The Unpredictable Popularity of "Squid Game"

"Squid Game" is the most-watched series in the history of Netflix. In the first 28 days since its release, the show was watched by viewers from more than 142 million households. As a frame of reference, this makes the show more than 7-times more popular than the most-watched episode of "Game of Thrones."

Why did "Squid Game" become so popular? Both television critics and psychologists have written extensively about the appeal of the show’s brutal violence and its timely social commentary. As someone who had to ration out the show’s episodes so that I wouldn’t finish it in the first two days, I can vouch for the show’s quality. But there are many good shows being released every month. Why should this one, in particular, become a global phenomenon? Why not "Teenage Bounty Hunters" (a one-season show from 2020 about a pair of psychically-connected twins who are both teenagers and bounty hunters)?

Predicting whether any one piece of media will become popular is extremely difficult ahead of time. Although we like to think that success is due to the qualities of the program itself (for example, the acting, the story, or the cultural themes), it’s more likely that popularity is the result of largely random social processes. Consumer behavior is shaped by information about the choices and behavior of other people, and these processes can help turn any (decent) piece of media into a global phenomenon.

Social Information

You are traveling in a foreign city and have a choice between two restaurants. Both restaurants are rated well on Yelp (4 stars), but the first restaurant has three reviews and the second has eight hundred. In this situation, most people would choose the restaurant with more reviews. This is because people use social information about what others have chosen in the past to guide their decisions. All else being equal, people choose products that are more widely used.

Research suggests that this type of social process explains why it is so hard to predict (ahead of time) the media that will become extremely popular. Salganik and colleagues (2003) conducted a series of experiments to examine how social information influences the popularity of online media. In their experiments, users visited a website where they could listen to (and download) a catalogue of randomly selected songs. The researchers then tested how adding social information (the number of times each song had been downloaded by other users) would influence which songs would become popular.

Presenting users with social information had two effects:

• First, when participants could see how often songs had been downloaded by other users, the market became more unequal. In other words, social information made it more likely for a small number of songs to receive a large number of views. It became more likely that one piece of media would completely dominate the market. This is because small early advantages in popularity can quickly accumulate and grow exponentially.
• Second, social information also made it harder to predict (ahead of time) which songs would become popular. It became less likely that the objectively “best” songs (based on users’ quality ratings) would also end up being the most popular ones. The correlation between quality and popularity gets weaker when we have social information about other users’ choices.

These results mean that any show with an initial boost in popularity can quickly become extremely successful. At a certain point, people will begin to watch "Squid Game," or any other show, because of its status as a highly popular show (the #1 most-watched show on Netflix in most countries). In real life, the effects of social information are likely to be even stronger than the effects in Salganik’s experiments, where the manipulation of social information was relatively subtle. On Netflix, the promotion of popular shows like "Squid Game" is (obviously) more direct.

There are still probably limits to the effects of social information. "Squid Game" is, admittedly, a very good show (with 93 percent critic approval and 84 percent audience approval on Rotten Tomatoes). But the show is still probably not 142 million households good. Without meeting some baseline level of quality, it’s unlikely that a show would be able to attract initial interest from users. To be the best possible show of all time, a show merely has to be good enough.

Explaining Versus Predicting Popularity

It is easy to explain why something became popular after the fact. This is true for television shows, hit songs, movies, politicians, and businesses. It’s much harder to predict the future. Seeing convincing explanations might lead us to think that it’s relatively easy to make predictions about the future. But this isn’t the case.

Media Essential Reads

Do Celebrities Make Ads More Effective?

The Often Overlooked Positive Side of Social Media

For every show on Netflix (or shows on any of the other major streaming platforms), there have been rooms full of people who have predicted that it will be successful. Making anything as complex as a television show requires the ability to convince dozens of people that it will succeed. But success, in the end, depends on the randomness of social information. And under the right circumstances, any good show could become the next "Squid Game."