The 2008 Presidential Election (official)

[cincibuck]

Originally Posted by Gatorubet

20-12. I'm a yellow dog this election cycle because the Supreme Court Justice replacement issue trumps all others

Amen

that and if we are to have any idea of the depths to which this administration has subverted the constitution, abused veterans, misused our armed forces--especially the Soldiers and Marines-- damaged our foreign policy, visited havoc upon the purpose and honor of our intelligence professionals, damaged the environment and fattened the coffers of Halliburton and big oil at the expense of our children, there will have to be a Democrat in the White House.

[shetuck]

Originally Posted by AKAKBUCK

These aren't exactly polls but more like futures markets.

Sure, but I guess my point is why would a futures trader's position be at significant odds with polling data?

For example, if you were a gold futures trader and you hear news that people's demand for gold is on the uptick, wouldn't you go long?

Why would a futures market think that there's a 60% percent chance of a dem taking over in the Oval office when polls indicate that percentage is actually on the wrong side of the long/short equation?

EDIT: It's interesting that our own poll is giving Obama about the same chance...

['BusNative]

Originally Posted by shetuck

Why would a futures market think that there's a 60% percent chance of a dem taking over in the Oval office when polls indicate that percentage is actually on the wrong side of the long/short equation?

Irrational exuberance?

[shetuck]

Originally Posted by 'BusNative

Irrational exuberance?

Figures that traders would peg the dems for being yield starved...

Searching for Alpha

[shetuck]

Originally Posted by AKAKBUCK

These aren't exactly polls but more like futures markets.

Reuters: Traders bet Obama will win US Democratic nomination

Traders bet Obama will win US Democratic nomination

By David Alexander

Tuesday, April 8 08:28 pm

WASHINGTON (Reuters) - Democratic presidential contender Hillary Clinton will win several state nominating contests in the coming months but has little chance of becoming the party's candidate for the November 2008 election, traders were betting on Tuesday.

Traders in the Dublin-based Intrade prediction market gave Democratic front-runner Barack Obama an 86 percent chance of being the Democratic presidential nominee, versus a 12.8 percent for Clinton, the New York senator and former first lady.

Results were similar on the Iowa Electronic Markets at the University of Iowa, with traders giving Obama an 82.9 percent chance of winning, versus a 12.8 percent chance for Clinton.

Intrade traders were betting the Democratic nominee would ultimately become president. They gave the Democrat a 59.1 percent chance of winning, versus a 48.8 percent chance for the Republican. Iowa traders gave the Democrat a 57.1 percent chance of winning, versus 46.3 percent for the Republican.

Prediction exchanges let traders buy and sell contracts on the likelihood of future events. Contracts are structured so the prices can be read as a percent likelihood of an event occurring. Studies of prediction markets have shown they have an accuracy comparable to that of public opinion polls.

cont'd...

[AKAKBUCK]

Originally Posted by shetuck

Sure, but I guess my point is why would a futures trader's position be at significant odds with polling data?

For example, if you were a gold futures trader and you hear news that people's demand for gold is on the uptick, wouldn't you go long?

Why would a futures market think that there's a 60% percent chance of a dem taking over in the Oval office when polls indicate that percentage is actually on the wrong side of the long/short equation?

EDIT: It's interesting that our own poll is giving Obama about the same chance...

Who said they were at odds? Just a different measure then a usual poll... "Percent Chance" of winning is different than 60% "will vote for" Obama, 40% "will vote for" Clinton.

Its not who you're gonna vote for, its "who do you think is going to win?"

[shetuck]

Originally Posted by AKAKBUCK

Who said they were at odds? Just a different measure then a usual poll... "Percent Chance" of winning is different than 60% "will vote for" Obama, 40% "will vote for" Clinton.

Its not who you're gonna vote for, its "who do you think is going to win?"

I undrestand your point, but, in practice, it's not "who do you think is going to win" vs "who are you going to vote for". If you reduce it down, it's more like "who do you think people are going to vote for" vs. "who do you think is going to win". I know that sounds idiotic, but that's the only way to break it down - candidates win elections because people vote for them. There's no other relevant factor there that I can see (but I'm open to suggestions ).

There are a boatload of factors that go into WHY people vote for one candidate over the other, but poll results ultimately serve as the *best* indication/basis on which to make a bet about whose going to win.

