How does monte carlo statistical analysis and genetic programming perform using randomness versus not using randomness?
Re-ask that question in a way that shows you understand anything at all about either technique.