Connecticut and Kentucky entered the NCAA tournament as No. 7 and No. 8 seeds. Participants in the ESPN Tournament Challenge weren’t wild about the Wildcats or the Huskies. Just 2.1 percent of them picked Kentucky to make it to the finals and just 0.4 percent had Connecticut. Tomorrow, these two teams will compete for the national championship. Just how unlikely is this bracket-busting matchup?
The question is a little trickier to answer than you might think.
In one sense, at least two things about tomorrow’s game are unprecedented. Every national championship game since the men’s tournament first began to be seeded in 1978 has featured at least one team seeded No. 3 or better; this year broke that record. Furthermore, Connecticut is the first No. 7 seed to make the finals out of 144 teams who had the opportunity. (Kentucky is the fourth No. 8 seed to do so, out of 144 attempts.)
The two teams were stronger than their seed lines might suggest, however. Before the tournament began, the FiveThirtyEight forecast model had Kentucky ranked as the 11th-strongest team nationally — equivalent to a No. 3 seed. It had Connecticut as the 22nd-best team, which would have slotted the Huskies as a No. 6 rather than a No. 7 seed.
Of course, underseeding these teams didn’t make their route to the championship any easier. Instead, each one had to overcome obstacle after obstacle. In our pre-tournament odds, Kentucky had just a 3.9 percent chance of reaching the national championship game and Connecticut’s chances were just 2.0 percent. The probability of both Connecticut and Kentucky reaching the championship was even more remote, of course. It can be calculated by multiplying their odds together: 3.9 percent times 2.0 percent is 0.078 percent, or about one chance in 1,280.
But any individual matchup is inherently unlikely because there are many possibilities — and there is so much parity — in the tournament. The championship matchup that the FiveThirtyEight model deemed most likely to occur before the tournament began was Louisville against Florida. But even that had only a 6.3 percent probability of happening (about one chance in 16).
Instead, the question we’re really interested in is how unlikely it was that two relative longshots would make the championship game together, regardless of the identity of the teams. (If the championship featured, for instance, No. 7 seeded Oregon against No. 9 Pittsburgh, or No. 7 New Mexico against No. 8 Gonzaga, those would also seem pretty amazing.)
There are a couple ways we might approach this calculation. For instance, what were the odds that no team seeded No. 3 or higher would make the championship? Despite this never having happened, the FiveThirtyEight model put the probability at 11.6 percent before the tournament. That’s partly because there were some exceptionally strong No. 4 seeds this year in Louisville and Michigan State.
What about the prospect that both championship teams would have been seeded No. 7 or higher? Those odds were much longer; just 0.9 percent.
Or, we can ignore the seeds since Connecticut and especially Kentucky were atypically strong for their seed lines. Nevertheless, they were among 56 teams that began the tournament with less than 5 percent chance to make the title game, according to the FiveThirtyEight model. We can think of this group of 56 as “The Field.” It includes some teams like Connecticut, Kentucky and Ohio State that the model gave a fighting chance despite their middling seeds, some others like No. 3 seed Iowa State that the model regarded as overseeded but nonetheless viable, and a couple dozen truly hopeless cases like Eastern Kentucky.
The Field’s prospects of advancing at least one team to the championship weren’t so bad: 40.6 percent. But having both finalists come from The Field was more unlikely: the chances were only 4.9 percent.
In other words, Connecticut and Kentucky’s individual accomplishments are impressive but far from miraculous. Teams that faced longer odds, such as Butler in 2011 and Villanova in 1985, have made the championship game. But having two such teams do so in the same year is more remarkable. Connecticut and Kentucky have already made history together, regardless of Monday night’s outcome.