In this blog post, we explore the secrets of these connections through scientific experiments and data analysis to determine whether the “six degrees of separation” theory—which originated in offline human relationships—still holds true in the digital world and web networks.
In the novel *Chains* by the renowned Hungarian author Frigyes Karinthy, the protagonist says: “I can connect to anyone in the world through a chain of five acquaintances.” He explains that he can even connect to a Ford Motor Company worker he has never met, describing the process as follows: “Last year, I met a friend named Pastore. Pastore happens to know an executive at Hearst Publishing. That executive knows the president of Ford Motor Company, and the president is connected to a company manager. And that manager knows a factory worker very well. So, I can ask that worker to build the car I want.”
Although this novel did not receive much public attention at the time and faded from the spotlight, the claim that everyone in the world can be connected through these “five degrees of separation” was an important insight. This idea later became the starting point for the concept of “six degrees of separation.” About 30 years later, in 1967, Harvard University psychologist Stanley Milgram experimentally revisited this theory.
Milgram conducted a “letter-delivery experiment” to measure how close any two people in the United States actually were. He asked people living in places like Wichita, Kansas, and Omaha, Nebraska, to deliver a letter to a specific woman living in Massachusetts. However, instead of delivering it directly, they were instructed to select one acquaintance who seemed best suited to pass it on. That acquaintance would then pass the letter to another suitable person, and the experiment would be complete once the letter reached the final recipient in this manner.
After excluding letters that failed to reach their destination, an analysis of the number of intermediaries required for a letter to be delivered revealed an average of approximately 5.5 people. This figure closely matched the original hypothesis; since it rounds to 6, it became the basis for the term “six degrees of separation.” This result demonstrates that social distance is much shorter than we might think.
Does this concept hold true in the digital environment? Albert-László Barabási, a scientist at the University of Notre Dame, applied Milgram’s idea to the virtual space of the web. He sought to measure the distance between websites—that is, the “number of clicks” required to navigate from one web page to another. For example, the webpage of one of his graduate students is directly linked to his own webpage, so it can be reached with a single click. However, it took an average of about 12 clicks to reach the webpage of a randomly selected philosopher.
An analysis of the entire University of Notre Dame intranet revealed that the average number of clicks between documents was about 11. Surprisingly, this distance was also much shorter than expected. Of course, the University of Notre Dame’s web is only a tiny fraction of the entire Internet. In 1999, the entire web was at least 3,000 times larger than this. Does this mean that the number of clicks between two randomly selected web pages on the entire web would reach 33,000?
To verify this, Barabási introduced a sampling method. He first analyzed small units consisting of 10 nodes (documents) on the web, then measured the average distance between any two nodes within those units.
He then expanded the sample size to 100, 1,000, and 10,000 nodes, measuring the average distance in the same way, and confirmed that even as the number of nodes increased, the rate of increase in distance remained very low. This trend could also be expressed mathematically using a specific formula.
Based on these analysis results, it became possible to predict the average number of clicks between documents simply by knowing the total number of web documents. According to a survey by the NEC Research Institute, as of the end of 1999, the entire web consisted of approximately 1 billion documents (nodes). Applying this to the model leads to the conclusion that, on average, any two web documents can be connected with only about 19 clicks.
Although we generally think of the Internet as vast and complex, the path from one document to another is actually very short. This implies that the Internet is also a “small world,” and that the structure connecting information is surprisingly tightly woven. Ultimately, this serves as an example demonstrating that the “six degrees of separation” is a valid insight not only in the offline world but also in the online world.
