A long time ago I had a bet with someone who claimed that he could influence the outcome of a die roll. So we set up an experiment to see if he could roll a 6 m

A long time ago I had a bet with someone who claimed that he could influence the outcome of a die roll. So we set up an experiment to see if he could roll a 6 more often than expected. He would do the rolls and I did the scoring. Statistics predict that with a fair die one in six rolls would give a positive result. We did 300 rolls and it turned out that exactly 50 of them were sixes. After some thinking I found this extremely suspicious, because it is far too close to the expected result. I am certain that I had not been consciously manipulating the results, because I was pretty certain about the outcome. But during the experiment we did intermediate counts so I knew if there were too many or too few positive results. So I guess I am not a very reliable observer.