from Amazon
Friday, August 21, 2009
Near-sighted Marcels
John Burnson calculates the Marcels according to the monkey and compares it to the Marcels according to how he thinks humans would do it. While Marcel weights things at around 1.00/.80/.60, John figures that humans weight things at 1.00/.19/.06. He then gives us a list of the “sentimental” picks (led by Jason Bartlett) that basically means humans will think more of him than the monkey will.
I love the idea. I don’t think that John’s guess is accurate though in terms of the weighting. It’s easy enough to figure out what it is though, as I did here a few years ago:
Running a regression analysis on the individual Forecasters, and averaging the results, the weightings used by the Forecasters was 60/21/18 (for years 2002/2001/2000). Taking the Primer Readers as a group, their weightings are 64/19/18. Essentially, not that much of a difference.
So, taking a little sabermetric licence, I would say the weighting of humans would be on the order of 1.00 / 0.50 / 0.25. But, like I said, it’s easy enough to figure. Get the community forecasts, run a regression, and you get your answer. I posted the results of the community forecasts a year or so ago on Google Docs, if someone is tempted to do the work…
Recent comments
Older comments
Page 1 of 344 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date