Very smart (and funny) post on the HubDub blog about prediction market journalism —which should aim at demonstrating the real social utility of the prediction markets on news events. Some prediction markets are more useful than others, and, sometimes, some portion(s) in the life of one particular prediction market is/are more useful than other portions.
Let’s have this lofty goal in mind when blogging about the prediction markets. Are they really useful to society? Why? How?
When they establish a track record for accuracy, as something like the Iowa Electronic Markets have, then they can provide the service that macroeconomics has tried and failed to provide for decades: forecasting.
Not the crystal-ball kind, not ridiculously precise predictions about every detail of a circumstance. But rather, indicators of general trends; in favor or against the success of a particular candidate, whether or not the Fed will cut its rate within a certain timeframe, etc. Like all prices, this helps people adjust their expectations for the future; not the way they would if they could know exactly what was going to happen of course.
But there if there is value in decreasing uncertainty than effective prediction markets can be a tremendous asset.
OK.
FYI, I was referring to what Robin Hanson said about the value of the PM information. Look the last paragraph of the definition:
http://www.midasoracle.org/predictions/
The idea is to go beyond accuracy. Accuracy is not enough. See what Hanson said.
Ah, I see what you’re saying. It isn’t a question of whether or not they are accurate, but whether or not they are accurate <i>relative to other forecasting tools that already exist</i>, taking into account the cost of using each tool. Right?
To that end, I think they do add value. Take polling–polling requires you have the resources to go out and, well, get a large sample of people to answer your questions–meaning you also have to get through all the people who simply turn you away. Getting quality poll results is a high cost endeavor.
Prediction markets give people an incentive to be accurate because they lose money if they are not. Moreover, the more people putting money into it the more money the people running the markets make, so they have every incentive to try and draw in more traders, which in turn brings in more of the partial information distributed across those traders.
And (correct me if I’m wrong) the big prediction markets tend to have at least as good a track record as the polls–meaning you have a tool that is low-cost to implement with the potential for very high quality output.
Is this the angle you meant to look at it from?
Polls can be very good forecasting tools. Often however the competitor to one of our markets is the most low tech of species, the pundit. The problem is that the pundit provides forecasts with a 100% confidence and is a fixture of the news media scene. The challenge for prediction markets is to sell the benefit of providing more accurate but slightly more complex forecasts. We have already seen this happen in weather forecasting where (in the UK at least) the forecaster will talk about a ’50% chance of rain’. Maybe recent financial events will make people realize that the world is less certain than the pundits predicted.
Adam,
Yes, relative accuracy and the cost/benefit analysis are what I meant. But PMJ should demonstrate social utility with specific instances, which is easier to say than to do.
-
Nigel,
Too much is said about PMs versus polls and not enough of PMs versus experts. I am very interested by the issue that you have raised here.
I agree with Chris that we should focus on PM vs. experts, because that is where most of the added value is.
Â
Experts usually don’t agree, more often they don’t. Will the earth warm up or will we enter an ice age ? No-one can tell me. No-one knows, although a lot of people think they know. We are now at a stage where some of the experts are betting against each other using real money. AÂ prediction market can give me a reliable estimate of what is likely to happen, that is all I need to know. If it turns out to be 50-50 then so be it. I don’t want to support one theory or the other, I don’t have to, it doesn’t make any sense. You’d only be kidding yourself. All I want to know is the consensus.
Â
Same thing with the economics picture. Will stocks go up or down 12-24 months from now ? Is a real disaster still waiting for us ? No-one can tell me, the experts disagree. And as we have learned recently, a lot of them are lying too, sometimes because they don’t have a choice (telling the truth about the troubled Iceland banks would have caused a run on the banks 6 months ago, yes, the central banks already knew about these troubled banks but didn’t tell us anything, because they couldn’t).
I read a lot about this stuff, but I don’t even know if we’re in an inflationary environment or a deflationary environment any more. You tell me. No, actually don’t tell me because your opinion is probably just as worthless as all the others. All I need is a consensus estimate provided to me through a PM. People are firing opinions all over the place but where is the bloody consensus ? I can look at the stock market but that won’t tell me anything. Yeah, yeah, there will be a 60-90 day technical “rally” but where do we go from there ? How far down ? Nobody has a clue, they are all GUESSING. Well, we don’t have to.
Â
I suppose I will have to settle for learning the odds of Holland winning the World cup soccer thanks to betfair offering this market, which is nice to know because the “expert” opinions suck as usual. But surely there has to be more… This isn’t rocket science. I’ve known people who were driven to the exchanges out of despair. Because whether you know it or not, you really are a mug if you put a lot of trust in expert opinions, especially these days as I haven’t even mentioned vested interests yet. Maybe the most important question is, do you really want to know the truth, or do you want to be a believer ? It’s probably the latter for most, but there is a price attached to it.
Â
A whiff of philosophy, as usual. Hey, you can be the expert.
Â
“I agree with Chris that we should focus on PM vs. experts, because that is where most of the added value is.”
Thanks.