grouped into fours called “seed lines.” For instance, in our bracket, Ohio State, Davidson, New Mexico State and Montana are all equal on the No. 14 seed line.

The committee can “scrub” seeds and vote to shift a school up or down upon review. We considered scrubbing our initial rankings of North Carolina, Kansas and Missouri as the conference tournaments played out.

Seeding begins before at-large selection ends, and it does not have to occur in numerical order. We seeded through Louisville on the No. 5 seed line and then filled the bottom fourth of the bracket with at-large teams as we waited to decide the fates of bubble teams.

Placing teams into the bracket is an art guided by subjective principles. The committee must spread out teams from the same conference, which is a challenge with Big East schools. Host schools cannot play where they are hosting, which means Ohio State is not allowed to play in Columbus this year.

Regular-season rematches are avoided in the early rounds, as are rematches of previous years’ Tournament games (Syracuse and Vermont, for instance). The top teams cannot have a home-court disadvantage on opening weekend, and all teams must remain as close to their “areas of natural interest as possible.”

To facilitate this, schools can move either one seed line up or down.

Among the top teams, the committee assigns sites in order of true seed. We assigned Kentucky, our overall No. 1, the Louisville site for games Thursday and Saturday. Syracuse received Pittsburgh, Kansas received Omaha and North Carolina received Greensboro.

How did we do? Compare The Pitt News’ bracket on page 12 to the real bracket on page 21.

Brian Batko, Dustin Gabler, Lauren Kirschman, Roger Sepich and Jasper Wilson contributed to this story.