« ALEA Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next

Re: existence of bad pollsters

By: Cactus Flower in ALEA | Recommend this post (0)
Tue, 06 Nov 12 8:10 PM | 101 view(s)
Boardmark this board | The Trust Matrix
Msg. 11404 of 54959
(This msg. is a reply to 11400 by DigSpace)

Jump:
Jump to board:
Jump to msg. #

exactly. drew linzer at votamatic has already shown that the pollsters are clustering, which he suggests means some of the wilder pollsters are trying to save their reputations by staying close to the mean.

so really this means only the dumbest bad apples stand out at the end.




» You can also:
- - - - -
The above is a reply to the following message:
Re: existence of bad pollsters
By: DigSpace
in ALEA
Tue, 06 Nov 12 8:01 PM
Msg. 11400 of 54959

alea, I know that Nate uses a complex weighting method in his poll aggregations , which must totally irk the hell out of the polls trying to shift the average of the polls ... not a big deal when he was less followed, but a real thorn in their method when he is followed. They would have to become more random in their behavior as opposed to outright biased in order for him to not sniff them out.


« ALEA Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next