Wednesday, August 1, 2007

Predictions: A Success and a Failure - How do they compare?

What does Olson's prediction of PCs not making it into the home environment and Gordon Moore's Law have in common? Well, it appears that in both cases a "swag" or a guess was taken. It seems that neither individual used a lot of data and they operated from their instincts. In Moore's case, he was right on the mark. Olson missed on his prediction and later stepped down from leading Digital into the future.

I thought that this was interesting and very different than Dr. Halal's approach that says you use the data from various sources and build models to make more accurate predictions. I tend to agree with Halal, but Moore's Law is a great example of how a brilliant person can have insight into the future without using a lot of data.

These two approaches are very different. As stated in our text book - when dealing with economic data, most of the predictive models tend to fail - regardless of complexity.

I am left with the following question: How do we reconcile the insights of brilliant people and also leverage predictive models? What do you think? Please leave a comment as I would to hear your thought - maybe you can help me mold my own.

No comments: