Monthly Archives: October 2011

Computer programmes and modelling

Computer programmes and modelling

Basic flaws

1. anything  that can be modelled will rebel against the model and modeller.

Hence the two places that modelling would be very useful are the stockmarket asport results. Despite  all the high computer power available  outcomes in these two areas are not reliably predictable as all the possible events affecting them cannot be adequately programmed into a computer and when they are and are acted on this then leads to changes  in the behaviour of the observed event away from the expected outcome.
eg races are fixed  and company results are falsified.

2. Modeller bias is always present  eg set up a  company called Joe’s best business and set out to evaluate it.  Amazing “Joes’ is the  best business.

Worse set up a panel on Climate change. Amazing  finding “The climate changes”,  never would have expected that in a million years from that panel.

Early days yet and this is just a kindergarten attempt at Blogging. The sophistication is yet to come.