The world knows UConn is the favorite. But the Huskies' edge is less than most might think.
ESPN's brand-new Basketball Power Index (BPI) for women's college basketball gives Geno Auriemma's team a 37 percent chance to win this year's NCAA tournament, one year after the Huskies came up short in the Final Four. (You can read a little bit about the methodology behind our model here).
That would be overwhelmingly favorable odds for virtually any other team in any sport but for UConn? In women's basketball? It's ... just 37 percent?
That's more than double the next team's chance to win -- Notre Dame at 16 percent -- but it probably doesn't dull the shock value for the many who figure the Huskies hanging another banner this year is essentially a foregone conclusion.
So why is our model, relatively, down on the Huskies?
Mainly: their draw. Though the top overall seed in this year's field after a thus-far undefeated season earned the Huskies the top rank in Strength of Record, Auriemma's team somehow managed to land reigning national champion and BPI's sixth-ranked team South Carolina in its region. While UConn has virtually no chance of losing its first couple of games, its road gets considerably tougher at the Elite Eight stage. The Huskies have "only" a 78 percent chance of beating the Gamecocks, and waiting for them in the Final Four is likely Notre Dame, the second-best team in Division I, per BPI.
Though there's a solid chance that the Irish and Huskies would have squared off no matter where they were located in the bracket, being on the same side of the Final Four puts the likelihood of them meeting at 42 percent. It would be the fifth time they have met in the national semifinals, with Notre Dame winning three of the previous four meetings.
Ultimately, the Huskies have a 51 percent chance to reach the title game. Waiting for them there could be any number of teams, but Baylor, a No. 2 seed on the other side of the bracket, is the most likely option at 17 percent.
Despite what may seem like negative rhetoric above on UConn, the Huskies are still far and away the best team in women's college basketball. Athletic forward Gabby Williams continues to be productive and has emerged as a star while Katie Lou Samuelson battled through injuries to earn AAC player of the year honors.
A group of contenders
Because Connecticut only has a 37 percent chance to win it all, that leaves quite a bit of the pie for everyone else. Notre Dame's 16 percent is closely followed by Baylor, also at 16 percent. Winners of 28 straight, Kim Mulkey's squad won its ninth Big 12 tournament title with ease (+27.7 average margin of victory) despite losing senior point guard Kristy Wallace to an ACL tear in the regular-season finale. Since winning their second title in 2012 the Lady Bears have failed to reach the Final Four, despite earning a No. 1 or No. 2 seed every tournament. They share a region with Louisville and Oregon State, teams that ended Baylor's hopes of a third title in 2013 and 2016, respectively.
The Irish also look to make a run without their star, or stars, as four of their players have fallen victim to ACL tears this season. Last year they managed to reach the Elite Eight despite losing their leading rebounder and second-leading scorer, Brianna Turner, in the second round to the knee injury. Notre Dame's three losses have come to other No. 1 seeds, with two games decided by single-digit points. With her team down to seven scholarship players and three walk-ons, Muffet McGraw's battle-tested group looks to reach the title game for the fifth time in the past eight years.
Mississippi State, which took down UConn a year ago, has an 11 percent chance while Louisville -- which has to wrestle with the Lady Bears in its own region -- is at 8 percent. The Gamecocks have only a 3 percent shot due to the high likelihood of running into the Huskies and losing. A'ja Wilson, the 2017 Final Four Most Outstanding Player, put up 20 points in South Carolina's Feb. 1 meeting against the Huskies, but her teammates shot 24 percent from the field as the Gamecocks lost by 25 points. Aside from UConn and Tennessee, USC is the only team in women's NCAA tournament history to successfully defend its title, winning the second and third editions of the tournament in 1983 and '84.
Women's hoops has a reputation of being more favorite-heavy than the men. So will all four No. 1 seeds cut down the nets en route to Columbus?
It's possible, but unlikely. BPI says there's only an 8 percent chance of that happening, partially due to the fact that the third-best team in D-I, Baylor, is not a No. 1 seed. However, there is a 76 percent chance that at least two of the Final Four teams are No. 1 seeds.
On the other end of the spectrum, could any of the No. 1 seeds fall in the first round? If it's going to be anyone, it likely would be Louisville, which has a 4 percent shot to lose to No. 16 Boise State.
Scoring title at risk?
While BPI can't explicitly predict whether Ohio State's Kelsey Mitchell will score the 31 points needed to pass Jackie Stiles for second among all-time scorers, or the 165 needed to beat Kelsey Plum for the record, we can make a guess about her chance to play in the requisite games. Mitchell has scored 24.5 points per game this year, so she has a pretty strong shot of reaching the first milestone if Ohio State can beat George Washington in its opening game (83 percent chance).
Mitchell would have to outperform her per game average to break the record even if the Buckeyes reached the title game, and they only have a 0.5 percent shot to get there.
The Women's College Basketball Power Index is an ELO-based measure of team strength designed to predict team performance going forward. Entering the season it assesses each team's predicted strength based on a regressed version of performance in prior seasons. Once games start, each team's rating is adjusted up or down based on game results. Winning a game by more points and/or against a better opponent results in a larger adjustment.
Strength of Record (SOR) is a measure of team accomplishment based on how difficult a team's W-L record is to achieve, derived from opponent's ELO ratings. Game predictions account for opponent strength and site and are used to simulate the season 10,000 times to produce season and tournament projections.