Behavioral New World
May 1, 2023
Is there a success jinx?
You may have heard of the Sports Illustrated (SI) jinx: After an athlete (or team) appears on the cover of SI, their performance subsequently declines. For example, Serena Williams appeared on the cover on August 31, 2015, when she looked (almost) destined to win the Grand Slam of tennis, a feat not accomplished by a woman since 1988. What happened? She lost in the semi-finals to an unranked player.
Was she jinxed by appearing on the SI cover? (There’s a whole Wikipedia page devoted to the SI cover jinx.)
Cathie Wood is a well-known investment manager. She appeared on the June/July 2021 cover of Forbes magazine. Why? Her best-known investment fund over the previous 12 months earned about 89% (!) for her investors. But over the next 12 months, it lost approximately two-thirds of its value (!).
Was she jinxed by appearing on the Forbes cover?
No. These examples are part of a broader phenomenon called “reversion to the mean,” also known as “regression to the mean.” The underlying logic is straightforward: When something extraordinary is happening, it is unlikely to continue. But magazine editors want to attract readers, so it makes sense for them to highlight people whose recent performance has been extraordinary (called “peak performances” for a reason).
Here's your next cocktail party vocabulary: Post hoc ergo propter hoc. This phrase refers to the belief that because one event preceded another event that the first event caused the second event. The fallacy here is assuming that the sequence of events necessarily proves causality. This is one example of “Correlation does not mean causation.”
I’m pretty sure my dog doesn’t quite grasp this concept. When someone is walking down the street, he barks at them, and they move on. I can’t be sure, but based on his satisfied look, I think he believes he has caused the person to keep walking.
Reversion to the mean is not limited to appearing on a magazine cover—it occurs out of the spotlight as well as in the spotlight. Let me illustrate with a few more examples.
Imagine a city concerned about increases in crime.[1] Reviewing the data, the city’s decision-makers see that there has been a sharp increase in crime in one neighborhood. A natural reaction: Devote more police resources to that neighborhood. And guess what? The crime rate comes down.[2]
But was the heightened police presence the cause? We must be careful here. Although absolute proof is not possible, we are likely to be observing reversion to the mean. We can’t know for sure, because we don’t know what would have happened if there hadn’t been an increase in police presence. But reversion to the mean is a characteristic of crime rates, so crime might well have gone down without the policy change.
I think you get the idea. Here are a few more examples without detailed explanations:
A movie producer has a huge success, and their next movie is not as successful.[3]
A corporate manager is fired, and the company’s performance subsequently rebounds.
A medication is administered, and the patient gets better.
The last nine teams to have the NHL’s winningest regular-season record have all failed to win the Stanley Cup at the end of the season (true as of the date of this newsletter).
Wow, the world is sometimes not as simple as it looks. What can be done? In some cases, we can design experiments to assess causality versus reversion to the mean. For example, we can give some sick patients the medication and others a placebo—this will help us assess whether the medicine is working. In other cases, such as the SI jinx, we just need to decide which explanation is most likely.
So by all means, be successful! Enjoy your peak performances! It beats the alternative. But don’t be surprised if there are ups and downs in your success (however measured) that don’t seem related to your genius and hard work. The good news is that you are not jinxed.
Behavioral New World is a reader-supported publication. To receive future posts and support my work, consider becoming a free or paid subscriber. You can also read the Archives, three years of monthly newsletters.
[1] Over the long term, crime has diminished in the United States and around the world. See Steven Pinker’s book The Better Angels of Our Nature, 2011.
[2] https://link.springer.com/article/10.1057/palgrave.cpcs.8150017?. Ideally, we would compare similar neighborhoods, some without increased police presence and some with. The difference in the changes of crime rates would give us a good idea about the actual influence of the increased police presence.
[3] As discussed in Leonard Mlodinow’s book The Drunkard’s Walk: How Randomness Rules Our Lives, 2008. Recommended.
I think the "Cathie Wood is a well-known investment manager" is an interesting one. In something as unpredictable as speculative investments, random bets will also win for a short period of time. Perhaps I am just an oldie, but when I listened to her talks, I could not believe she was getting as much attention as she did. Same hype as the cryto-bros but with better phrasing I guess...