Behind The ScenesESPN College Baseball

ESPN researcher Jeremy Mills makes another bid at college baseball Bracketology ‘perfection’

ESPN Stats & Information's Jeremy Mills. (Courtesy of Jeremy Mills)
ESPN Stats & Information’s Jeremy Mills. (Courtesy of Jeremy Mills)

Perfection is imperative for ESPN researchers responsible for supplying facts and nuggets for use on air or in print. Jeremy Mills from ESPN Stats & Information is seeking perfection in another, more forgiving area: college baseball Bracketology. This is the fifth year he has projected the field of 64 teams in the NCAA Division I Baseball Championship.

Mills’ most recent Bracketology is available on ESPN.com. His final projection of the season will be available Sunday before the official brackets are released Monday at noon ET on ESPNU. He spoke with Front Row about how it all started, the projecting process, fellow bracket generator, Joe Lunardi, and more.

How did you come up with the idea of Bracketology for college baseball?
It seemed strange that the network with the most postseason coverage wasn’t predicting the brackets. So I pitched the idea to ESPN.com editor Lauren Reynolds and went from there.

What is your process for bracketing out the 64 teams?
There are only five delineations — eight national seeds, eight additional No. 1 seeds and 16 each Nos. 2-4 seeds — so I usually start by bucketing teams rather than creating a true S-curve like you see for basketball. With the committee, there’s no difference between being the 17th and 32nd best teams. In baseball, they can pair any No. 1 seed with any No. 2 seed, as long as they aren’t from the same conference, which is a stark contrast to the balancing of strength across regions in basketball.

I’m looking at a variety of numbers to create the buckets. RPI drives the seeding a lot more in baseball than in other sports, so the original grouping starts there. But other considerations are records against various RPI categories, record in weekend series (which carry more weight than midweek games since that’s the postseason format) and standing within the team’s conference.

As for particular location assignments, the NCAA prefers to keep teams as close to home as possible. Even knowing the 16 sites, it’s hard to get all four teams at a location correct. Usually I get one or two of the 16 regionals completely correct and three of the four teams in a handful [of other regionals].

What is your success rate?
I’ve had 63 out of the 64 twice. Two years ago, I missed two and both were cases of the committee rewarding teams from mid-major conferences for good seasons over lower teams in the power conferences.

Do you exchange ideas and concepts with Joe Lundardi about his Bracketology?
I’ve had the opportunity to work with Joe on shows as part of my “day job” in research, but we’ve never discussed compiling brackets. Maybe I need to pick his brain the next time we’re in the same room.

Back to top button