With the NCAA selections coming up Sunday night, a recent discussion thread on Ultimate College Softball reminded me of how the tournament seedings don't always match up with the media/coaches' poll rankings. I thus decided to make the following chart, which compares the past three years' final regular-season ESPN.com/USA Softball poll rankings (which I chose based on their nice archiving; see 2010, 2009, and 2008) and NCAA tournament seedings (2010, 2009, 2008). You can click on the chart to enlarge it.
As can be seen, I divided teams into three categories: comparably high in the poll rankings and NCAA seedings (+/- 1 of each other); seeded markedly higher than their poll rankings; and seeded markedly lower than their poll rankings. At the very top levels, the seedings and poll rankings tracked very well in 2008 and 2009, but much less so last year.
Though we obviously have a small sample, it seems that of the two, poll rankings may be a better indicator of likely tournament success than the seedings. Most notably, the 2008 Arizona State squad that won the Women's College World Series was ranked No. 2 in the poll, but only seeded sixth. That same year, Virginia Tech, which made a surprise appearance in the WCWS, had a poll ranking of 16th (potentially warranting a No. 16 seed), but was unseeded. Louisiana-Lafayette, another surprise 2008 WCWS participant, was ranked 17th, but unseeded.
Arizona, which made the final championship series last year against UCLA, had its finish better forecasted by its No. 6 poll ranking than its 10 seeding. The Wildcats were also seeded somewhat worse than their ranking in 2009.
One team, the University of Georgia, was seeded more favorably than its final regular-season poll position would warrant in all of the past three seasons.
Some have suggested that the RPI rankings, a computer-based system that takes into account teams' strength of schedule, have informed the NCAA seedings more strongly than the weekly polls (here supposedly are last year's RPI rankings).
I don't claim to know all that is going on. But, as you're watching the Selection Show, don't be surprised to see your favorite team seeded higher or lower than what you may have been expecting based on the polls!
As can be seen, I divided teams into three categories: comparably high in the poll rankings and NCAA seedings (+/- 1 of each other); seeded markedly higher than their poll rankings; and seeded markedly lower than their poll rankings. At the very top levels, the seedings and poll rankings tracked very well in 2008 and 2009, but much less so last year.
Though we obviously have a small sample, it seems that of the two, poll rankings may be a better indicator of likely tournament success than the seedings. Most notably, the 2008 Arizona State squad that won the Women's College World Series was ranked No. 2 in the poll, but only seeded sixth. That same year, Virginia Tech, which made a surprise appearance in the WCWS, had a poll ranking of 16th (potentially warranting a No. 16 seed), but was unseeded. Louisiana-Lafayette, another surprise 2008 WCWS participant, was ranked 17th, but unseeded.
Arizona, which made the final championship series last year against UCLA, had its finish better forecasted by its No. 6 poll ranking than its 10 seeding. The Wildcats were also seeded somewhat worse than their ranking in 2009.
One team, the University of Georgia, was seeded more favorably than its final regular-season poll position would warrant in all of the past three seasons.
Some have suggested that the RPI rankings, a computer-based system that takes into account teams' strength of schedule, have informed the NCAA seedings more strongly than the weekly polls (here supposedly are last year's RPI rankings).
I don't claim to know all that is going on. But, as you're watching the Selection Show, don't be surprised to see your favorite team seeded higher or lower than what you may have been expecting based on the polls!