Mark Frederick · @markYfrederick
2nd Apr 2012 from Twitlonger
What Mega Millions Can Teach Us About Trayvon Martin Backlash
I don’t have terribly many Facebook friends (120), but a good dozen of them mentioned purchasing lottery tickets this week (10% of them). They purchased lottery tickets for what news sources cited over and over as a 1 out of 176 million chance. So at least 10% of my friends decided it would be a good investment to risk a dollar on a 0.000000568% chance to win hundreds of million dollars. And maybe they were right, and it wasn’t such a bad investment.
But if there’s one thing I’ve come to understand in studying and researching pathological gambling in America, it’s that people are really terrible at understanding probability, odds, and percentages. On top of that, they are emotionally biased in assuming they have a better chance of winning then they actually do, even if they intellectually understand the likelihood. Admit it: when you bought your Mega Millions ticket this week, you allowed your mind to consider the possibility of winning. How would you react? What would it feel like? Who would you call first? How would you make sure the ticket wasn’t lost, stolen, or dropped down a gutter on your way to collect the winnings? And of course, what would you spend the money on?
Furthermore, when people hear that there is a 1 out of 176 million chance of winning the lottery, they are often focusing far more on the ‘1’ than the ‘176 million.’ They know that SOMEONE has to win, so it might just be them. But if we really took a hard, mathematic look at it, the truth is you have NO CHANCE of winning. 0.000000568% rounds quite easily to 0%. You will NOT win the lottery. Someone WILL win the lottery, but it won’t be YOU.
Now, what does this have to do with the Trayvon Martin case, you ask? Take a look at Pat Buchanan’s column. Now this column is stupid for a number of reasons that many others can make much stronger arguments about than I. But I’d like to focus on the stupidity of his statistical logic. Essentially, his thesis is that, because African-American and Latino men are responsible for more crimes than other Americans, it is only logical that they be profiled. This is not a new argument, and while it is certainly offensive, it is actually quite compelling to people’s poor understanding of probability. We see Pat Buchanan cite that black males between 16 and 36 are only 3 percent of the population but commit 33 percent of crime and think: Oh my god! That’s 10 times more! No wonder we profile! If I see a black person on the street there’s a good chance they might try to murder me!
But we’re not thinking about what these numbers actually mean. If you bought 10 Mega Millions tickets instead of just 1, you were 10 times more likely to win! And your chance of winning was still ZERO.
So lets look at some crime numbers closer and think about what they ACTUALLY mean. First of all, according to 2010’s crime statistics, 22.2% of homicide victims were killed by people they’d never met. 22.2! That’s a lot higher than I would have thought, seeing as how I’m always hearing how much more likely you are to be killed by someone you know. But then I thought about what that means. There were 12,996 murder victims reported in America in 2010. If 22.2% of these victims were killed by a stranger, that would give us 2,859 random murders The US census reported a 2010 population of 308,745,538. That means that in 2010, as an American citizen you had a 0.0009% chance of being murdered by a stranger. Virtually ZERO.
Now let me ask you this: What sounds scarier? That you are virtually assured of not being murdered by a stranger in America, and you’re even more assured to not be murdered by a black stranger. Or that you are one thousand six hundred and thirty times more likely to be murdered by a stranger than win the lottery! I assume you’ll agree it’s the latter that sounds worse, but it doesn’t matter, because NEITHER ONE is going to happen to you.
Human beings are extraordinarily biased in our thinking, and especially when it comes to the meaning we make from statistics. We FEEL like we have a much better chance of winning the lottery, getting murdered, or becoming a professional athlete or Hollywood superstar than we ever will. The fact of the matter is, Pat Buchanan and others who cite statistics to rationalize profiling are introducing the same bias into what they are claiming is objective facts. There’s quite a big difference in asking the final questions based on the exact same statistical data: Should young African-American men learn to put up with being profiled because we whites are 15 times more likely to be attacked by blacks than vice versa, as Buchanan claims (I have no idea if these statistics are accurate or where they came from)? Or rather, should we stop profiling African-American men out of fear in light of the fact that there is an essentially ZERO PERCENT CHANCE that we will be attacked by them.