How Many Murderers Will You Walk Past?
Note: the below is an exercise in Fermi estimation, which is a great way of making accurate, if not precise, estimates based on very little data.
Yesterday, as I was browsing Facebook, I saw a picture stating that “The average person walks past a murder 36 times in their life.” Surely, this is an extraordinary claim, and extraordinary claims require extraordinary evidence. If there’s one thing that clickbait pictures from 9gag do not contain, it’s extraordinary evidence. So, let’s test their claim. How many murderers will an average person walk past in their life?
First, let’s unpack that question:
- The scope of the question is murderers (let’s assume unique murderers, so if one of your coworkers were a murderer you would only count them once no matter how times you interacted)
- The scale of the question is average people, which means that we can use aggregated data and only have to worry about the population described in aggregate
- Walk past implies a very low level of interaction, but is inclusive of higher levels. That lady at the grocery store with whom you were competing for the last jar of your favorite peanut butter counts, but so does your mom.
- In their life indicates that the question is not iterative, and is limited only to the time frame in which the observer is alive.
Great, now let’s set some assumptions and examine if they are reasonable:
- You live in the United States – our crime rate is rather high, which may be a function of reporting, but the rate for murders is actually below the average for the world (but still higher than Europe, Asia, and Oceania). North and South America actually have the highest intentional homicide rate in the world. Let’s keep it simple.
- You can only walk past a murderer on the street if they got away with murder – Popular TV shows say that once two days has elapsed since the murder, it is unlikely to the solved. In reality, the truth is unlikely to be that easy, but there’s also very little data on the topic otherwise. We may have to give this one a pass.
- The average murderer only kills one person – Dr. Mike Aamodt at Radford University has compiled a dataset on the basic statistics of serial killers. It contains data on 3873 serial killers and 11,187 victims, or approximately 2.89 victims per serial killer. Such a low number brings into question the definition of serial killer, but it also tells us that this is a safe assumption.
- The average person walks past some number of people every day. This number is vague and contentious, so we’ll deal with it later.
And finally, let’s get our basic facts in order:
- The average lifespan in the US is 78.74 years
- The rate of murders in the US in 2013 was 4.5 murders per 100,000 people, or 14,196 estimated in total
- The population of the US in 2014 was estimated at 322,583,006 people
- One third of murders go unsolved
This is a stock photo, I swear.
Out of 78.74 years, let’s say that the average person is ambulatory and social for 75 of them. That’s 75 years, or 27,375 days (not counting leap years) during which you might walk past a murderer.
One third of murders go unsolved, so 1.5 murders per 100,000 people go unsolved. Extending on the basis of our third assumption, that means that the rate of unapprehended murderer in the US is 1.5 per 100,000 people, or 4838 at any given time. That is 0.0015% of the population.
That just leaves the fourth assumption: how many people does one walk past every day? This depends to some degree on lifestyle. If you’re a trucker that only works night shifts, you probably have many-fold fewer interactions than does someone who works at an airport or a carnival, to give two examples. In any case, this number is contentious, so we’ll test it two different ways.
First, let’s try out a logarithmic scale distribution. Just from the previous example, it’s clear to me that some people have a lot more interactions than others. Let’s test the resultant range when we test a range that varies in a logarithmic fashion. If you meet 1 new person per day, you meet 27,375 new people throughout your entire life, of which 0.41 of them are likely to be unapprehended murderers. Of course, this means that if you meet 10 new people per day, you’ve met 4.1 murderers throughout your life, and if you meet 100 people new people per day, then you’ve met 41 murders throughout your life.
This is a very interesting result because it is actually quite close to what the picture told us. If you live in or around a city and walk somewhat frequently, it’s very conceivable that you could meet 100 new people per day. Of course, if you live in a small town, or in a city but you do not walk around a lot, 100 is less likely to be achievable. The power of the logarithmic scale is that you get a wide and diverse range without a lot of work.
Let’s try another approach to finding this number. I know that most people meet an average number of people per day, but that some meet an unusually large or small amount. Using a normal distribution, we can compute a “typical” number that the aggregated average of people might meet. Let’s pull forward the values from the logarithmic distribution and assign a quintile (or 20%) each for 1 and 100 new people per day, and the remaining three quintiles for 10 new people per day.
By calculating the distributed average, we find that (20% * 1) + (20% * 100) + (60% * 10) = 26.2 new people per day. This seems reasonable, so let’s finish the calculation.
At 26.2 new people per day, the average American meets 717,225 people throughout their lives, of which 10.76 are likely to be unapprehended murderers. Not bad. 9Gag calls their fact a fun fact because they didn’t kill you… but 10.76 murderers over 75 years really just means that they (and you) are so busy meeting people that any specific person is unlikely to be one of their victims.
Well, that was dark.
But people get out of prison after murdering, so the number would be much higher.
Most people that are in prison for murder are in there for their whole life
Not necessarily.. Murderers are let out all of the time. A life sentence is 25 years. I know a few people who have killed someone, not by accident and are now free.
You know people, not one person, who have intentionally killed other human beings and are now free? Do you know how crazy that sounds and how unlikely it is?
I also know two people that have killed someone. Both very young and both spent around 20 years in prison
It’s not crazy. In some places life sentence means life no parole. In others it means life with possibility of parole. In others it means maximum 25 years, but really out in 12.
It’s not unusual for people who commit second-degree murder to be out of prison within 10-15 years.
