We’ve blogged fairly regularly about prediction markets, so the next step would seem pretty logical: make our own.
Enter the folks at Predictify (see the latest news here), who have been kind enough to create the official Freakonomics Prediction Center. It can be found in the right-hand column of our home page.
We’ll post questions and you’ll supply the predictions. You need to register to play and the money, sadly, isn’t real. (We tried, but the Times seems to have some crazy prohibition against gambling!)
Since there’s no money to be won, we’ll at least give away a piece of schwag to one person each month who submits a good new prediction question. Send your suggestions for questions to editor@freakonomics.com.
and 
. You can also watch it on 
has anyone heard of the National Scanning Board-it was alive and well in the 1970-80 period- provided
future planning info to State governments. Was excellent source for those of us that consider ourselves futurists.Dynmaite massaging of stats and great algorythms. Is is still around. Thanks
Before we get rid of all the cars, we could easily make them run with natural gas, cheaper and cleaner combustion. Evey gas station will sell natural gas, like in other countries, and since we all have gas at home, we could also refill in the garage. It is already done in Europe. Why are we forced to buy only gasoline? Leave the gas tank, add the cylinder for the compressed gas, and voila, we will choose what to use.
already saw commercially viable, non-fossil fuel cars from many companies, including GM and toyota.
see “who killed the electric car?” for details.
Let’s get serious about this. The word, prediction, to me indicates before words, that is what goes before what we put into words. If we put all of our accumulated information into a time variable sequence and then apply what we think is going to change that sequence, we have a prediction. Some may call that a forecast, as contrasted to a hindcast, where we take all that we have found out about something changing over time and see if we can project that forward to a known time a short time ago.
What we usually find from hindcasts is that either our “known data” were not accurate or the projection method was lacking. Once we can get hindcasts to be within a certain percentage of acceptable verification, say 75%, more or less, i.e. that our projection method and our known data symbiotically produce an acceptable result within 25% more or less, we can then try a forecast with the same acceptable limits.
I expect that the great Isaac Asimov had much more than a clue when he considered the large number of participating predictions, with the large amount of known data emanating from the participating predictors, to make Harry Selden’s Psycho-history conform to the desirable contention that large ensembles of samples behave more predictably than small ensembles.
Isn’t one of the base requirements for a prediction market that real money be used so that participants have a tangible stake in the outcome?
You can’t just have something to win – you need to have something to lose.
You’ve got a prediction system here, but it’s no market. Is that why you’ve dubbed it a “Center”?
Katie’s right, there’s no risk here. I made four guesses I wouldn’t have bet money on.
As it is, this approaches other unscientific Internet polls.
Maybe you can get around the NYT policy by creating a fictitious currency. Give everybody 100 ‘accuracy’ points to start and let them wager whatever they want on each prediction. Combine that with Predictify’s reputation calculation to score the payout. A user’s accuracy would be a bragging point, and we all know anybody at 100 points of lower is a dope, so there’s disincentive to be wrong.
I do believe you mean “swag.”
“Schwag” is a slang term for low-quality marijuana.
Feudalism will accelerate the replacement of democracy in the USA.