# More Uncorrelated Mistakes lead to more success More uncorrelated mistakes happen because there were more uncorrelated samples of some abstract [[Idea space]]. By analogy, the same way that Monte Carlo sampling routinely beats human intuition for a sampling distribution, more uncorrelated samples lead to a better exploration of the idea space. In this context, [[Uncorrelated samples of an idea space means that some people are working on something that other people would find absurd]]. For mistakes to be truly uncorrelated, the different attempts should have literally no idea about each other. This kind of sampling is expensive in time and resources. [[Mistakes can remain reasonably uncorrelated even if the two groups know about each others approaches]]. <!-- #evergreen --> [Web URL for this note](http://notes.benjaminreinhardt.com/More+Uncorrelated+Mistakes+lead+to+more+success) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/More+Uncorrelated+Mistakes+lead+to+more+success)