Monday, August 24, 2009

A Flip of the Coin

Well, I have nothing but bad news for you today, Company. I know that you are a big fan of doing "Train, train, Number Nine, rolling down Chicago line, when the train goes off the track, do you get your money back? Yes, no, or maybe so" whenever you have to make a decision. But that only really works when you are choosing between a big group of things, like when you are trying to pick which apple to throw at that clown or which of your gentleman callers to let into your pants. It doesn't work when there are only two apples in the bowl or maybe when it's just a little Navy ship that came into port and there aren't that many sailors floating around. In times like those I know that you always turn to the coin flip. Yeah, bad idea, Company, because recently it has been determined that the odds in a coin flip aren't really fifty-fifty.
"Wait, what? Did I read that right? Did those words really appear on my screen just now? I mean, what the hell is wrong with our coins that I don't have an even chance of getting heads or tails? Is it those state quarters? I bet it's those damn state quarters screwing everything up. Well I am through. No more of that coin flipping for me." I can hear you saying that, you said it out loud. In fact, I could hear it in the next room. And I don't blame you for freaking out...this is bad. As it turns out, nothing is really as random as it seems, including the coin flip, which is exceptionally shocking because the coin flip has always been the metaphor for pure chance, for even odds, for complete and utter randomness in something. That's why it is used for NFL games, and that's why the refs just make up the call on the coin flip anyway.
The problem here is that recent scientific research has come to the conclusion that coin flips are controlled less by the laws of chance as they are by the laws of mechanics and the physics of the actual act. Here's the deal: for natural flips, whatever those are, it turns out that the odds of the side facing up when you begin the flip coming up when the coin lands are actually about 51%. For instance, if you start with heads facing up, slightly more than one half of the time heads is going to win. Now that is startling. As the article I read states, head facing up predicts heads; tails facing up predicts tails. That's about the best way to explain it.
So why...why...why would this be so? I mean, is there something about the way coins are struck so that maybe the heads side carries less weight than the tails side or something? No. The odds don't say anything about whether heads or tails is more likely to come around, all it speaks to is which side is up to begin with. Three academics: Persi Diaconis, Susan Holmes, and Richard Montgomery did the research on this. As to who did the funding, I have no idea. They could have used that money to feed the poor or to buy me new shoes. But I digress. The researchers went ahead and constructed a coin with balsa wood on one side and metal on the other and the same results came out with the coin always landing on the side it started on when flipped by a machine.
Yeah, the three researchers sort of figured out the very complex mathematics governing the flipping of the coin by building a machine to flip a coin with the same amount of force time and time again, filming it with a high speed camera like they use on The Discovery Channel. What this did was confirm that flipping a coin really isn't random. The machine could make the coin come up tails a million billion times in a row if they wanted it to.
As it seems, the relative randomness of the coin flip comes when we humans do the flipping. Way to screw it up, Company. The odds were 100% until you got involved. Anyway, there are several reasons for this. First of all, and more intuitively, is that from one flip to another the force is never exactly the same when a human is doing the flipping. It can be close - damn close - but it can never be the same. When was the last time you flipped a coin the exact same height on multiple occasions? Yeah, I knew that shit doesn't happen. The other thing that makes the coin flip in human hands much more random is that coins flipped by a human tend to rotate on more than one axis, sort of tumbling through the air like a rag doll as opposed to just flipping one way heads over tails.
In the end, it all really doesn't matter. After all, as Homes pointed out, the odds of the coin flip are really not that far off of being 50-50 when done by a person. I mean, it's 51-49 and that's close. Plus, in most coin flip situations the person calling the flip is not the same person flipping the coin are not the same person, if that makes any sense, so that helps to keep anyone from taking advantage of that one flip out of one hundred advantage. Also, the more the coin flips on multiple axis the more random the results become. So it's not all so bad. I, and all the three researchers who were part of this study, think that it's okay to go ahead and keep using the coin flip as a basically random way of making decisions between two choices. So don't sweat it, Company. Flip away, but only if there's two. If you have three or more you are stuck with "Eenie, meanie, miney, moe." I'm just saying.

No comments: