SkepticblogSkepticblog logo banner

top navigation:

Coincidences and Certainties

by Michael Shermer, Dec 04 2012

On the morning of Friday, November 16, 2012, I wandered out of my hotel in Portland, Oregon—The Crystal Hotel, an exotic boutique hotel with rooms decorated in the theme of a musician, poet, or artist (I stayed in the Allen Ginsberg room staring at a portrait of the beat poet and realized why I write nonfiction). In search of breakfast, I could have turned left or right as I exited the lobby. I turned right. At the first intersection I could have continued straight, gone left, or gone right. I went left. There were breakfast restaurants on both the left and the right side of the street. I chose one on the right. The hostess asked if I wanted to be seated near the window or next to the wall. I chose the window. About half way through my breakfast I happened to look up to see a man walking by who looked familiar. He looked at me with similar familiarity. I waived him into the restaurant. He spoke my name in recognition. I stuttered and stammered and hemmed and hawed and finally admitted, “I’m sorry, but I can’t remember your name.” He said, “Uh, Michael, it’s me, Scott Wolfman, your agent!”

After I recovered from my embarrassment and momentary fear that I’d never get another speaking engagement, we had a laugh about it all, but then got to thinking—what are the odds of something like this happening? I’m from Southern California and Scott is from Connecticut. And we happened to run into each other in Portland, Oregon, a city neither of us normally has any business being in. I was randomly walking about the town, as was Scott. We were stunned. It sure seemed like something more than a coincidence, and we both joked about how there must be some sort of scheduling god who makes these things happen.

But Scott and I are good skeptics. We know how to think about such events. Even though such coincidences as this really stand out as unusual—and they are when I describe it in this manner—most people forget to consider all the other possibilities: the thousands of people I know who didn’t happen by that diner, the delay at the diner talking to Scott when I might have left earlier and had something else unusual happen that now didn’t, all the other cities I’ve traveled to and dined in when I didn’t see anyone I knew, and so on. And the same for Scott: he has hundreds of clients and knows thousands of people in the lecture business, any one of which he would ever happen to bump into in any given city he happened to travel to, would stand out as unusual.

In other words, after the fact we construct all the contingencies that had to come together in just such a way for one particular event to happen, and then we only notice and remember (and later tell stories like the above) about the events that we noticed as extraordinary, and conveniently forget to notice all the other possibilities. Here’s an article opening you’ll never read:

“A remarkable thing happened to me this morning. When I went out for breakfast I didn’t see a single person I know.”

And yet I’ve had thousands of breakfasts just like this one in which I see nothing but strangers. And, of course, I don’t bother to take note of that uninteresting fact, and I do not give it a second thought. The main cognitive bias at work here is the hindsight bias.

The hindsight bias is the tendency to reconstruct the past to fit with present knowledge. Once an event has occurred, we look back and reconstruct how it happened, why it had to happen that way and not some other way, and why we should have seen it coming all along. Such “Monday-morning quarterbacking” is literally evident on the Monday mornings following a weekend filled with football games. We all know what plays should have been called…after the outcome. Ditto the stock market and the endless parade of financial experts whose prognostications are quickly forgotten as they shift to post-diction analysis after the market closes—it’s easy to “buy low, sell high” once you have perfect information, which is only available after the fact when it is too late. In this story, the hindsight bias was my noticing after the fact all the particularities that had to come together in just such a way for Scott and I to run into each other.

What would have been truly and extraordinarily beyond coincidence is if I had computed ahead of time the odds of running into my lecture agent at that very time and place, and then it happened. But that’s not what happened. My account here is a post-diction—an after-the-fact analysis—instead of a prediction. Unfortunately, most people who are not aware of such cognitive biases fail to consider all the other possibilities, and how the sum of all these possibilities is certainty—something must happen, and 99.99% of the things that happen are uninteresting and unimportant and so we don’t notice or recall them later. This cognitive shortcoming is, in part, the basis of a type of superstition and magical thinking that finds deep meaning in coincidence, while ignoring entirely the certainties that must happen according to the laws of nature and contingencies of history.

