Noel Brewer writes to the SJDM mailing list about a job posting where applicants are requested to
… flip a coin 50 times, record the pattern of Hs and Ts, and send it in with application their materials. Presumably the responses were checked for randomness, allowing the employer to detect people who faked the flipping by making up a faux random series. Of course, really professional cheaters would presumably do it with code and get a truly random sequence. Their method ferrets out the merely lazy cheaters, but it keeps the dutifully honest and the artful dodgers.
Fun stuff. My theory is: it’s just a joke. J-O-K-E. But actually it might be more revealing about applicants’ cognitive apparatus to request that they try to make up a random sequence themselves.
P.S. I just discovered that a coin (well, a US nickel) will land on its edge in around 1 in 6000 tosses (Murray and Teare, 1993)—assuming a frictionless, rigid, level surface, and ignoring internal vibration of the coin.
Murray, D. B. & Teare, S. W. (1993). Probability of a tossed coin landing on edge. Physical Review E, 48, 2547-2552