Who needs pundits track records when we have prediction markets?

No Gravatar

Emile Servan-Schreiber:

Mr. Kristof, if you want to keep yourself accountable and track the success of your predictions in the long run and in real-time, why not simply participate in a prediction market such as NewsFutures?

You could suggest that particular stocks be listed in relation to particular new stories and their possible outcomes. Then, as you invest in particular outcomes, your prediction portfolio would either grow or shrink, providing us all with an objective measure of your foresight. You could feature on your blog a widget displaying in real-time the &#8220-net worth&#8221- of your various predictions.

Other advantages of this approach would include:

Forcing a detailed specification of possible outcomes-
Having you compete directly (bet against) the general public-
Measuring how much your columns can influence price movements for various predictions-
Leading by example to show other pundits how it&#8217-s done.

There are various types of prediction markets out there, so you can pick the venue where you&#8217-d feel most comfortable:

– Play-money only, like NewsFutures– [or HubDub :-D ]
– Real betting (illegal) like Intrade-
– Charity-driven, like Bet2give.

If the idea intrigues you, please contact me at [email protected] and we can get you started right away!

Emile Servan-Schreiber
CEO, NewsFutures


Readers, do click on the link (which will bring you to the New York Times), and do click on &#8220-Recommended&#8221- under Emile&#8217-s comment &#8212-so that his pitch for the prediction markets will be more visible to all the people reading the comments there. Thanks. Appreciated.

HUBDUB PUNDIT WATCH: TechCrunch is the bottom of the pool, while VentureBeat and Pat Buchanan are stellar.

No Gravatar

I am not surprised at all by the results.

Maybe a non-profit organization should sponsor PunditWatch.

Robin Hanson (mister &#8220-Track Records&#8221-), take notice.

FAMOUS SCIENTIST TO ROBIN HIGH IQ HANSON: Science, which is a very long-term endeavor, does not need your stickin idea about scoreable predictions and track records. Please, go back to minding economic issues in your Ivory Tower, and let us run science our way, on our timing. Thanks. Appreciated.

No Gravatar

Overcoming Whatever:

I don&#8217-t really think the comparison with sports/business/weather forecasters really holds up, for a prosaic reason &#8212- in particle physics, the timescale for experiments is years and decades, not days. There is no way to efficiently grade/reward people on the accuracy of their predictions, and correspondingly no real incentive for anyone to make very quantitative predictions.

On the other hand, it&#8217-s not as if there is no incentive to be right. If you devote your life to working out the ramifications of low-energy supersymmetry and it&#8217-s not there, you won&#8217-t get fired (if you have tenure), but on the other hand your life&#8217-s work will be useless. Which is a pretty big incentive.

Posted by: Sean Carroll [from Cosmic Variance] | August 11, 2008 at 12:25 PM


Sean, I don&#8217-t understand the relevance of the timescale to the efficient grading of predictions. Given enough forecasts we can see a signal of accuracy above the noise of luck in individual forecasts. I agree that the longer the timescale the weaker are incentives from any given reward tied to scoring. But I&#8217-m not really focused on incentives in this post – I&#8217-m focused on whether it is reasonable for folks to crow about being vindicated when they weren&#8217-t willing to make scoreable forecasts.

Posted by: Robin Hanson | August 11, 2008 at 12:35 PM

Scientists don&#8217-t want to make scoreable forecasts.

Hence, it is impossible to collect track records.


Robin Hanson&#8217-s idea has no application &#8212-over than vanity blogging.

Let&#8217-s go back to our prediction markets (where traders work, for free, as info collectors).

Let&#8217-s not waste our precious time on fruitless ideas.

Track Record Collecting vs. Prediction Markets

No Gravatar

Robin Hanson&#8217-s false good idea: collecting track records.

But his post is the living proof that he is wrong:

  • Prediction markets incentivize traders in researching issues (reading the experts&#8217- works), making probability bets, and delivering a collective verdict-
  • Experts don&#8217-t like to state publicly their home-made probabilistic predictions &#8212-as his post shows.

And if experts are not used to express scoreable forecasts, then, by essence, you can&#8217-t collect anything. Hence, the superiority of the prediction market method.

Another false good idea from Robin Hanson.

Predictify is about building track records of human predictors.

No Gravatar

Robert Scoble interviews their CTO.


Predictify is not a prediction exchange. We think prediction markets are superior to polls and surveys, don&#8217-t we? :-D

With Predictify, the mechanism delivering the collective verdict is simplistic: it&#8217-s a poll &#8212-with possibility to get down to each individual answer.

Their conversation is very interesting, nevertheless &#8212-in great part due to Robert Scoble&#8217-s intense curiosity.

Technically, the video is awesome and plays well &#8212-even with my old computer and slow DSL line. Kudos to the Fast Company techies. :-D

UPDATE: I don&#8217-t like that their video starts off automatically, though. With YouTube, we decide to play the video &#8212-it is not imposed on us.

UPDATE: Alas, their embedded video does not go into the blog feed. :(

UPDATE: I e-mailed my remarks to Robert Scoble, and he&#8217-s asked to his techie to look into the issues. :-D

UPDATE: I see that the video does not start on its own, now. They managed to correct that. :-D Rest the fact that their videos don&#8217-t go into feeds.