Saturday, June 07, 2008

Current Odds: 59.7% Obama vs. 36.3% McCain



At 6/07/2008 9:40 AM, Anonymous Anonymous said...

In 1936 the magazine Literary Digest predicted from their poll that Alf Landon would beat FDR. FDR won with 62% of the vote.

What happened? The Literary Digest poll questionaire was mailed to names taken from the phone book and auto registration lists. So, their survey went to people who, in the middle of the depression, were well off enough to have phones and cars.

Today we have Intrade as polling authority. Intrade polls a self selected group of geeks. How many people in the wider world even know Intrade exists?

At 6/07/2008 12:01 PM, Blogger juandos said...

"Are you saying that marketplaces work better when more stupid people participate?!"...

Well the 'FREE' marketplace works better with more people involved...

'Stupid people' as you call them provides quick liquidity in the marketplace due to their possible inability to make critical thinking part of their buying experience...:-)

Someone is buying all that silly crap that's being sold on the 24/7 infomercial cable stations...:-)

Regarding the pinko Neville Chamberlain vs. the clueless Manchurian Candidate I think some will remember this from a previous Professor Mark link: Now that the two parties have finally selected their presidential candidates, it is time for a sober-- if not grim-- assessment of where we are.

Not since 1972 have we been presented with two such painfully inadequate candidates. When election day came that year, I could not bring myself to vote for either George McGovern or Richard Nixon. I stayed home

At 6/07/2008 12:57 PM, Anonymous Anonymous said...

Dear Kevin,

You have ironically provided support for my thesis.

At 6/09/2008 2:25 PM, Blogger seadogsf said...


Intrade is not a poll, but a futures market. It is not asking whom they would vote for, but what outcome they think is most likely (backed by actual money). This is a huge difference. While the users at intrade may very well be tech-savvy "geeks" (more likely professional handicappers), the question is not whom they will vote for but whom they think will win the general election. One doesn't take the "non-geeks" into account in the former but one does in the latter. Therefore it is much more likely to arrive at the correct answer than the poor sampling method you highlighted with the Literary Digest reference.


Post a Comment

<< Home