Forgot your password?
typodupeerror
Network Math Stats The Internet Technology

You Can Navigate Between Any Two Websites In 19 Clicks Or Fewer 185

Posted by Soulskill
from the too-close-for-comfort dept.
An anonymous reader writes "A study done by a Hungarian physicist found that of the billions of websites and over a trillion objects on the web, any given two are separated by no more than 19 clicks. 'Distributed across the entire web, though, are a minority of pages—search engines, indexes and aggregators—that are very highly connected and can be used to move from area of the web to another. These nodes serve as the "Kevin Bacons" of the web, allowing users to navigate from most areas to most others in less than 19 clicks. Barabási credits this "small world" of the web to human nature—the fact that we tend to group into communities, whether in real life or the virtual world. The pages of the web aren't linked randomly, he says: They're organized in an interconnected hierarchy of organizational themes, including region, country and subject area. Interestingly, this means that no matter how large the web grows, the same interconnectedness will rule.'"
This discussion has been archived. No new comments can be posted.

You Can Navigate Between Any Two Websites In 19 Clicks Or Fewer

Comments Filter:
  • by mark-t (151149) <markt@lynx.b c . ca> on Wednesday February 20, 2013 @02:54AM (#42953303) Journal

    ... this one is quite obviously false. And illustrates one of the dangers about assuming that extrapolation is equivalent to actual supporting data.

    I mean, there are objects behind paywalls that, all by themselves, can be more than 19 clicks away from a highly unrelated web page elsewhere online There are objects which are online that have no external links to them at all. And those are just the obvious ones.

    It's an interesting notion, but it's incorrect.

    • Have you read the article? The blog entry is entirely plausible, sounds like research I have been seeing for a good decade or so. I also found the article via the author's website: http://www.barabasilab.com/pubs/CCNR-ALB_Publications/201302-18_RoyalSoc-NetworkScience/201302-18_RoyalSoc-NetworkScience.pdf [barabasilab.com] Interesting citations it has. One can also derive things with statistical methods. I would also note that this is talking about websites, not pages.
    • by lxs (131946) on Wednesday February 20, 2013 @04:30AM (#42953699)

      The guy is a physicist so it is only true for spherical websites in a vacuum.

    • by wvmarle (1070040)

      On the other hand, it's said that every person on this planet is connected to anyone else in no more than six links. So any web site linked to any other web site in less than 19 clicks - especially with sites like Google in the mix - sounds rather plausible to me, if not on the high side for number of clicks even.

      • by mark-t (151149)
        There are, believe it or not, websites that are not indexed by *ANY* search engine. Many of these are private, and only a limited number of people would typically even know of any particular one's existence, but they still have web pages.
    • by matunos (1587263)

      Is an object with no external links to it truly online? Is it offline? Is it in a superposition of online and offline?

      The mind boggles.

      • by Sique (173459)
        What ever it is, it surely is not part of a world wide web. It's an island all of its own.
      • by tompaulco (629533)
        If you can get to it by typing it into your address bar from any computer with an internet connection, then it is online.If you save it as a bookmark, is it now considered linked? I guess that depends on whether your computer has been hacked.
    • by Hentes (2461350) on Wednesday February 20, 2013 @05:51AM (#42954013)

      Many times it's not the author being overenthusiastic but the media reports.

    • by pla (258480)
      Like most overgeneralizations... this one is quite obviously false. And illustrates one of the dangers about assuming that extrapolation is equivalent to actual supporting data.

      Strange, I would have called it trivially true: click on the URL bar, click the keys "goo.gl/yc2lK", click enter.. Done in 14. :)

      I mean, there are objects behind paywalls

      I see your point, but actually feel fairly comfortable with the author excluding paywalls and the "dark" web. If I can't get there at all without special a
      • by mark-t (151149)

        I see your point, but actually feel fairly comfortable with the author excluding paywalls and the "dark" web.

        So do I, but the author said, referring to the estimated 14+ billion individual pages on the web, "Like actors in Hollywood connected by Kevin Bacon, from every single one of these pages you can navigate to any other in 19 clicks or less." (emphasis mine). If he had meant most, or even "practically all", he should have said that.

        Yes, I'm being pedantic. But this is slashdot.

