Prediction markets harvesting the Wisdom of Crowds's
Prediction markets, according to Binyah Kesselly, director of strategic planning and process excellence at Janssen Ortho, trade forecasts to determine the highest probability outcome for any given scenario much like stock markets. The forecasts are based on collective intelligence or the...
By on Aug 14, 2007Prediction markets, according to Binyah Kesselly, director of strategic planning and process excellence at Janssen Ortho, trade forecasts to determine the highest probability outcome for any given scenario much like stock markets. The forecasts are based on collective intelligence or the wisdom of crowds the notion that a group of individuals acting together make better assessments than the most expert individual acting alone.
This wisdom of crowds concept drives Wikipedia, Google and other Web 2.0 sites's success in sorting information. It is why friends, acting as lifelines on the popular television gameshow Who Wants to be a Millionaire?, are right 65% of the time, while the audience working as a whole is correct a staggering 91% of the time.
And it works.
Data related to prediction markets forecasting 49 elections in 13 different countries collected one week before the actual vote took place, predicted outcomes with a margin of error of 1.5%, compared with a Gallup opinion poll average of 2.1%.
The bottom line: the group is often wiser and more accurate than the best expert within the group.
Putting prediction markets to work
Intrigued?
It's difficult not to be and you'sre not alone. Kesselly is pioneering the use of prediction markets with Janssen Ortho and the rest of Johnson & Johnson, because he believes they improve the forecasting of future events.
Pfizer, General Electric Healthcare, GSK, Novartis and Johnson & Johnson are all using internal prediction markets for forecasting. And Eli Lilly performed a pilot prediction market that beat its official sales forecast 85% of the time.
Kesselly says prediction markets add value by improving forecasts of future events, complementing or displacing conventional decision making processes and imparting learning about the role of information in an organization. Prediction markets explain why the average consumer picks a specific product what motivates them and that's where the true nugget lies, he says.
By depending on the rank and file from the janitor to the C-suite to predict outcomes, the institutional bias on given topics by experts is removed from the equation, Kesselly says.
Robin Hanson, professor of economics at George Mason University and an expert on prediction markets, agrees and says that prediction markets also offer numerically precise estimates and the ability to continuously update estimates. Prediction markets can be used in the pharma industry, for example, to predict drug quantities that would be demanded at different price points, he says.
They also can be used to examine sales as a function of marketing strategies for allocating and setting budgets, he suggests.
Getting it right
Prediction markets work best in situations where data is dispersed among the many, rather than the few, says Joe Miles, a mathematician from Oxford University who has been studying the application of prediction markets to the pharmaceutical industry for eyeforpharma.
Miles says prediction markets not only exploit the wisdom of the crowd, but also provide incentives to share information and reward successful participants with greater influence on the market.
Prediction markets provide a means of gathering and exploiting untapped knowledge so companies can learn more about their own products and processes, as well as those of their competitors, Miles says.
An ideal [R&D] trading group is made up of a diverse group of scientists, working with the management and forecasting departments, he says. The huge variety of perspectives come together to form an extremely insightful picture.
Using prediction markets effectively, however, requires a department wholly dedicated to forecasting, Kesselly says, that exists in an environment with a desire or need for change and complete support from senior management. The impact of prediction markets will be greatest in an organization that is bold, aggressive and a true market leader, he stresses.
You can't be overly conservative or risk averse, Kesselly says. But prediction markets will take pharma to new places to move us into areas we haven't been before.
To learn more about the application of prediction markets to pharma, listen to our free 15-minute video podcast featuring Hanson, Miles and Kesselly at www.eyeforpharma.com/virtual/ForecastingPM/podcast.htm.
But for the best Prediction Market intelligence, be sure to take part in eyeforpharma's Forecasting Excellence and Prediction Markets webinar on 5 September at 15.00 GMT. Kesselly, Hanson, Miles and Dr. Carol Gebert of Thermofisher Scientific, who introduced the first prediction market to the pharma industry at Eli Lilly, will all be participating and offering attendees keen, hands-on information regarding this cutting-edge forecasting technique. To register for a reduced price of only $99, visit www.eyeforpharma.com/virtual/ForecastingPM.