The Media’s Shameless Bias To Try To Create Competitive Race
[Quote:]
On a perfectly related point, all you have to do is look at the current electoral vote projections from the independent groups — versus those by the media — to see how they shamelessly distort the facts to support their own faulty narrative of the race (or inexcusible bad political analysis):
Obama McCain Net
Electoral-Vote.com 304 221 Obama +83
FiveThirtyEight.com 300 238 Obama +62Real Clear Politics 238 190 Obama +48
Rasmussen Reports 260 240 Obama +20MSNBC 200 200 Tied
CNN 190 194 McCain +4While the four independent groups see Obama leading in the electoral race by an average of more than 50, the two media organizations that have put out their own analysis show the race tied, or McCain even leading narrowly.
June 17th, 2008 at 19:13
Electoral Vote marked Virginia as being won by Obama, despite the fact that farther down the page the only recent VA poll is reported as 45%-44%, doubtless within the margin of error and thus should be a toss-up. Given that AST had VA as a toss-up until this poll, I’m a little confused about his criteria.
June 17th, 2008 at 19:16
PS: a lot of the difference between E-V and CNN is that CNN reports a lot of states as toss-ups, and for a lot of those E-V’s latest reported poll is 3-4 weeks old, i.e. from before the Dem primaries finished, i.e. pretty much junk.
June 17th, 2008 at 19:23
“the only recent VA poll is reported” - Andy never uses just one poll. That in itself may color his view, of course.
June 17th, 2008 at 19:43
Well, I’m not going to dig into what’s written on E-V about Andy’s methodology right now, but take a look at the VA chart and tell me why he called VA for Obama when Obama declined from the last known poll, and the latest poll has them statistically even? Makes no sense to me.
June 17th, 2008 at 19:49
Same reason why NC is called for McCain, even though McCain declined from the last known poll and the latest poll has them statistically even as well. If his method doesn’t make sense, at least it appears to do so consistently
June 17th, 2008 at 19:50
“never uses just one poll”
Bzzzt. Site says: “# If no other polls were taken within a week of the most recent one, only the most recent poll counts.”
Furthermore: “On the map, the states with a white center are those where the candidates differ by less than 5% and are certainly statistical ties.”
Oddly, the converse does not seem to be true, that states than are certainly statistical ties are shown white on the map.
I get the feeling that either the description of the methodology is out of date, or statistical ties are not being reported correctly in the current map.
June 17th, 2008 at 19:51
I’ll send Andy a summary of your remarks…