This question of trying to figure out whether a book is good or bad by looking at it carefully or by taking the reports of a lot of people who looked at it carelessly is like this famous old problem: Nobody was permitted to see the Emperor of China, and the question was, What is the length of the Emperor of China's nose? To find out, you go all over the country asking people what they think the length of the Emperor of China's nose is, and you average it. And that would be very "accurate" because you averaged so many people. But it's no way to find anything out; when you have a very wide range of people who contribute without looking carefully at it, you don't improve your knowledge of the situation by averaging.
I think tarot cards are a lot simpler and more accurate. Mine suggest that Barack Obama should beware of the tall tower with the lightning bolts, and that a hooded figure carrying three sticks will bring him five golden discs. Meanwhile, Hillary Clinton will be menaced by a demon-like creature with the face of Rudy Giuliani, and will fight back by swinging swords at it while blindfolded. Or something.
Your point? If they have no predictive power, you are guaranteed a profit by shorting front-runners and/or buying long-shots. If you believe that -- what are you waiting for?
What does "no predictive power" mean? The Emperor's Nose technique has a certain amount of predictive power: it's very unlikely that it will be off by more than an inch, for example.
Similarly, in the hands of a conscious reader such as myself, my tarot deck has lots and lots of predictive power. Right now it tells me that Mike Gravel does not have a 50% chance to become President this year, that Barack Obama will not live to be 200 years old, and that Bill Clinton, with 99.999% certainty, has not invented a perpetual motion machine.
I'm not saying these markets are wrong. I'm suggesting that, at best, they're as correct as an off-the-cuff guess by the average reader of the New York Times. That's not all that impressive, but it's just good enough that I can't consistently outguess them without doing some research -- and, given that the total size of most of these penny-ante markets is less than a postdoc's annual salary, I'm never going to make any money doing that.
People in this market aren't a random sample of people in the street. They are self-selected based on their willingness to bet money on the event. If they are poor predictors they will lose money on average and if they are good they will gain money. After many events the good predictors will tend to prevail over the poor ones because they have more money to bet and they'll be more eager to bet... Oh screw it I give up... some people just don't get odds or markets.
If there's one thing I know about gamblers, it's that their willingness to bet money on something generally has no correlation whatsoever to their knowledge about it. Their ability to continue betting over their lifetime certainly does, which of course means that any prediction market that doesn't run on real dollars is essentially useless.
But that doesn't mean the for money ones are useful. This particular one is not. The infrequent nature of bets of this type, plus the small number of people able to get money onto sites like that in any reasonable timeframe keep it (at least where politics is concerned) from becoming anywhere near efficient. It's simply too small for much in the way of smart money to even bother with, so what you see is mostly just the result of sports bettors who watched an hour of CNN yesterday and have nothing else interesting to wager on.
I wouldn't be surprised if they were extremely accurate for European sports, but for American politics they've been way off recently. The good news is, they're very easy to exploit in any close election year. Perhaps I'll blog about how after I do it in November.
I admit the volume isn't huge. Maybe 10K per day on the presidential nominations. So you might be able to average placing 1K on per day without destroying your odds. If you had even a 10% edge then that's $100 gain which most people wouldn't be giving up their day job for. There are other markets though. Betfair had turned over $3 million on the Dem Nom.
The funny thing is the more accurate the market is the less 'smart money' would be attracted to profits and vice versa.
Good luck with your trading. By participating you will presumably increase the accuracy of the market. But I think the average trader might be more knowledgeable than you give them credit for (even ones from those non-US places). And you never know... there might actually be something in this 'Wisdom of the Crowds' thing.
My trading in November (assuming a close race) is more predicated on knowledge of what happens to those betting markets throughout the day of the election than any political knowledge.
As you pointed out though, I can't make retirement money on that small of volume. I might be able to make vacation money though.
Also I should check to see if there's a difference in odds between betfair and intrade. Might be able to run some sort of cross-site arbitrage if possible.
I assume the technique involves betting on whoever's losing, and inversely so based on how much they're losing by assuming that the numbers will come to 50/50 in the end?
I have a reasonable understanding of Brownian motion. Your comment indicates that you don't really understand markets. In an efficient market system (one which is perfectly predictive of chance of an event occurring based on observable information) it should not be possible to know if the price is going to go up or down. If could know for certain then obviously the price is wrong (and you could make a risk-free profit). So yes this would obviously look a lot like a random walk. However if you look at the graph of the odds through time (say for Obama winning the democratic nomination) you'll see it is responding to new information (such as poll results etc). It represents the aggregate opinion of the participants. If you were confident that your opinion is more accurate than this group then you could profit by betting.
This question of trying to figure out whether a book is good or bad by looking at it carefully or by taking the reports of a lot of people who looked at it carelessly is like this famous old problem: Nobody was permitted to see the Emperor of China, and the question was, What is the length of the Emperor of China's nose? To find out, you go all over the country asking people what they think the length of the Emperor of China's nose is, and you average it. And that would be very "accurate" because you averaged so many people. But it's no way to find anything out; when you have a very wide range of people who contribute without looking carefully at it, you don't improve your knowledge of the situation by averaging.
I think tarot cards are a lot simpler and more accurate. Mine suggest that Barack Obama should beware of the tall tower with the lightning bolts, and that a hooded figure carrying three sticks will bring him five golden discs. Meanwhile, Hillary Clinton will be menaced by a demon-like creature with the face of Rudy Giuliani, and will fight back by swinging swords at it while blindfolded. Or something.