Bracketology

[dc_x]

The bracket matrix site has their final rankings of all the bracketologists. They compared the final projections of 89 different bracketologists. The use a point system which gives points for correctly including teams in the field and the accuracy of the seeding.

Here are results for the more notable sites:

1. Foxsports.com (Jordan Schwartz)
13. CNNSI.com (andy Glockner)
37. Yahoo sports
49. CBSsports.com (Gary Parrish)
53. ESPN.com (Joe Lunardi)
58. Bracketology101.com (these guys were ranked #1 the last few years)
73. CBSsports.com (Jerry Palm)
77. Rivals.com
89. Collegerpi.com (dead last)

Keep this in mind next year when looking at projections.

[sirthought]

Thanks for the info. Link please?

[XUglow]

Ole Joe gets a lot of pub, money, and air time to be #53.

[dc_x]

http://bracketproject.50webs.com/matrix.htm

[drudy23]

Doesn't Palm run College RPI? So does he have a different bracket for CBSports.com than he does for CollegeRPI.com?

[whitesox]

It will be interesting as we get more and more years of data, but my gut instinct is that most of the difference between the bracketologists here is luck. With somewhat subjective standards and a committee that changes every year, no one can get it down to a science. For example, look at the previous #1 dropping all the way down to 58th.

I think the lesson to get out of this is that no bracket prediction will be without flaws- it is best to look at a mix of them to get different perspectives.

I think the problem with a lot of the guys on ESPN is that they get too wrapped up in their own world- they only look at Lunardi's bracket projections. Had they looked at anything else, it would not have been surprising that UAB and VCU got in- several people projected them to be in or right in the mix. Same goes with Colorado being left out- it should have been clear that they were right on the fringe, but they saw that Lunardi thought they were safe and didn't think about it beyond that.