Hi, author here... you can find more information about the context of this project here: <a href="https://medium.com/@thi.ng/evolutionary-failures-part-1-54522c69be37" rel="nofollow">https://medium.com/@thi.ng/evolutionary-failures-part-1-5452...</a>
Interesting approach. From their Github page (<a href="https://github.com/postspectacular/rnd.farm/blob/master/README.md" rel="nofollow">https://github.com/postspectacular/rnd.farm/blob/master/READ...</a>):<p><pre><code> We'd like to do a little experiment and see if a large (or even not so large)
group of people can collectively generate better quality random numbers than
an algorithm. Intuitively one would think and hope so [...]
</code></pre>
It's not clear that individuals can reliably generate true random numbers when asked to do so (see answers at <a href="http://cogsci.stackexchange.com/questions/3591/how-well-can-a-human-generated-random-number-be-predicted" rel="nofollow">http://cogsci.stackexchange.com/questions/3591/how-well-can-...</a>), but this is asking whether a group of people can do better than that, and is capturing mouse and keyboard keyboard input. It'll be interesting to see how using a group compares to using a single individual's input.
Very nicely done!<p>This reminds me of the tiny project I built where you basically click one bit at a time to collectively generate an abstruse binary stream:<p><a href="http://www.zeroone.io" rel="nofollow">http://www.zeroone.io</a>
Interesting. However, when "recording," the `00` column is always significantly larger than all the others. Doesn't look normal... (Chrome 47, Win7 x64)