The Statistics Trap

If I can give you any advice as an instructor, it’s to not get stuck in the statistics trap. As a person who’s just living their life, in this age of disinformation, where we have way too many people telling us way too many things, it’s very easy to fact-check something over and over and over again, even if the source is incorrect. What I mean by the “statistics trap” is very simple. A lot of instructors will read fancy statistics, memorize those statistics, and use those statistics constantly, regurgitating them, trotting them out like a proud parent, but without understanding where the stats come from, what the research is, or any other factor that makes that statistic true. Very simply put, statistics can be used for anything across the board.

The first statistic I used when I started teaching was related to stabbings. We are a very knife-oriented gym; we teach a lot of blade work. The city I come from, Edmonton Alberta, is lovingly referred to as “Stabmonton” Alberta, due to the fact that gun violence is low, but knife and machete violence is very high. We used the general stat that everybody used, which is that 80% of knife attacks come underhand, and that’s how the attack lines work. When we built our first curriculum, we designed it off that statistic, since I don’t go around knifing people. I’ve been stabbed, I have friends that have been stabbed, but the studies showed that most attacks were coming underhand to the lower body, side, kidney region. I ran with that stat for two years. Every day of those two years, people would come to me with anecdotal evidence, saying other things, like, hey when I was stabbed, this happened, or, I’m a paramedic, and this is happening, etc etc. And I held steadfast and true to the statistic that I had read, in a book from a country that I’m not from. When I delved deeper, I realized that the survey that the statistics were based on had included prison stabbings. If you know anything about prisons, the weapons that are used are generally point-oriented weapons. So, of course, the study was skewed towards stabbing at that low angle, because so many prison stabbings come at that angle.

Why bring this up? I had fallen into the statistics trap. It was ridiculous of me to tell people who worked as EMTs, to tell people who had been stabbed, to tell people who were doormen, that stabbings happened a certain way, when all their experience didn’t line up with the statistic. It was ridiculous of me to disregard my own story. My favourite joke is that I’ve been stabbed two and a half times – I’ve been stabbed once in the leg, once in the face, and once by a fork (that’s the half). Every single time I was stabbed, it was an overhand stab, it looked like a monkey dance with a knife, an overhand swing coming at me. This was also the evidence I was getting consistently from EMTs, people I trusted and respected, but because I had this fancy statistic and I was an “expert”, all of them turned their brains off and stopped arguing with me because of the fact that I could quote a statistic.

Understand the information that you’re using, understand where it comes from. You can use it as an example, but nowadays most statistics on the internet are written as clickbait. These sites want to give you a stat like, “1 in every 7 males with blue eyes is attacked by foxes”. That’s a crazy stat! Obviously, you have to understand how surveys work, and how sample sizes work. There was a great article by the Huffington Post on the statistic that 1 in 4 women will be sexually assaulted while in places of higher education. That article then breaks it down very well … if you read the article, you see it’s not 1 in 4; that number is exceptionally high. That number is the 1 in 4 people who took the survey, and it just means that a large number of the people who took the survey were been people that had already been sexually assaulted. Now, I am not saying that sexual assault is not a horrible thing. What I’m saying is that the number that everybody’s throwing around, the 1 in 4, was used by the New York Times as clickbait, without the full study being released.

One of my favourite bits from the great comedian Bill Burr is on stats, and how he hates stats. He does this whole bit on how you can go to iamright.com and use a stat to prove that you are right or wrong, as long as it lines up with your vision of things. That’s the problem with most of this stuff; the statistics you’re using just happen to line up with your view of the world, and so you take that stat at face value without going further into the information. Then you’re disseminating information to people that is not intelligent, it’s not making people more powerful, it’s just making people more crazy about number crunching, just blurting out things to make it sound like they are more intelligent, again just becoming really proud parents.

The stat Bill Burr uses is about shark attacks. His bit is, “did you know, most shark attacks happen in shallow water?” And he pauses, and everybody’s thinking, “yup, yup, yup, that makes sense” and then he states it very clearly, “why do you think that happens?”

Because that’s where the PEOPLE are, people are in SHALLOW water, so of course shark attacks are going to happen in shallow water.

This problem of the statistics trap is becoming more and more prevalent in the marketing of self-defense programs. I’m not saying don’t use statistics, and I’m not saying change your marketing plan. I’m saying that every one of those statistics should have a little star beside it, and the star should say, “as of the study here, where they use a sample size of this”. Because there is no study in the world that takes 100% of the population, who then all send their surveys in, and who then don’t lie, so then that stat is completely true. You have to take all these stats with a grain of salt.

Rory Miller sums this up succinctly as well, just to hammer this point across one more time. His favourite saying is, “Correlation is not causation”. The stat he drops is very simple. “Did you know, that the more churches in a city, the more violent crime that happens in that city?” The wheels start turning in everyone’s heads, “oh yeah, that makes sense, uh, obviously more churches means there’s more religious tensions …” and then Rory breaks it down just as simply as Bill Burr does. He says, “No. The fact that there are more churches in the city means there are more PEOPLE in the city, and more people in that city means there’s more violent crime in that city”. More in this case not being a per capita rate, but simply more total cases. The language used and the statistics used are very important.

So. Do me a favor. Do your due diligence on statistics! Find the stat you like, if it fits how your brain works, if you think that this stat is true, read the actual study. Don’t take the clip, don’t do the thing that happened on IFLS (the science publication), where they put up a headline that said “Cannabis Proven Not From Earth!” People shared and shared and shared and shared, and loved it, and said “yeah, cannabis isn’t from Earth, that makes sense!” If you had clicked the link instead of just reading the clickbait title, instead of just reading the 1 in 4 – the article actually wasn’t about that at all. The article was about people only reading headlines.

As an instructor, it’s your duty not to fall into the statistics trap.

Randy King

2017-11-15T22:19:50+00:00

Leave A Comment