APPENDIX 8: The Falsification of Evolution by Natural Selection

After a great deal of work the penny finally dropped mid June 2019. All my computer simulation work was really doing was trying to reliably predict the mean or average work required (total number of mutations) to evolve a genome. The Second Law of Thermodynamics is a statement about the average behaviour of a system. As genome size increases the model should show both the power and the limit of what natural selection can do under the Second Law. The problem was to predict the mean of a distribution where the distribution has an unknown law and a large standard deviation. But now thanks to Richard Dawkins this problem has been solved! It is the model performed by Richard Dawkins in one of his Christmas lectures in the 1990's at the RI!

In the original demonstration an audience of about 120 children were asked to all stand and privately guess heads or tails. Another boy at the front was asked to toss a coin. All who got it wrong were asked to sit down. Honesty was as in all science basic to the success of the whole thing. This was repeated until there was only one child left standing at which point Dawkins pointed out he had just guessed seven H-T tosses correctly in a row demonstrating that improbable things are not so difficult to observe. It was of course a simulation of selection's power over pure chance but I don't recall Dawkins actually claiming that.

In this simple coin toss model there is NO TARGET! All simulations of evolution with a predetermined target are invalid because evolution has no target. Instead an external random event becomes the next test of survival for the last guess (mutation) which in a very simple way mimics response in a population to a change in an external fitness landscape. The model is purely a test of selection uncomplicated by other constraints. This model finally achieved my objective of having a model of mutation and selection in a population and I found it has an equation for the mean total number of mutations to evolve any code sequence of any given length. I believe this model has an equation because the true entropy cost is paid for up front by starting with a population which all gets culled so the outcome is predictable but only in entropy terms which is the cost of performing the total number of mutations.