If the "line" on the bet is different than what polls are indicating, then there's some other factor that's being taken into account that I don't understand. That's what I'm trying to get at (what is that *other* factor that I'm not seeing).

Let me explain my thinking with a [gasp] hypothetical scenario:

Let's say there's a huge box in front of you filled with about 10,000 marbles. Some of the marbles are blue, while others are red (there are no other colors).

Just this morning somebody came and dipped a cup into a random spot in the box and pulled out about 100 marbles. When we looked at the breakdown of the colors of the marbles in the cup, it turned out that they were split fairly evenly (52/48) between the two colors.

The box has been sealed since...

You're in the business of buying and selling marbles. You know how much you'd be willing to pay today for each color marble. Let's say the going price is 10 cents for red marbles and 12 cents for blue marbles.

Now... the question is... how much would you pay for a randomly selected set of 1000 marbles from that box?

Do you think that your "bet" on the color breakdown would have more to do with the sampling that was done yesterday than some other factor?

If so, what other factor would influence your bet away from simply predicting that the 52/48 split would hold up for the rest of the marbles?

[MaxBuck]

Originally Posted by shetuck

...Let me explain my thinking with a [gasp] hypothetical scenario:

Let's say there's a huge box in front of you filled with about 10,000 marbles. Some of the marbles are blue, while others are red (there are no other colors).

I'm not sure where this analogy is going, given that both parties no longer have any marbles with which to work.

[shetuck]

Originally Posted by MaxBuck

I'm not sure where this analogy is going, given that both parties no longer have any marbles with which to work.

Great post. I agree...

[AKAKBUCK]

Originally Posted by shetuck

I undrestand your point, but, in practice, it's not "who do you think is going to win" vs "who are you going to vote for". If you reduce it down, it's more like "who do you think people are going to vote for" vs. "who do you think is going to win". I know that sounds idiotic, but that's the only way to break it down - candidates win elections because people vote for them. There's no other relevant factor there that I can see (but I'm open to suggestions ).

There are a boatload of factors that go into WHY people vote for one candidate over the other, but poll results ultimately serve as the *best* indication/basis on which to make a bet about whose going to win.

If the "line" on the bet is different than what polls are indicating, then there's some other factor that being taken into account that I don't understand. That's what I'm trying to get at (what is that *other* factor that I'm not seeing).

Let me explain my thinking with a [gasp] hypothetical scenario:

Let's say there's a huge box in front of you filled with about 10,000 marbles. Some of the marbles are blue, while others are red (there are no other colors).

Just this morning somebody came and dipped a cup into a random spot in the box and pulled out about 100 marbles. When we looked at the breakdown of the colors of the marbles in the cup, it turned out that they were split fairly evenly (52/48) between the two colors.

The box has been sealed since...

You're in the business of buying and selling marbles. You know how much you'd be willing to pay today for each color marble. Let's say the going price is 10 cents for red marbles and 12 cents for blue marbles.

Now... the question is... how much would you pay for a randomly selected set of 1000 marbles from that box?

Do you think that your "bet" on the color breakdown would have more to do with the sampling that was done yesterday than some other factor?

If so, what other factor would influence your bet away from simply predicting that the 52/48 split would hold up for the rest of the marbles?

Yeah, I understand exactly what you're saying... though I don't think that's exaclty the measure.

I wouldn't pay anything for a randomly selected set of 1000 red and blue marbles. I'd pay some amount for some color of marbles I thought would be more valuable in the future. Or If I thought they'd be less valuable later, I'd borrow them from you, sell them and then buy them back cheaper later and give them back to you.

That said.... this Rassmussen Markets game has proved very accurate at predicting winners in close elections...

Somewhere BM80 or someone got off track and compared it to "Polling" in general...

Guys like Zogby ask questions like, "If the election were today, would you vote for A or B"

Whereas this says, "Is A or B going to get more votes"

That's all I'm saying.

Meaning if the Zogby Poll comes back and says 75% for A and 25% for B. The likelihood of A getting more votes than B is much greater. Probably 99% or something. (The outcome for this particular Rasmussen "poll" isn't the accuracy in how close it is... its who the winner is)

[shetuck]

Originally Posted by AKAKBUCK

Yeah, I understand exactly what you're saying... though I don't think that's exaclty the measure.

I wouldn't pay anything for a randomly selected set of 1000 red and blue marbles. I'd pay some amount for some color of marbles I thought would be more valuable in the future. Or If I thought they'd be less valuable later, I'd borrow them from you, sell them and then buy them back cheaper later and give them back to you.

Well, I guess I thought it's implied that a long/short position can only be excercised in the future (hence the name), based on a spot rate. My analogy about the value of the marbles to you now (in order to sell in the future) uses that 52/48 split as the spot rate.

Not sure how that analogy would work if you were short, but I think there's probably some kind of put/call thingy that might be applicable.

My analogy will disintegrate into dust if I try and make any more out of it than that, so I'll just concede to your take.

Originally Posted by AKAKBUCK

That said.... this Rassmussen Markets game has proved very accurate at predicting winners in close elections...

I've also heard that. It would be interesting to know how that accuracy hold up as we get farther away from the election.

Originally Posted by AKAKBUCK

Somewhere BM80 or someone got off track and compared it to "Polling" in general...

In all fairness to BM80, I think the way I titled my post about the Rasmussen Futures Market might have led him off track (he said he didn't open the link). The heading of the article actually is something like "Latest Results from Rasmussen Market..." but I didn't put it in (neither with their disclaimer/explanation) b/c I figured people those would be seen as "nuances".

Originally Posted by AKAKBUCK

Guys like Zogby ask questions like, "If the election were today, would you vote for A or B"

Whereas this says, "Is A or B going to get more votes"

That's all I'm saying.

Meaning if the Zogby Poll comes back and says 75% for A and 25% for B. The likelihood of A getting more votes than B is much greater. Probably 99% or something. (The outcome for this particular Rasmussen "poll" isn't the accuracy in how close it is... its who the winner is)

So, what you're saying is that the futures traders are betting on the likelihood that the poll numbers hold up?

[AKAKBUCK]

So, what you're saying is that the futures traders are betting on the likelihood that the poll numbers hold up?

Yeah... and... as such... in the end they are good at predicting the winner.

But, in reality...all I am saying is "Who are you going to vote for" doesn't equal "who's going to win."

In the end a 60% chance of winning, though, isn't nearly as good as a 20 point lead. (60/40 polling result). I'm not sure that's what we're comparing though, all I'm saying is that its not the same thing.

But on the topic of the rasmussen market, they are really betting on their ability to predict the value of the "contracts" between now and the election... that's why its fun... and of course the polls are a big part of that.... maybe we should go corner the market on Frozen concentrated McCain. (Probably too cryptic a reference to Trading Places there) now I'm thinking of Young Jamie Lee Curtis' cans.

But... this still is a game... and... maybe there's gonna be an early frost for Obama or a drought for McCain.

[shetuck]

Originally Posted by AKAKBUCK

Yeah... and... as such... in the end they are good at predicting the winner.

But, in reality...all I am saying is "Who are you going to vote for" doesn't equal "who's going to win."

Sure... but all I'm asking is, "Why?"

Originally Posted by AKAKBUCK

In the end a 60% chance of winning, though, isn't nearly as good as a 20 point lead. (60/40 polling result). I'm not sure that's what we're comparing though, all I'm saying is that its not the same thing.

But on the topic of the rasmussen market, they are really betting on their ability to predict the value of the "contracts" between now and the election... that's why its fun... and of course the polls are a big part of that.... maybe we should go corner the market on Frozen concentrated McCain. (Probably too cryptic a reference to Trading Places there) now I'm thinking of Young Jamie Lee Curtis' cans.

But... this still is a game... and... maybe there's gonna be an early frost for Obama or a drought for McCain.

Good stuff Mortimer!

Now on to pork bellies...

[methomps]

Originally Posted by shetuck

Sure, but I guess my point is why would a futures trader's position be at significant odds with polling data?

For example, if you were a gold futures trader and you hear news that people's demand for gold is on the uptick, wouldn't you go long?

Why would a futures market think that there's a 60% percent chance of a dem taking over in the Oval office when polls indicate that percentage is actually on the wrong side of the long/short equation?

EDIT: It's interesting that our own poll is giving Obama about the same chance...

Maybe the traders believe the polls will change. The Dem position might strengthen once one of them captures the nomination. Maybe McCain will suffer from being in the shadow of the Obama/Clinton race. Maybe fundraising will be an issue.

[cincibuck]

Has no one commented on the fact that Hillary is shown wearing an apron bearing the words "the other white meat?"

I for one find that laden with joke potential and Freudian implications.