27 Responses to “Coincidences and Certainties”

  1. James says:

    I think about this all the time. All of the little and insignificant decisions and timing that goes into where you are and what you are doing at any given point of a day. It perplexes me to try and grasp all of the relativity involved in how the world around me takes shape and moves forward each day/week/month/year etc. I wonder what the future will bring while at the same time having a somewhat accurate idea based on passed events and statistical probabilities. A very interesting discussion worth deliberating on. I always find your blogs, articles, books, and lectures very intriguing Mr Shermer. Thank you for making me think in a different way.

  2. Max says:

    One day, I saw a car with a license plate 638KS2. What are the odds of that!

  3. Chew says:

    You should submit that story to The Odds Must Be Crazy, a blog that collects strange coincidences like this.

    • badrescher says:

      Shameless plug, but “Skepticality” is the official podcast of “Skeptic Magazine” and it features stories like this from It’s the most common kind of story we receive, to be honest.

  4. Jeremy says:

    The fact that these encounters do happen all over is a testament to the fact that we can get about a lot; I had 2 of these within a few weeks happen to me:
    1) I’m on a break from a coach tour to USSR (this was 1988) prior to starting a ~3 year stay in UK from my home in New Zealand. I’m at a cafe in Copenhagen meeting with a girl who came to NZ for a holiday job a year or so earlier. A friend of mine from university in NZ happens to walk past while we’re having lunch.
    2) A week or so after the coach trip, I’m walking around a London suburb & bump into an Iranian schoolmate of mine from 1977-1979, when we went to school in Bahrain in the Persian Gulf.

    I’m sure millions of things have happened to me before & since, just that those two are a bit more notable than the rest…

  5. MadScientist says:

    “… I wondered out of my hotel …”

    If you can prove it, Randi owes you $1M. :) I haven’t heard a good astral travel story in years.

    I think this is a great example of how our biases (expectations) can let us down. The brain says “no, it can’t be that person, what on earth would they be doing here?” and we struggle to identify people, among other things. It’s happened to me a few times, so now I just go “hey, is that you …” just to make sure I’m not imagining things, especially when I hadn’t seen someone in years.

    • tmac57 says:

      Seeing somebody out of context like that can really throw you.

      My wife and I were vacationing in Las Vegas,and while walking through Caesar’s Palace,I saw a guy who looked just like my best friend,leaning against a handrail,but he was wearing glasses,and my friend whom I had known for 25 years never wore them.But I had to ask…”Larry”…I tentatively said as I approached him,and slowly a look of recognition came to his face.It was him,and he had just started wearing the glasses the week before.He and his girlfriend had flown in the day before on a last minute whim (he didn’t know we were also in Vegas),and also coincidentally,they were staying at the same hotel that we were in.

  6. Max says:

    Does hindsight bias make events seem more likely or less likely? On the one hand, once you know that some catastrophe like 9/11 happened, it seems almost inevitable. On the other hand, when you consider all the decisions Shermer and his agent had to make to run into each other, it seems very unlikely.

    • tmac57 says:

      The difference from my perspective,is that in the case of an organised plot,there is intention to make the event happen,whereas in the case of coincidence,the different parties are not intent on meeting (to use Shermer’s example).
      The conspirators leaving a trail of suspicious activity which could logically be interpreted as a red flag,is different than two people going about their daily business,and then just randomly crossing paths.

      • Max says:

        It’s not just organized plots, it’s also a rare disease that kills a patient after a doctor misdiagnosed it. After you learn the cause of death, the doctor seems downright negligent.

      • tmac57 says:

        In the case of rare disease,or even a series of factors coming together for a bad outcome,I would agree that hindsight bias misleads people who do not understand the concept, into wrongly attributing too much fault on the medical community.
        I have a very personal story that involves such an instance,and I struggled to put it into perspective.I came to the conclusion that the factors were too random and complicated to assign any real fault to the medical staff.
        Despite the fact that the loss to me was so personal and devastating,I could not rationally blame it on the doctors involved.I accepted that they did their best,and it was just not enough.

  7. Max says:

    Just for the hell of it.
    This is what happens when your car has the same make, model, color, and almost the same license plate number as that used by armed robbers, and shows up at the same intersection where a SWAT team set up an ambush.

  8. David says:

    I read years ago a similar account written by a mathematician who while reading a paper at a train station was reminded of a friend he hadnt seen for 20 years, then looked up as his train arrived and that very person hopped off the train. Later the mathematician worked out such a thing would only be expected to happen to him by chance alone once in 50 million years ( or something – I forget exactly) But this meant that in the USA, population 250 million, every year such a rare thing would happen to 5 people! Saying a thing is extremely rare or unlikley or statistically improbable or assigning a probablility like 1 in 50 million is saying something about how often it does actually happen, not, as people tend to think that it could never happen unless something other than chance is at work. The problem of course is that a rare and unlikely experience has a powerful subjective effect on an individual and if youre already inrare event than it is to accept it as something thats purely random.

    • David says:

      That last sentence was supposed to read “…if youre already inclined to magical thinking of one sort or another, then it is much easier and it makes you feel important or special to discern something “other” in a rare event than it is to accept it as something thats purely random”

    • Max says:

      Extremely rare events that happen to a few people, like winning the lottery, can be explained by the large size of the population. But events like what Shermer described happen to everyone, and are explained by the sheer number of interactions any one of us has during our lives, and the sheer number of things that qualify as coincidences or interesting events.

      • tmac57 says:

        So what’s the difference? Those sound like exactly the same thing to me.Both are explained by the large size of the population…no?

      • Max says:

        You don’t need a large population to explain the kinds of events that happen to everyone, like running into someone you know. You need a large population to explain events that happen to a few people, like winning the lottery or developing Guillain–Barré syndrome from a flu shot.

  9. Roy Niles says:

    You also meet strangers that cause you to think and do things that add more interest and value to your life than having the unlikely meeting with your agent did. Everything that happens in one way or another is astronomically unlikely to have happened in the way it most certainly did.

  10. BKsea says:

    Aaaaaaaah!!!! I mistakenly read that as the Alan Greenspan room. Now I have to get the vision of Alan Greenspan leering over my bed out of my head.

  11. Phea says:

    Great story. I’ve sometimes wondered about odds, anomalies, luck, and coincidence in this plausible scenario. With 7 billion people I believe it’s actually probable that at least one person has led a charmed life and been very lucky, (I mean, freaky lucky, paranormal lucky, scary lucky), virtually their entire life,(so far). Think about what that reality must be like for this poor sucker. He HAS to believe he’s special, (actually, he IS), perhaps in possession of some para-normal, or at least extra-normal ability.

    At a certain point, he most likely would start taking his luck for granted, as his bias has over the years, time and again, been confirmed. Unfortunately,the fact that past events have absolutely no influence on the odds of a future event occurring, (even if you flip a coin heads five hundred times in a row, the odds for the next flip are still exactly 50/50).

    One day, this poor guys luck will probably run out, maybe when he’s counting on it, leaving him confused, lost, and not having a clue how to deal with this strange turn of events. Anyone know what the odds are this guy is out there somewhere?

    • tmac57 says:

      This could explain Donald Trump.

    • Nicholas says:

      The charmed life guy’s name was John F. Kennedy, Jr. He took to many risks and didn’t listen to his mother’s biggest fear (that he would die in a plane crash). His luck ran out in July 1999. But, oh what a great life he had!

  12. Mr Pogle says:

    The correction question to ask here is not “What are the odds of that happening?” but “What are the odds of anything happening that make me ask “What are the odds of that happening?””

    • tmac57 says:

      I was just thinking the exact same thing!!!

    • Max says:

      Notice that people ask, “What are the odds of that?” for coincidences like running into someone you were just thinking about, but not for any seemingly rare thing, like running into someone with a unique name. But then, everyone is unique in some way.
      Similarly, flipping a coin and getting 7 Heads in a row is more interesting than getting HHTTTHT, even though it’s equally improbable with a fair coin. The reason the 7 Heads in a row stands out is that it brings to mind the alternative hypothesis that the coin is two-headed, which would make the outcome 100% probable. Likewise, coincidences stand out because they bring to mind an alternative hypothesis of some paranormal force that makes them more likely.

  13. John says:

    Michael not recognizing his agent was the interesting part to me. Did that happen because he was not expecting to see him in that environment? Is this similar to the Invisible Gorilla phenomenon?