4
Sep

Forecasting

Jon-Kulok3By Jon Kulok on September 4, 2014

Forecasting

I’ve always loved to predict and often rely on a bevy of different forecasting systems to assist me.  This dates back to betting on anything and everything with childhood friends.  Having spent most of my lifetime in this arena, I now consider myself a strong prognosticator.   This serves me well more often than not, but I’ve had some spectacular misses that may be the subject of future posts.

 

As a result, the Good Judgment Project which recruits pools of forecasters intrigued me from the moment I learned about it.  I thought about signing up last year but, like many good intentions, I never followed through due to ongoing commitments.  http://www.goodjudgmentproject.com/.   However, a recent segment on NPR rekindled my interest in the project and I signed up for the next season.  If selected (and I may not be due to the stampede to sign up after the story was produced), I will use this blog to keep a tally of my performance.   NPR on Good Judgment

One of the major objectives of the study is to see if the collective wisdom of the crowd can predict future events.  I’ve spent the last decade thinking about this topic and there are many examples of its efficacy.  The Iowa political markets, in particular, have often proved prescient. http://tippie.uiowa.edu/iem/ (the program has sadly been shut down though).  More recently a group of research scientists were able to predict NFL over/under bets 65% of the time by using twitter data.  NFL bets using Twitter .  Generally, you need to win 52% to break even (or cover vig money), and the very best have a hard time regularly winning over 60% of the time.

However, I’m not 100% sold that tapping into the masses works in every instance.  Why? During the dot com bubble I became obsessed with trying to understand stocks and discovered Charles Mackey’s classic “Extraordinary Popular Delusions and the Madness of Crowds” Wiki on Madness of Crowds

My favorite chapter was on the Dutch Tulip bubble of 1637 – due to a surge in demand the price of tulips increased 100-fold.  Throughout the book, Mackey unquestionably demonstrates that the crowd is not always right.  While this book was written in 1841, it carries lessons for the modern age.

What does this have to do with marketing research?  I think our industry is in the midst of a major transformation.  There is so much more data in the public domain.  Over time, this may lead to a drop in stand-alone survey research with a concurrent increase in tapping available data sources (yes, one of my predictions).  As a result, if we plan to continue to offer insights we need to be able to harness additional data streams.  And prediction oriented markets may be a sandbox we play in.  And the ten year old in me can’t wait.

 

 

Share

Testimonials

I have worked with Edge Research for over 5 years, as a client at two very different companies. Throughout that time, they have always given me very personalized service — not only do I work directly with the top people at the firm, but I also get excellent staff support. They have been able to handle both the volume and complexity of work I need. They are experienced in a range of qualitative and quantitative methodologies, so they can meet my organization's diverse needs, and they always bring new ideas and approaches to the table.

Andy Hessabi Director Market Research K¹² Inc.