abstract
| - It has been generally demonstrated that, because human brains are wired toward pattern detection, we are lousy at intuitively interpreting statistics; this is the main reason why casinos are viable businesses. Trying to do anything to curb this problem often results in the worship of the Random Number God, or beliefs like:
* The hit/miss belief: "A hit ratio below 25% is hopeless and a hit ratio above 75% is guaranteed. Everything else is a crapshoot."Not so much. There are four groups of 25% in 100%. Go ahead and count them. We'll wait. There is a 1 in 4 chance of hitting any one of them.
* The Gambler's fallacy: All probabilities should somehow "even out" while you're playing. For example, if the computer has a hit chance of 50%, and hits, that's okay. However, if it then scores another hit right away, The Computer Is a Cheating Bastard. In truth, it just happened to be the way the "dice" fell. As is often stated, "dice have no memory."
* Naive Combination of probabilities: Given the probabilities of two events, people will often simply either add them or multiply them. Generally speaking, calculating the combined probability is much more complicated. For example, if someone accused a group of 100 people of taking drugs, each person would be 1%. Accusing 4% of adults, and 4% of children, if the group is half of each, would be 4 people, not 8.
* The Definition of Probability: There's two ways probability can be defined. The first is what should happen in a random process in the long run. The second is the degree of certainty with which a belief is held. The first definition applies to statements like, "The odds of rolling a six on a fair die are one in six." The second applies to statements like, "My favorite team will win this game." This difference can be very important. These two views are called frequentist and Bayesian, respectively.
* Decision-Making and Probability: Many make a mistake related to all these fallacies. When a decision-maker makes a decision to play the odds in a situation where he can calculate the odds, he's taking a measured risk based on what he knows at the time. This is his best decision based on what he could possibly know at the time. If this fails, calling it a wrong decision is fallacious because it would require knowing the less-likely alternative would happen. For example, if two gamblers agree to roll a fair die, betting 1:1 where Gambler A wins on a 1-5, and Gambler B wins on a 6, Gambler B is making an idiotic decision to bet - and the decision remains idiotic even in hindsight if he wins. Note that while this often makes fun of the developers messing up at statistics or the author having no clue how it works, this can actually be invoked or justified. Also see 20% More Awesome, which involves quantifying the unquantifiable. Examples of how this plays out in storytelling:
|