The No. 1 overall seed in the NCAA tournament bracket may seem like a reward. This season, however, the Duke Blue Devils were royally hurt by the way the bracket turned out -- and I have the numbers to prove it.
Before the bracket was announced, ESPN's Basketball Power Index projected that Duke would have a 56.5 percent chance to reach the Final Four. After the bracket release, that number drops to 49.5 percent, a decrease of 7 percent. There are several reasons Blue Devils fans should be upset about Duke's draw. First is the appearance of the Michigan State Spartans -- the fourth-best team in the country, according to BPI -- as the No. 2 seed in the East Region.
But even before facing Michigan State, Duke may have to face the No. 4-seed Virginia Tech Hokies in Washington, D.C. BPI views Virginia Tech as being as good as a 3-seed, and don't forget that the Hokies already beat Duke (which was minus Zion Williamson, granted) in the only matchup between the two schools. Should they reach the nation's capital, the Hokies would have fewer miles to travel to Washington than the top-seeded Blue Devils.
Virginia, on the other hand, was a winner in the bracket release. The Cavaliers' chances to reach the Final Four increased from 61 percent before the bracket release to 67 percent after, largely due to drawing Tennessee, the lowest rated 2-seed in BPI. To be fair, the model thought Virginia was the most likely choice as the No. 1 overall seed because the Cavaliers are first in every advanced metric listed on the team sheets. Despite Duke getting the No. 1 overall seed, Virginia got the No. 1 overall treatment.
Meanwhile, North Carolina's chances to reach the Final Four increased 11 percent to 41 percent after the bracket reveal and Gonzaga's chances increased 13 percent to 65 percent. These increases are largely explained by the fact that neither team has Michigan State in its region.
Outside of the 1-seeds, Maryland and Virginia Tech were both winners of the bracket reveal. Maryland could potentially play Temple, which is only the 72nd best team in the nation, per BPI, in the first round. In the second round, Maryland is looking at a potential matchup against LSU, which is only as good as a 6-seed in the eyes of BPI. Maryland's chances to reach the Sweet 16 were 20 percent before the bracket was announced, but up to 36 percent after.
Virginia Tech has a likely game against Duke, but as a 4-seed, getting to the Sweet 16 is what most people expect. To that end, Virginia Tech went from a 48 percent chance to reach the Sweet 16 before the bracket was revealed, up to 63 percent after it was revealed. The main reason for this bump is that Virginia Tech opens with Saint Louis, the 98th-best team in the country, according to BPI, which is worse than all other 13-seeds as well as two 14-seeds.
Methodology: BPI runs 10,000 simulations of the entire season. Included in these simulations is a selection and seeding model based on various metrics such as BPI, Strength of Record (SOR), adjusted net efficiency (similar to KenPom), quality wins, quality losses and conference championships. The model assigns probabilities to what the NCAA selection committee will do in bracket decisions, entirely based on what it has done in the past 11 seasons.
For this exercise, we compared the tournament odds of what the Basketball Power Index thought the committee would do after all conference championship games were completed, to what actually happened when the field of 68 was set.
For full BPI tournament projections visit espn.com/bpi and for other analytics content visit espn.com/analytics.