Coin Flipping Isn’t Exactly Random
Posted by Karl Withakay on August 3, 2009
If you’re like me, and I know I am (as my high school art teacher used to say), you’ll find the following article about coin flipping very interesting.
I always knew that technically, flipping a coin is not random, and that given enough information on the input conditions, you’d be able to predict the outcome of any given coin flip. The academics in the article showed this by building their robotic coin flipper that flips a coin to heads every time.
What I didn’t know, and learned from this article, is that the initial position of the coin (heads up vs heads down) when filipping inherantly affects the result of the coin toss. Apparently a coin flip is a 51-49 proposition if you know which side is facing up when the coin is flipped.
The article mentions that the coin flip shouldn’t be eliminated from sports such as football, because to get the advantage, you have to know the initial position of the coin. However, I wonder whether people wouldn’t naturally tend to start out with the coin in the heads up position when flipping. We’re not just talking about pulling a quarter out of you pocket and flipping it. In the NFL, the officials make a big deal out of showing the captains from each team both sides of the coin and explaining which is heads and which is tails. (They use special large NFL coins for the toss in the NFL.) I would think that it is natural human tendency to start out with the coin face up (heads) when flipping a coin if you’re paying attention to which side is up to start.
I always scoff when Jim Hanifan insists you should always call heads on a coin toss, but IF officials tend to flip the coin with the heads side facing up (and I admit I only have my intuition to support this hypothesis), than there is a slightly higher chance of winning the coin toss by calling heads, 51%.
Maybe old Jim knows what he’s talking about after all.