The 6 little things David Pennock didnt tell you about the Prediction Market Institute

No Gravatar

– #1. – It is not such a great idea to call it a &#8220-prediction market institute&#8221-, for the reasons that it excludes the non-market mechanisms and the other collective intelligence mechanisms. (See Daniel&#8217-s comment on the Pennock blog, here.) That said, it should focus on prediction markets &#8212-do you feel the nuance, doc?

– #2. – It shall not be a pure academic endeavor. It shall be a mix between academics, exchange operators, and other participants in the field of prediction markets. The term &#8220-applied research&#8221- conveys it. It is &#8220-applied&#8221- in the sense that it is not research eggheads bottling up in yet another Ivory Tower. The outputs of this PMI should be useful for the prediction exchanges and the software vendors.

– #3. – One of the purpose of regrouping the prediction market forces into a grand consortium would be to seek external alliances with some foundations, think tanks or educational organizations that might share our ideology &#8212-yes, I said &#8220-ideology&#8221-.

– #4. – It is not such a great idea to set up our own organization from scratch. It is more pragmatic to seek out the creation of a &#8220-unit&#8221- or &#8220-department&#8221- within an already existed organization that has a larger purpose than ours &#8212-for instance, one focused on &#8220-derivatives&#8221-, &#8220-wisdom of crowds&#8221-, &#8220-digital business&#8221-, &#8220-knowledge management&#8221-, &#8220-forecasting&#8221-, or whatever meta keyword you can think of and that encompasses the prediction markets and their cousin mechanisms.

– #5. – It shall have a clear strategy, game plan, and way to assess the results.

– #6. – It shall have David Pennock on top. That guy is our common denominator. He is our most sociable element. He never slammed anyone. (The only time he went on being bombastic is when, being a boy, he told his mother, &#8220-I don&#8217-t like broccolis, they taste like fractals.&#8221-)

Flu prediction markets can correct Google Flu Trends.

No Gravatar

2 practicing physicians laugh at using collective intelligence for nation-wide flu detection:

[…] Flu Trends tracks almost perfectly with data on influenzalike illnesses that the CDC obtains from doctors&#8217- offices. And as an added bonus, Flu Trends detects outbreaks up to two weeks earlier, when people are still sitting at home sneezing into their keyboards. […]

But if officials monitored only Flu Trends, it would be difficult to sort the signal from the noise —in addition to losing critical details on who is sick. Things besides an actual flu outbreak can cause people to search the Internet for flu information. We would imagine that Flu Trends would spike on the release date for a flu-related movie —maybe Outbreak 2: Electric Booga-Flu. And what happens if a pandemic flu scare hits the nightly news? Flu Trends&#8217- ability to detect when the real pandemic hits will be obliterated when people, including those without symptoms, start to search the Internet. Monitoring drugstore sales has the same issue: A jump in cold-medicine sales may mean a flu outbreak, but it could also mean that CVS is running a sale or that flu fear is causing people to stock their medicine cabinets. […]

They end their articles saying that Google can&#8217-t cure the flu, anyway. [???]

The response to the objections they jot down in the 2nd paragraph above is easy:

  • Informed by all other means, the event derivative traders can determine whether the spikes in Google Flu Trends are due to abnormalities (see the 2nd paragraph in the excerpt above) or due to the real spreading of influenza.
  • Hence, the flu prediction markets have a much higher social utility than Google Flu Trends. Chris Masse said so.
  • David Pennock, go writing another research paper about that.
  • History will retain that David Pennock was research scientist under Chris Masse&#8217-s reign in the field of prediction markets.

Google Flu Trends

Iowa Health Prediction Market

The “predict flu using search” study you didn’t hear about – by our good Doctor David Pennock


New York Times

WSJ Health blog

University College Cork (UCC) School of Medicine + Intrade

Dylan Evans&#8217- website

Previously: #1 + #2 + #3

Do give the Nobel prize to Yahoo! research scientist David Pennock. – Santo Subito! [*]

No Gravatar

He and his co-authors were the first to publish about the fact that flu-related searches on the Web are precise predictors of the upcoming influenza outbreaks.


