Skip to main content

View Diary: Rating the 2010 election forecasters (54 comments)

Comment Preferences

  •  But you're working with just a subset of races (0+ / 0-)

    You're only using the 49 races that Bowers selected using a seemingly arbitrary criterion. This selectiveness may not matter, but whether or not it does needs to be shown (as I observed earlier).

    There were in all 37 Senate contests and 37 Governors races -- 78, not 49. In addition, there were 435 House races that Silver called but the others did not. (I think Pollster called some of them but not all of them.)

    "Everyone is entitled to their own opinions, but not to their own facts" -- Daniel P. Moynihan

    by zackamac on Sun Dec 19, 2010 at 08:54:00 PM PST

    [ Parent ]

    •  I know... (0+ / 0-)

      I was just working with the data provided... I don't have a stake in this as I was just bored and felt like running some analyses.  That said, however, 49 is an adequate sample of 78 so the analyses should still hold for 538, Pollster, and RCP.  I don't know about Bowser as he hasn't indicated the reasoning behind his arbitrary cut-off.  

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site