Gambler’s Fallacy
If you flip a coin several times in a row, you’d expect to get an approximately even proportion of heads and tails. The more flips you do, the closer to 50% heads / 50% tails you’ll expect to get.
Let’s say that you flip a coin fifty times, and get heads every time. What is the probability that you’ll get a heads on a 51st flip?
If you said something other than 50%, then you’ve just fallen into a logical trap known as Gambler Fallacy.
This is an extremely common mistake, so don’t feel bad. Educated and intelligent people fall victim to the gambler’s fallacy all the time, because it defies our natural intuition. After all, what are the chances of getting 51 heads in a row? We all know that it’s extraordinarily unlikely. (Getting 51 heads in a row actually has a probability of 1 / 2^51, or about 1 in a quadrillion). But that’s the probability for the entire sequence. The chances for each individual flip remain unwaveringly at 50/50 regardless of the circumstances surrounding the flip itself.
Here’s a thought experiment which might make this more intuitive. Let’s say you flip a coin 50 times, getting heads each time, and then put the coin back in your pocket. Six months later you pull out the same coin and flip it again. What is the probability of it coming up heads on this flip? I hope it’s obvious that the chances are 50/50. If it is 50/50 in six months, then what if you wait 3 months? One day? One hour? How about if you wait only a few seconds until the next toss and don’t stick it in your pocket in between?
The coin does not contain any information about how many times it has been flipped before. It is a dumb random number generator with no memory capabilities.
It is this same fallacy that causes people to believe in “hot streaks” and “cold streaks” in card games, professional sports, and other places where probabilities are important. The results of one hand of poker have no bearing on the next, so continuing to play because you’re “on a roll” is just plain silly. You may want to continue to play because you’re enjoying yourself - winning is fun, after all. But it certainly is not in your best interest to do so in terms of maximizing your winnings.
In fact, casinos make a lot of money from the gambler’s fallacy. If you believe in hot and cold streaks, then you are likely to want to keep playing when you’re winning, and stop playing when you’re losing. Since the odds are in the favor of the casino, as long as you keep playing you’ll eventually lose some of what you’ve won, perhaps all of it, and perhaps end up in the hole.
December 21st, 2007 1:35 pm
what’s the difference between a mathematician, a statistician, and fool after red has come up 30 straight times on a roulette wheel?
the mathematician doesn’t bet, because he knows that each spin is independent of the last, and that the odds are in the house’s favor.
the statistician bets on red because he makes a case that 30 straight red numbers might constitute a bias towards red.
the fool bets on black, “because it’s due”.