They only stay in for life if their a minority. Whites get 2-20years.
I think it was my picture on Facebook 😉
just to go in the opposite direction .. would it be
2.89 * 1.33 = 3.837
3.837 * 4.1 = 15.7317
3.837 * 41 = 157.317
people you will pass by who will most likely be murdered?
Through some twist of fate I happen to know 3 murderers for sure, and one of them is a serial. One died in prison. The serial murderer is in prison and the other is free and clear with the Criminal Justice system. That is kind of scary.
Hi!
Wow what a very interesting stat and thing to read! One doesn’t usually know three killers in their lifetime. If at all interested in discussing, I would love to pick your brain for a report I am doing on the potential of one knowing a killer. If interested in speaking on this topic with me, you can reach me at chasercasting@gmail.com
Thanks!
Chase Jeffery
I know/knew 4 personally. 3 of them are in jail all for separate murders and 1 killed himself (murder-suicide). I went to school with the 3 that are still alive the other person was my best friend’s husband.
Not sure if I should count 1 more who was the younger brother of someone I knew in High School who did a school shooting since I didn’t really know him. Just passed him in the hall occasionally.
I agree with Roccito. You assume they had to get away with murder in order for you to have walked past them. But the prison system is always letting prisoners out, and that includes murderers. Plus, it doesn’t say specifically that they had already murdered someone. It just states that you will walk past that many in a life time. Does this not also include people who will eventually commit murder as well as those who already have? Or does that make the number even larger?
NBC did a study and they estimate that 6,000 people a year get away with murder. That’s not 6,000 murders that is not one person getting away with more than one murder. That is 6,000 people getting away with murder and possibly more than one murder a year. 6,000 people a year get away with murder. In their estimation. That is a lot of people. And those are only the ones that get away with it. And are never charged or even looked at.
It seems that Since most murderers only murder one person, we don’t have to worry about the murderers we pass because they’ve already met their quota. It’s everyone else that is a potential murderer that is disconcerting.
Well, these stats are waaayyyy off according to my own experiences. I don’t live in a city. I rarely meet many new people, but i personally know of 17 different people who got away with murder and were never suspected. Now, there was one group of 3 involved in 1 murder, and another group of 4 involved in another. There was 1 couple. But still, it just proves that many unsolved murders are never a statistic in any way. Some were passed off as OD’s, some as suicide, 1 passed off as a drowning, one a diabetic sugar crash, and 1 as “self defense.” In all cases, it just took a little planning, and BOOM, they’re dead, and the murderer(s) got away clean. So yea, no one could possibly know how many murders there really are. But it’s definitely A LOT higher than any of these “sources” could ever suspect.
There’s a flaw in your math. You said that 4.5 people per 100,000 were murdered every year, divided by 3 to get 1.5 people per 100,000 who get away with murder each year. Every year of their lives afterward (that they aren’t in prison for another crime), they’re walking around adding to the total percent of murders in the population. Violent offending peaks in the late teens and early twenties, and like you say, the average US life expectancy is in the mid-seventies, so on average most of the people who got away with murder in the last 50 years are going to be alive. Let’s adjust that downward to 40 years to reflect the fact that people with a history of violent crime (shockingly) have a shorter life expectancy than average. Assuming they each committed exactly one murder, that would give us .0015% * 40 = 0.06%, or about 1 in 1667 people. The estimate should probably be a bit more conservative, because some people do commit multiple murders and others will end up in jail later for unrelated charges, so my best estimate would be that 1 in 2000 of your fellow community members is a successful murderer.
Plugging that 0.05% estimate into your formula with your estimate for the average number of people Americans meet each day, you get P(not a murderer) = .9995^26.2 = a 98.7% chance you didn’t walk past a murderer on any given day. At that rate, it will only take about 53 days before you passed your first murderer.
What if i’m the murderer? :/
A few things not taken into account that would shift the number in different directions:
1. The average unsolved murder is much more likely to be the work of a serial killer than one that is solved. You can see MurderData.org for an interesting project that attempts to identify patterns in unsolved murders pointing to previously-unidentified serial killers.
2. Killers who are caught don’t immediately go away forever. As others pointed out, many are eventually released. But also, they continue to meet people even while they are detained, e.g. guards, police, lawyers, and other prisoners. This lays bare another feature of the distribution, which is that it is likely to be very lopsided.
An alternative approach to your calculation, after estimating how many killers there are, would be to work out how many different people each of them meets after the first murder.
A remaining problem is that the numbers change over the course of a lifetime: population is increasing over time, and the meeting numbers may not be entirely scale-free. Also, the US murder rate is decreasing. (But what about Americans who come from or go to other countries?)
It becomes fun if you count future murderers. People who don’t even know that they will commit murder at a later point in their life. This can statistically also be calculated, I’m sure …
.This article was fantastic. I am sure in your research you included the small amount of people, and this would have to be one in a trillion or a zillion, that a person met 2 serial killers in their life time., and also lived to tell about it.
And I am sure it would be an xtra million, trillion, that the above said person also got away from the two different serial killers. There are some things that statistics and research and numbers cannot answer.
If you are interested in anymore let me know. Oh, and I forgot to mention that I met a killer that was not serial. I was married to him.
I grew up with 3 people that are doing life sentences. I have been in jail in 4 Stats and have been around at least 6 to 10 more. I know these are not normal circumstances. Also growing up was surrounded by pretty bad group of people..