Monday, November 2, 2009

As the preseason pick to finish sixth, the signs are good for the Dutchmen (Or: I am a dork)

Some nerds break into sorority houses for fun. I crunch CAA numbers. I am a winner.

On Thursday, your good friend and mine, Noted Hofstra Hater Mike Litos, poked fun at himself for spending his spare time crunching the four-year statistics of the CAA’s coaches. I can’t say I set out to top him in terms of dorkdom, but as usual, whenever I’m in the room (virtual or otherwise), nobody else has to worry about being the least cool person in it.

For the last couple weeks, I’d been wondering about the accuracy of the preseason CAA polls that are released every year at media day. So there I was Saturday night, spending Game Three of the World Series digging through the archives of CAA athletic sites in search of those polls dating back to 2001-02 (thank you Delaware and—grumble grumble—George Mason; so this is what it feels like when your sworn enemy stops to help you change a flat tire) so that I could compare the predictions with the real thing.

(You’ll soon see what struck my fancy during Game Four last night. Crap, what am I going to do once the Yankees close this out tonight? Find a real job?)

My most notable finding (and I swear I wasn’t poring over the numbers until I found a thesis I liked): As the sixth “seed” in the preseason poll, history is on the side of the Flying Dutchmen. Since the CAA expanded to 10 teams in 2001-02, the sixth seed is the only team among the top eight whose average finish is better than its predicted finish. The average finish of the sixth seed is 4.88, which places it fifth overall.

The sixth seed has finished fifth or better five times, including first (VCU) in 2006-07. That’s one of only two times in the last eight years the regular season champion was picked lower than second in the preseason (UNC Wilmington was the fifth seed in 2005-06).

In addition, the sixth seed has finished as low as ninth just once (James Madison in 2001-02). Among the top seven seeds in preseason polls, only the one seed can also claim it has never finished more than three spots below its forecasted slot.

Most interesting of all: The preseason sixth seed has finished in eight different spots in eight years, but never actually sixth.

My methodology went like this: I went through the last eight seasons and awarded teams the difference between their predicted finish and their actual finish. So last year, when the Dutchmen were picked seventh but finished fifth, they earned two points. Delaware had the biggest negative gap between prediction and reality: The Blue Hens were picked fifth and finished ninth, so they received minus-four points.

I’m not sure, but this may have actually been a worthwhile, real-world use of algebra. If my ninth-grade algebra teacher Mr. Metro could see me now!

Of course, like all things me and mathematics, this is not perfect. The first seed obviously can’t finish any higher than first, and any finish below first would automatically mean the collective first seed has not met expectations. However, the first seed still has the best average finish.

Below you’ll find the identities of the preseason sixth seed since 2001-02 and their eventual finish as well as the overall rankings. There’s a lot more data where this came from—in addition to figuring out the projected and actual finishes for all 12 seeds, I also ran the numbers on all 12 schools to try and determine which teams most often exceeded and failed to meet expectations; that was a really imperfect yet still interesting endeavor—and I’ll unveil that less Hofstra-specific information later in the week.

2001-02: James Madison (9th)

2002-03: James Madison (7th)

2003-04: Old Dominion (4th)

2004-05: UNC Wilmington (3rd)

2005-06: Northeastern (5th)

2006-07: VCU (1st)

2007-08: UNC Wilmington (2nd)

2008-09: Georgia State (8th)


(Predicted finish followed by average finish)

1: 2.13

2: 3.38

4: 3.75

3: 4.38

6: 4.88

5. 6.5

7: 7

9: 7.38

8: 8.13

10: 9.13

12: 9.5

11: 9.75

Email Jerry at or follow Defiantly Dutch at

No comments: