The ten best women's programs take a third as many players from Ohio and Georgia as the rest of Division I, and half as many as the programs next door to them. The players who do come out of those states play as much as their own teammates. For every program outside the top 10, that is the same quality with less competition.
The gap
Each season we take the ten Division I programs with the most All-Americans as the top 10: North Carolina, Maryland, Florida, Loyola, Boston College, Syracuse and Northwestern most often. Since 2015, 0.7% of their players have come from Ohio and 0.9% from Georgia. Across the rest of Division I the figures are 2.3% and 2.6%.
Some of that is geography: the top 10 sit mostly on the East Coast. So we compared each top-10 program with the three non-top-10 programs nearest to it, which recruit from the same map. Those neighbours took 1.7% of their players from Ohio and 2.0% from Georgia — still more than twice the top 10's share.
Share of roster from Ohio and Georgia, women's Division I
North Carolina, in the top 10 for nine seasons, took no one from either state; its nearest neighbours took 3.0%. Maryland, also nine seasons: none, against 2.2%. Florida, next door to Georgia, took 4.3% against its neighbours' 12.1%. Northwestern: 0.5% against 6.1%. Draw any ten programs a season at random and compare them with their own neighbours in the same way, and the gap averages zero. In 2,000 random draws, one as large as the top 10's came up six times.
Measured against their teammates
A gap is only an opportunity if the players are good. Comparing states on raw playing time is misleading, though: a player at a weaker program starts more easily. So we compare every player with her own teammates in the same season. If a state's players start more often than the women they share a locker room with, that is not about which programs they landed at.
How often players start compared with their own teammates
Ohio's players start 0.8 points more often than their teammates, and Georgia's 3.0 points more. Neither is beyond chance, and neither is a warning sign. By the one measure that controls for where they play, they are ordinary Division I players — as good as the women around them.
Whether the top 10 are missing stars is a different question, and the data cannot answer it. Ohio and Georgia have produced two All-Americans each since 2015. Their teammates' rate would predict 2.4 for Ohio and 4.7 for Georgia. Numbers that small say nothing either way. The only way to find out whether Ohio produces top-10 players is for a top-10 program to recruit there.
What does stand out
Two states separate from their teammates by more than luck explains. New York's players become All-Americans 38% more often than their teammates — 173 selections where their teammates' rate predicts 125 — and they sit on the bench less. The top programs' appetite for New York is not a habit; it is paying off.
Illinois players start 11 points less often than their teammates. Texas and Michigan look similar, at about 9 points below, but with fewer players that could still be luck.
What to do with it
Programs outside the top 10: Ohio and Georgia offer players who hold their own, in a market the top 10 barely enter. That is the same quality with less competition for it.
The top 10: the gap is a habit the data can neither justify nor condemn. A few visits would settle it cheaply.
How we counted
- Top 10: the ten Division I programs with the most All-American selections each season, win percentage breaking ties; 2020 and 2021 left out (COVID).
- Neighbours: the three nearest Division I programs that were not top 10 that season, by distance between their cities; programs in the same city tie, and all tied programs count.
- Against teammates: for each player-season, her teammates' starting rate that season (excluding her); a state's figure is its players' starts minus that expectation, per season. Starter means started at least 60% of the team's games. Seasons under 10 games left out.
- All-Americans: IWLCA First, Second and Third Teams, 2015–2025, matched to rosters by name, season and division, and compared with the rate of the player's own teammates.
- Chance: each state against thousands of random draws of the same number of players. A result counts only if it holds in either direction after allowing for all the states tested.