Monthly Archives: July 2010

Paul The Octopus and Scientific Methods

A conversation about scientific research methods sometimes arises when I meet my dissertation supervisor.  Surely if methods and processes are solid then whatever comes out of them is equally solid? If only life was so easy and straight-forward.

Plenty of in-depth and critical predictive research is carried out on large sporting events. After all, where would the bookies be without them?  The problem is that predicting the future is impossible to do with anything close to total accuracy.  There are simply too many unknown variables that could come into play between the time of the prediction and the final whistle.

All is not lost, however.  The Germans have the solution. They have Paul the Octopus!  The fountain of knowledge, (or Oracle, if you will) has done a seriously good job of predicting the output of the matches played by the German football team.  According to Wikipedia (yes, I know) Paul has correctly predicted the outcome of 10 out of 12 important matches. That is significantly more than could be predicted by chance.

The question then is how.  What techniques and methods is Paul availing of.  A quick reminder that Paul is an octopus is needed.  Does an octopus have the cognitive skills to analyse a range of football performances?  Surely not.  So , what’s happening?  Does Paul have a canny zoo feeder who is the real football analyst and has trained his tentacled friend to accept one box of food over another that has nothing to do with football per se?   Or am I spoiling all the fun by trying to find a reason for this?  Can I now just leave well alone and enjoy watching Paul choosing his lunch?

More seriously, black box processes can produce interesting and plausible output. Yet, they are frowned on because of their lack of transparency.  In formal research, it seems that output is frowned on unless it can be traced back to its roots and each and every step in its production is documented in minute detail. Why is that?  Research has become more about process than product?  When and how did that happen?  Answers on a virtual postcard……