YARPT

[TomCollins]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Paul is in third place behind Giuliani and Romney in prediction markets, having surpassed both John McCain and Fred Thompson.

[/ QUOTE ]

I am totally obsessed with this. I stopped posting intrade updates because no one seems to really think much of it for some reason, but I think it's clearly awesome stuff. He's the THIRD most likely to win according to the betting market, and most msm outlets still treat him like he's a drunk Tancredo.

I love it. Everything is so encouraging of late.

[/ QUOTE ]

Can you explain something to me about how this works? Does Ron Paul being at 9 (or whatever he is at) mean that I can bet against Paul being nominated at this number (and not have all my potential winnings swallowed up by commission). In other words, can I place a wager that will cost me $100, which I will lose if Paul IS nominated, but where I will WIN $9 if he is not?

If you can't do this, then I don't think this is at all an accurate measure of his chances. Betting markets are generally good predictors when they attract an equal amount of money on both sides.

[/ QUOTE ]

You put up $91 if someone else has an offer for $9 and you want to bet against RP. I used to trade at this site when it was part of tradesports. I bought shares of Romney when he was at 9, but when the UIGEA went down, I sold. When it was part of Tradesports, it was $10 per share, so 9 meant .90, and you'd have to put up $9.10 to "short" RP. Commissions were like 2 cents per share.

[Borodog]

Here's what I don't understand, maybe you can help. Someone mentioned that the price listed was basically the percentage chance of the event happening, at least in the opinion of the market participants.

But it seems like the price is just determined by the marginal pairs, right? Whatever you are willing to bid at and buy at is fair game, right? So the price on each individual candidate should not depend on the other candidates.

But then you add up the prices of the candidates, and damned if they don't total exactly 100 (well, 99.8, but to within roundoff in the individual numbers).

Is that a coincidence, or is this the market working better than even I would have believed?

[img]/images/graemlins/confused.gif[/img]

Edit to add that the same is true of the Democratic nomination, although the total is a bit more off, around 97.

But the total for the election itself is again almost dead on 100 (~99).

Not sure I understand how the prices are all set independently yet they can all so nearly add up to 100, unless the market is FREAKING AWESOME, which I know it is, but come on. WTF?

[iron81]

[ QUOTE ]
Not sure I understand how the prices are all set independently yet they can all so nearly add up to 100, unless the market is FREAKING AWESOME, which I know it is, but come on. WTF?



[/ QUOTE ]
I think the market is freaking awesome. If the probabilities didn't add up to 100%, one possible explanation is that the remainder represents "the field" that isn't listed. However, since we're only two months from Iowa, just about every conceivable candidate is listed. If we assume that the nominees will be one of the people listed and the market were off by more than a couple percent, you could book a no risk win by just betting everyone to lose.

[Money2Burn]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
What's with the small number of voters in the straw poll? Are they always like that?

[/ QUOTE ]

It is a private club.

[/ QUOTE ]

Woah, I totally missed that, it's the same club that denied him access for a fundraiser?

[/ QUOTE ]

Yes.

[/ QUOTE ]

Wow, I feel like I have teh retard. This is so much more interesting and hillarious to me now.

[Borodog]

[ QUOTE ]
[ QUOTE ]
Not sure I understand how the prices are all set independently yet they can all so nearly add up to 100, unless the market is FREAKING AWESOME, which I know it is, but come on. WTF?



[/ QUOTE ]
I think the market is freaking awesome. If the probabilities didn't add up to 100%, one possible explanation is that the remainder represents "the field" that isn't listed. However, since we're only two months from Iowa, just about every conceivable candidate is listed. If we assume that the nominees will be one of the people listed and the market were off by more than a couple percent, you could book a no risk win by just betting everyone to lose.

[/ QUOTE ]

Doesn't that logic imply that you should be able to book an automatic win by selling both HOUSE.DEM.2008 and HOUSE.REP.2008, which together total less than 90?

[ALawPoker]

[ QUOTE ]
I think that people that poo-poo the prediction market don't really understand it very well. There was a great article I read about it recently (not actually about the prediction market per se, but used it as an illustration of a broader point):

Evolution and the Wisdom of Crowds

[/ QUOTE ]

Sweet article.

[ALawPoker]

[ QUOTE ]
Here's what I don't understand, maybe you can help. Someone mentioned that the price listed was basically the percentage chance of the event happening, at least in the opinion of the market participants.

But it seems like the price is just determined by the marginal pairs, right? Whatever you are willing to bid at and buy at is fair game, right? So the price on each individual candidate should not depend on the other candidates.

But then you add up the prices of the candidates, and damned if they don't total exactly 100 (well, 99.8, but to within roundoff in the individual numbers).

Is that a coincidence, or is this the market working better than even I would have believed?

[img]/images/graemlins/confused.gif[/img]

Edit to add that the same is true of the Democratic nomination, although the total is a bit more off, around 97.

But the total for the election itself is again almost dead on 100 (~99).

Not sure I understand how the prices are all set independently yet they can all so nearly add up to 100, unless the market is FREAKING AWESOME, which I know it is, but come on. WTF?



[/ QUOTE ]

Wow, good observation. I was arguing with a friend of mine about why this is a reliable indicator a while back, and I never even thought to look at it that way. Good stuff.

[iron81]

[ QUOTE ]
Doesn't that logic imply that you should be able to book an automatic win by selling both HOUSE.DEM.2008 and HOUSE.REP.2008, which together total less than 90?

[/ QUOTE ]
Depending on the vig intrade charges, yes, it does.

[AlexM]

[ QUOTE ]
[ QUOTE ]
I think that people that poo-poo the prediction market don't really understand it very well. There was a great article I read about it recently (not actually about the prediction market per se, but used it as an illustration of a broader point):

Evolution and the Wisdom of Crowds

[/ QUOTE ]

Sweet article.

[/ QUOTE ]

Really? The first paragraph was such complete nonsense that I stopped reading.

[ALawPoker]

I thought it was good. I also happen to hear a lot of people tell me Wikipedia is garbage and betting markets aren't reliable indicators, so I enjoyed seeing a nicely laid out explanation. And I thought the tie to evolution was a cool touch.

I bookmarked it and will have it on reference for when I feel like I'm not getting through to someone.