Nate Silver (Poblano, 538.com) will be on Countdown tonight

[phrigndumass]

about midway through the program. Watch, Tivo, spread the word ...

http://www.fivethirtyeight.com/2008/08/media-alert.html

[phrigndumass]

1. .

:kick:

[phrigndumass]

2. He should be on in about 10 minutes

[Coexist]

3. Thank you so much for the heads up!

I want to see this.

[phrigndumass]

4. on next

:kick:

[FrenchieCat]

5. I liked very much what he had to say.

He's gonna start getting a lot of hits......more than ever. And that site is so much better than Rassmusen and "them".

Twas cool.

[phrigndumass]

6. Nate's got game!

I believe his model to be the most unbiased in existence. He deserves the good things coming to him! Keith Olbermann was the perfect person to interview him, with sports as the common denominator.

:hi:

[Awsi Dooger]

7. Hit and miss

I watched Silver and he was impressive in some areas.

Notably he pointed out that the VP choice can have much greater impact depending on the size of the state. That's a point I've made on DU since joining in 2002. It's generally 3-3.5 points but it can be much higher than that in a smaller state, and somewhat smaller than that in a major state. Depends on other variables like how recently the state has been represented on a national ticket. He made a very good point that someone like Sweitzer would have massive impact in Montana, 7-8 points.

His comment that a VP is like a demotion for a governor in large states was bizarre. What does that have to do with voting impact? The voters will still back a favorite son. Texas is about as big as it gets and there's little question LBJ had a major benefit for JFK in '60.

I had to chuckle when Silver wanted to believe sites like his are driving the betting markets. I swear, these guys must think 10 years ago was the dark ages. I'm in Las Vegas and talk to political bettors virtually every day. These guys wager tons, much higher than me. Many of them have wagered on politics for decades. I've participated since '96. I can tell you exactly what they are emphasizing. They hope the market sites and fixed odds prices are overly influenced by state poll junkies and websites that overreact to current polling at the expense of situational trends. That's where the bargains are. Very similar to the primaries when the odds on California and Ohio were out of whack, compared to the demographic realities.

Overall he was impressive; calm and made all his points without straying into minutiae, stuff the audience wouldn't grasp.

[phrigndumass]

9. Good analysis

I didn't quite agree with the VP/big state/small state thing either. A one-point bump from a California VP would garner more votes in California than would a seven-point bump from a Montana VP garner votes in Montana.

1% of 30-some million is much more than 7% of 600,000 or so.

But I do believe Nate Silver's model to be the most unbiased. I wouldn't be surprised if he nails the total electoral college projection within 5 electoral votes for either candidate.

I also agreed that the margins of error for polling are more like seven or eight percent at this point, rather than the two to five point range advertised when the polls are released. That will change after Labor Day.

:hi:

[Awsi Dooger]

13. There's no doubt margin for error is understated at this point

In 2008 I love it because we have the situational edge and Obama is near his low point. Minus implosion by Obama there's no way McCain can surge. In 2004 it was just the opposite. When everything was going Kerry's way via months of bad news for Bush, yet Kerry couldn't manage more than a 2 point lead, I knew damn well the balance point was actually a Bush lead of 2-3 points, once normalcy restored. Kerry needed that election to land at a point with Bush in freefall. Instead, Bush was approaching 50/50 approval.

Also, from my previous post, obviously LBJ was a sitting senator in '60, not governor. That was Connally. The topic was VP impact depending on the size of the state. Then Silver threw the governor aspect in there.

As phrigndumass pointed out, 1 point in a huge state can be more critical than many points in a smaller state. Particularly if it means the difference in a winner take all scenario of dozens of electoral votes.

I've started to sample 538. Many impressive aspects to that site. No doubt he's on the right track. I am puzzled that there never seems to be any emphasis on how polls err depending on the state. He's got plenty of scrutiny on how polling companies differ from each other, but otherwise he seems to assume that polling in every state is more or less the same caliber. It doesn't work that way. Many of my supreme wagering bargains have resulted from understanding which states are difficult to poll, and in what direction and margin they are likely to be wrong.

[writes3000]

12. So what are the political bettors saying these days?

The situational trends favor Obama somewhat? Strongly? Or hard to tell?

[darius15]

14. Bets are on Obama, around 60-65 %

[Awsi Dooger]

16. That's right

It basically hasn't moved in months. Even before Obama was isolated as the nominee, the Democratic ticket was 60/40 chalk.

That's what I mean about situational influence. It looks and feels and is logical as a Democratic year and the group opinion is that equates to 60/40 advantage. IMO, that's about right.

In 2004 even when Kerry led the polling, Bush was -140 chalk, or bet 7 to win 5. That's slightly below Obama's current theoretical lead.

Keep in mind there have been changes since 2004, offshore wagering under attack by the bill that Frist attached to the port security bill at the last second. Consequently, sites like Intrade have lower volume particularly from American speculators. It's a little bit more of a foreign view of our elections.

Frankly, I kind of like it. There seem to be more soft spots, overall waves impacting odds in areas where the wave doesn't apply.

[Awsi Dooger]

15. Strongly but not to the point of 2006

That was a second term midterm, virtually a guaranteed avalanche particularly when it's a GOP president.

Normally an open race after one party has held office exactly 2 straight terms is a 50/50 armageddon, like '60, '68, '00. You can even include '76 although technically Ford was the incumbent. Those are among the closest outcomes in history. The one recent exception was '88 when voter wanted a 3rd Reagan term.

Let's put it this way, most of the political bettors I know are right leaning. That seems to be true of gamblers in general. Hell, most of them are white men so it makes sense. In 2004 without exception they were sure Bush would prevail and bet accordingly. I was scolded when I tried to make a case for Kerry. They correctly denounced that as my partisanship wanting to paint a landscape that wasn't there.

This time none of them are backing McCain, not overall. I can't think of one of them who has asserted McCain will win and he's backing that opinion with cash. They hope for a turnaround. Otherwise they are looking for isolated spots to bet McCain, an individual state or a proposition bet. They seem to agree with me that the market sites are tilted to Obama. State after state even when Obama is rightfully favored the odds seem to be giving him too much of a benefit of a doubt. Same thing in the primaries when I got bargains on Hillary in California and Ohio.

Gamblers look for value. That can be a problem in political wagering because one of the variables is holding your cash. The longer they hold it the more of an edge you need. It seems like the value on betting Obama was early, but at that point you had to donate dollars for 6 months or more.

[darius15]

8. Does anyone have any video?

I missed it

[phrigndumass]

10. Someone might post it in the Political Videos forum soon

Give that a try after awhile :hi:

[Gabi Hayes]

17. olbermann rerun is on now

[phrigndumass]

18. Here's the link to the video on FiveThirtyEight.com

http://www.fivethirtyeight.com/2008/08/538-on-countdown.html

[Robbins]

11. 538

Assuming there Isn't a huge shift he has Obama keeping all the Kerry states,taking back the Gore
States of Iowa,and New Mexico,and winning In Nevada,Colorado,and Ohio.It also has Landrieu preveling
In Louisana,taking the open seats In New Mexico,Colorado,and Virginia,taking Stevens' seat In
Alaska,and knocking off Snununu In NH.And will bring us to 54 Democrats and Bernie Sanders and
then 44 Republicans and Lieberman.