Best wishes for the rest of their scientific career.

[*] &#8220-Make him a Saint, now!&#8221-, in Italian. :-D

New Yahoo! News election dashboard

No Gravatar

Cross-posted on Oddhead Blog.

The Yahoo! News Political Dashboard has re-launched for the general election stretch run of the 2008 US Presidential election.

Yahoo! News political dashboard for the 2008 US general Presidential election

From the main map you can see the status of the election in every state according to either polls or Intrade prediction market odds. Hover your mouse over a state to see current numbers or click on a state to see historical trends. On the side, you can see search trends, blogs, news, and demographic breakdowns at national and state levels.

You can also &#8220-create your own scenario&#8221- by picking who will win in every state. You can save and share your prediction and compare against markets, polls, history, or celebrities. More on ycorpblog.

Readers will be happy to see more thorough and prominent integration of prediction market prices compared to the primary election dashboard. Is that good enough to quiet Chris&#8217-s whining?

In the PM view, states are colored either bright red or bright blue, regardless of how close the race is in that state. To see a visualization that blends colors to reflect the tightness of the race, see

Yahoo! News also offers a candidate badge that you can display on your blog declaring your choice. The badge features national-level polls, prediction markets, search buzz, and money raised.

Prediction Markets for the 2008 Electoral College = US Electoral Map

No Gravatar

Interesting blog post from Lance Fortnow on the VP prediction markets. (I will soon blog about those.)

InTrade – Electoral Markets Map

Their brand-new widget:

US ELECTORAL MAP: Prediction Markets for the 2008 Electoral College

No Gravatar

ELECTORAL COLLEGE MARKETS: Probabilistic predictions for the 2008 US presidential elections based on market data from InTrade Ireland &#8212-(

By Lance Fortnow, David Pennock, and Yiling Chen. :-D

For more on probabilistic predictions, go to our &#8220-Predictions&#8221- page, or visit the prediction exchanges.

Alternatively, if you want an electoral map made of polls, go to


No Gravatar

If US laws were gambling compatible, would a FaceBook betting application solve the chicken-and-egg problem that any brand-new prediction exchange is facing? (Short sellers will come to the exchange only if there are enough backers, who will come only if there is enough liquidity, etc.) Could MySpace, FaceBook and LinkedIn (who have registered people by the millions, already) provide a starting launch for future prediction exchanges?

YooPick for FaceBook

YooPick @ FaceBook

PS: Yet another hit for Robin Hanson&#8217-s MSR. :-D

BetFair is experimenting a phenomenal revenue growth with sports prediction markets, as I am typing this post. Maybe our good friend the pragmatist could look into that, and report his research findings to the CFTC. Im sure hell do. The delicious sound of a red-hot cash register, with the dollars (o

No Gravatar

BetFair’s bet matching process stopped for one full hour on May 7, 2008, and their P.R. people are over-apologizing for it. They are apologizing for the problem, and for their slow response to affected customers &#8212-good move. They are also giving out 10 bundles of ?1,000 each to 10 affected traders chosen at random &#8212-a move I have mixed feelings with.

OK. Now, the beef.

General Formula for Equation of Exponential Growth

During the same period we’ve continually seen record volumes of bets. For example, the value of bets placed on Betfair last week was greater than for any previous week including Cheltenham or Aintree [which are big British horse racing events].

Three different expontial growth functions

Third Workshop on Prediction Markets, July 9 in Chicago

No Gravatar

Please join us at the Third Workshop on Prediction Markets, in Chicago on the afternoon of July 9, 2008. We plan academic talks, an industry panel, and open discussion. Participation is open to anyone, and we hope to be as flexible as possible in allocating speaking time. See the workshop website or the full Call for Contributions and Participation for details.