The Atlanta Journal-Constitution
Top-seed upsets indicate parity in women’s hoops
In just the second year of using the March Madness branding, the women’s NCAA Tournament lived up to that name in a way it rarely has during its opening weekend. No. 8 seed Mississippi knocked off No. 1 seed Stanford at Maples Pavilion on Sunday night. Less than 24 hours later, No. 9 seed Miami pulled off its own stunning upset, eliminating No. 1 seed Indiana in a 70-68 thriller at Assembly Hall.
The early ouster of two No. 1 seeds is shocking not only because Indiana and Stanford were popular picks to make Final Four runs but also because No. 1 seeds rarely get eliminated at this stage in the tournament. This year is only the second time since the tournament expanded to 64 teams in 1994 that multiple No. 1 seeds failed to make the Sweet 16 (the first time was in 1998).
Part of the reason upsets haven’t occurred often in the women’s tournament is the home-court advantage enjoyed by the top four seeds in each region. It’s hard enough beating a top-tier opponent, never mind trying to accomplish the feat in front of thousands of opposing fans. Before last weekend, only two No. 1 seeds since 1994 had not made the Sweet 16 while playing on their home court. Further, teams seeded No. 13 or lower have won only two true road games in the tournament’s history. Another factor is the historical dominance of a handful of programs, most notably Connecticut and Tennessee.
Are the events of last weekend
an anomaly that fans should expect once every 25 years, or does this represent a new normal? Data measuring parity suggest it’s the latter. The gap between the upper echelon of women’s teams and the middle of the tournament pack is shrinking, creating an environment tailor-made for last weekend’s upsets.
There are multiple ways to measure parity. One method involves using Her Hoop Stats Rating, a predictive metric of team strength based on machine learning. It predicts a team’s pace-adjusted performance on a neutral court against a hypothetical average team. For example, South Carolina boasts a Her Hoop Stats Rating of plus-55.5. This suggests that South Carolina would be expected to outscore a hypothetical average opponent — with a Her Hoop Stats rating of zero — by 55.5 points per 100 possessions on a neutral court.
The difference between the average Her Hoop Stats Rating of No. 1 seeds and No. 8 seeds in a single tournament offers insight into the relative strength of mid-tier tournament teams that year. The smaller the difference, the greater the parity and the higher the likelihood of headline-grabbing upsets like those this weekend. There has been a downward trend in this difference since 2010 (the beginning of the Her Hoop Stats tournament database), culminating in a significant decrease this year. The talent gap between a No. 1 seed and teams in the middle of the tournament pack is as small as it has been in recent years.
The Associated Press weekly poll provides another measure of parity. The amount of movement in the polls week to week is a signal of how similar the best teams in the country are. If there were wide gaps between each team’s talent, there would be no upsets and therefore no movement — in the same way South Carolina has been No. 1 all season. If the top 25 teams all had identical talent, fans would observe a constant reshuffling of the AP poll as teams win and lose all season against comparable foes.
AP polls have shown noticeably more week-toweek movement in the past two seasons, reinforcing the idea that teams in the top 25 (and just outside the top 25) are more closely matched now than in any season since at least 2009-10.
The AP polls and Her Hoop Stats ratings each indicate that parity among the country’s top 30 teams is at its highest level in Division I women’s basketball since at least 2010. This finding helps make sense of what transpired last weekend.