t*b
发帖数: 850 | 1 If you carefully look at the past 5 presidential elections (two Clinton, two
Bush and one Obama), one clear trend is that likely voters that back
republican candidate are always at least 2% higher than registered voters
that back republican. It is also interesting that likely voters that back
democrats are always the same or slightly lower (1% or less) than registered
voters that back democrats.
Another trend is that the final results are always closer to polls based on
likely voters.
The conclusion: poll results (most polls are using registered voters)
artificially gave Democrats a 2% margin. |
t*b
发帖数: 850 | 2 Some people here cite intrade, which shows Obama is heavily favored 70% vs
30%.
I am not sure who are likely to bet on intrade. I assume there will be more
democrats-leaners. I guess it is always true that high-tech based toys are
more likely used by democrats.
Intrade shows Scott Brown is less likely to be elected than Warren the
Democrat. But that clearly opposes the polls. The majority polls actually
show Scott leads. Very few polls show Warren lead. |
d**p
发帖数: 510 | 3 check the data of Bush vs Kerry in 2004. |
c******g
发帖数: 1217 | 4 Warren leads in recent polls.
more
【在 t*b 的大作中提到】
: Some people here cite intrade, which shows Obama is heavily favored 70% vs
: 30%.
: I am not sure who are likely to bet on intrade. I assume there will be more
: democrats-leaners. I guess it is always true that high-tech based toys are
: more likely used by democrats.
: Intrade shows Scott Brown is less likely to be elected than Warren the
: Democrat. But that clearly opposes the polls. The majority polls actually
: show Scott leads. Very few polls show Warren lead.
|
