- Joined
- Jul 9, 2008
- Messages
- 30,277
- Reaction score
- 17,796
- Gender
- Male
- Political Leaning
- Centrist
Re: FiveThirtyEight Senate Forecast: GOP Is Slight Favorite in Race for Senate Contro
Perhaps we have a different way of looking at this. My base assumption is people are often selfish, deceitful, manipulating creatures. As such I rarely take at face value that social science results are merely promoted on the basis of sound reasoning and data (including this thread). While some were much more willing to like Nate because of his approach with data (frequently these were people on the periphery or inside the social sciences), I think that was in the minority.
He has a pretty good track record, but as Democrats are much more willing to point out now, his legislative branch predictions had some issues. I was interested that there was much less play about Silver's Senate race predictions, and his follow-up response to issues with his model, than the Presidential race. Few were willing to go out and tease out his statements that his model relies on a multitude of polling data, and that some of these races had far less substantial polls than desired. What seemed to matter to these people more was that at least the President's race was covered to their liking, and it got to turn Republican nay-sayers sour.
That being said, a substantial portion of people looking at polling data of any sort will often revert to basic cliched responses, because it either validates or contradicts one's hopes, aspirations, and expectations.
While I am sure there was some of that, I think the biggest thing was a proven track record. While I am not happy at this latest projection of his, I can't argue against it based on anything more concrete than I have my doubts this far out. I still have a great deal of respect for his ability and models.
Perhaps we have a different way of looking at this. My base assumption is people are often selfish, deceitful, manipulating creatures. As such I rarely take at face value that social science results are merely promoted on the basis of sound reasoning and data (including this thread). While some were much more willing to like Nate because of his approach with data (frequently these were people on the periphery or inside the social sciences), I think that was in the minority.
He has a pretty good track record, but as Democrats are much more willing to point out now, his legislative branch predictions had some issues. I was interested that there was much less play about Silver's Senate race predictions, and his follow-up response to issues with his model, than the Presidential race. Few were willing to go out and tease out his statements that his model relies on a multitude of polling data, and that some of these races had far less substantial polls than desired. What seemed to matter to these people more was that at least the President's race was covered to their liking, and it got to turn Republican nay-sayers sour.
That being said, a substantial portion of people looking at polling data of any sort will often revert to basic cliched responses, because it either validates or contradicts one's hopes, aspirations, and expectations.
Last edited: