One out of 64. (Although I have read reports that it was 6 flips out of 7 - if so, some reader can calculate those odds for me).
The cartoon suggests that Republicans are upset; Salon's "Coingate" article says Bernie Sanders' supporters are crying foul.
Time to move on.
Cartoon via Jobsanger.
Amateur mistake. The question is not "what are the chances of this (or any other) precise string of coin flips?" (HHHHHH is no less likelier than HTTHTH) but "Given what we know about the situation, is it likelier that the observed result is due to chance (1/64 for 6 flips) or that the game is rigged (unknown but we can make an educated guess) ?
ReplyDeleteHurray for statistically-savvy people
Delete*cough* This seems to be following the Texas sharpshooter fallacy:
ReplyDeletehttp://www.npr.org/2016/02/02/465268206/coin-toss-fact-check-no-coin-flips-did-not-win-iowa-for-hillary-clinton
It would follow a Poisson distribution with 6 successes with an average success of 3.5 times for 7 trials - http://stattrek.com/online-calculator/poisson.aspx , about an 8% chance. So that could happen. I'm surprised no one is talking about the six coin flippers who all died in accidents on the way home that night. Now that's really unlikely.
ReplyDeleteIt has to mean something....
ReplyDeletehttps://youtu.be/NbInZ5oJ0bc