    • by Darinbob (1142669)

      And for any number N, it is possible to create a set of pages linked together in sequence that requires N+1 clicks to traverse.

  • by Biff Stu (654099) on Wednesday February 20, 2013 @02:54AM (#42953305)

    It's got to be a reasonably good, well-liked site, but not a mega-site like Google or Facebook.

    How about Salon.com or theonion.com?

    I would say /., but by its nature, /. has too many connections to be used for a Kevin Bacon number equivalent. Conversely, The Onion probably doesn't link to enough stuff.

    I vote for Salon.com

    • I don't think Slashdot is the great hub you think it is. Every time I click a link here to somewhere else, the page never loads and is clearly broken.

  • So how many clicks does it take to get from Tub Girl to Goatse?

    Wait...don't answer that.
  • if everyone links to Google/ Facebook/twitter/Skype and a thousand other social media sites.
  • After all is said and done, it really is a small, small world. [wikipedia.org]
  • by tsa (15680)

    I often go to a random video on YouTube and then try to get to a certain video just by clicking the suggested videos on the right side of the page. You must try it with videos you normally never watch, otherwise it's too easy. It's fun, you never know what you find on your way.

    • The problem with that game is the suggested links are based on your history nowadays, so if it's a video you're likely to want to see, it will be biased to come up early. If it's a video you've never watched and are never likely to want to watch, it may well end up hidden.

      I once switched off the adult filter on Youtube, because a video that used the F-word wouldn't play with it on. It's amazing how my search results changed that day.

      • by tsa (15680)

        It's true, even if you don't have a Google account it does that now. But still it's a fun game to play, and it should still find related videos, even of stuff you never watched before.

  • by Anubis IV (1279820) on Wednesday February 20, 2013 @03:36AM (#42953473)

    Not a chance, unless you're counting the number of clicks it takes to turn on the on-screen keyboard and type enough characters into Google search for a reliable suggestion to show up. Up until two years ago when I left academia I was with an Internet research lab at a major university, and I saw diagrams of some of the graphs collected by decently large web crawls of the time. None of them would have been clustered enough to allow jumping between two arbitrary sites in 19 clicks or less for three primary reasons:
    1) Most links are unidirectional, not bidirectional (e.g. you might link to a news story, but the news site is unlikely to link back to you). As a result, it's rather difficult to reach sites on the fringe of the graph, since many of them have few or no links pointing to them.

    2) Domains (as in domains like medicine, technology, and automobiles, not domain name like google.com) tend to be segregated from one another and oftentimes have long chains before they reach more clustered/common parts of the Internet (e.g. if you start at a particular site for a niche topic, there may be only one other site pointing to it, and then only one pointing to that one, and so on for quite awhile).

    3) Many sites don't have any links to other sites. It's not as uncommon as you might think, and they'd all count as a dead end, which would obviously end your traversal if you were starting from that site.

    When I used to see those graphs, most of them would exhibit chains that would dangle off of the main cluster and would stretch out for dozens or hundreds of sites in length, meaning that if you started from one of those sites in the middle, you'd have to go half that distance in either direction before you'd make it back to the main cluster. Even with as far as we've come in recent years, I seriously doubt that all of those chains have been eliminated.

    • by CAIMLAS (41445)

      A more interesting observation would be how many links it takes to get a link to Wikipedia, or a Google search result link. That proves closer to truth in my experience.

    • I have a sneaking suspicion there was a tpyo, and they actually meant 9! clicks.

    • Sorry, I know this is extremely late to this game, but I'm a little behind. There was a correction posted later stating that the study was from 1999, not recently. In 1999, I'd believe that the entire WWW was connected by 19 clicks. It was in an infantile state compared to today with "portals" like Yahoo! linking to tons of different content. GeoCities pages were all the rage and would often link back to the GeoCities homepage which was one of these magical portals. MySpace was humming along, and eBay
      • That makes a LOT more sense. Thanks for taking the time to point out the update, even if you are "late to this game". :)

  • by dcollins117 (1267462) on Wednesday February 20, 2013 @03:37AM (#42953477)
    If you read the actual article you'll find that his findings "involved a simulated model of the web that he created to better understand its structure. " So this article has nothing to do with the actual internet, but a simulation of it. It's not a noteworthy study, and I'm wondering why I wasted my time reading about it.
    • by SeaFox (739806)

      It's not a noteworthy study, and I'm wondering why I wasted my time reading about it.

      Because you didn't do enough surfing to find the superior time-wasting content that was only a click (or 18!) away?

    • You must be new here. R'ing TFA is frowned upon.

  • by twocows (1216842) on Wednesday February 20, 2013 @03:44AM (#42953503)
    <html></html>

    HAH!
  • "A study done by me has found that of the billions of websites and over a trillion objects on the web, any given two are separated by no more than 2 clicks. Distributed across the entire web, though, are links to search engines such as Google —that are very highly connected and can be used to move from area of the web to another. Google serves as the "Kevin Bacon" of the web, allowing users to navigate from most areas to most others in less than 3 clicks."

    I need my PhD. Now.
  • Why does the summary claim "sites" when the TFA clearly says any two pages? Oh well...

  • I dont need to navigate to any website. It takes three clicks to Chasey Lane and that's all that's important:
    - click on my bookmark to The Pirate Bay
    - click on Search after I type her name
    - click on the magnet link

    Alright there's a fourth click to start the movie, and I have to reach over to the box of tissues, but we're talking about clicks here.

  • Some of you have probably heard about this already, but there is this fun game... With your buddy, you both open a random article in Wikipedia. Then you decide some common article that you both try to reach by clicking only Wikipedia article highlighted words. The one who reaches that article first, wins.
    • by ledow (319597)

      I do the same with movies and IMDB. You have to link actor to movie to actor to movie, and get from anyone (or any movie) to a particular one.

      Aliens, I find, is particularly fun to try to get to and almost always the last few links involve Terminator or some such 80's action movie to get there.

      I've not YET found a movie I know that I can't link in my head to Aliens even without IMDB's help, but I'm sure the "Kevin Bacon" number for movies must be lower than 19.

      • The all knowing Wikipedia states:

        As of 6 February 2013, the highest finite Bacon number reported by the Oracle of Bacon is 9.

    • Some of you have probably heard about this already, but there is this fun game... With your buddy, you both open a random article in Wikipedia. Then you decide some common article that you both try to reach by clicking only Wikipedia article highlighted words. The one who reaches that article first, wins.

      I once proposed a similar game...called "Six Clicks to Britney's Snizz". The rules are pretty much self-explanatory, I think.

  • by Arancaytar (966377) <arancaytar.ilyaran@gmail.com> on Wednesday February 20, 2013 @05:47AM (#42953995) Homepage

    *click on URL bar*

    *enter URL*

    *hit enter*

    There, one click.

    • PS: There's actually a key combination to focus the URL bar. Ctrl-L in Firefox, apparently. So 0 clicks, I guess. :P

  • And my girlfriend wonders how it can be so easy to end up at a porn site several times a day...

    That said, it's a fairly incredible claim. That's not that many deviations of Bacon, considering how many 'deadend' sites there are out there which don't link anywhere. How many of these sites are simply referral to search engines?

  • How is this news? The author has a book called Linked (published in 2002 and actually a very good book) that already mentioned, in chapter 3, that the degree of separation is 19 (18.59 to be exact). It's interesting that it has not changed in 11 years but it's certainly not news !!!!

  • by Shavano (2541114) on Wednesday February 20, 2013 @10:01AM (#42954997)
    Because: 1. new sites are created every day initially having no links to them. 2. websites can be and are created that have no external links in them.
  • by allo (1728082) on Wednesday February 20, 2013 @10:06AM (#42955035)

    There are pages with no outgoing links. Before anyone yells "thats not part of the web", there are ingoing links, so its linked to the web.

  • Create three pages, one of which contains only links to the other two. Now "click" from one of the child pages to the other. On a more serious note - Google may well be a massive generator of links to other sites, but I fail to see its usefulness in being able to click through it to other sites - it does rely on textual input.
  • by wisnoskij (1206448) on Wednesday February 20, 2013 @10:22AM (#42955153) Homepage

    Because no one has ever created a website with no external links.

  • I have a tablet. I can't click.

  • I can do it in zero clicks.
    Press alt-left, to go back to google, type in new search, tab to result, press enter.

    • I was about to say something along the same lines, however,
      I think they are trying to say like 6% of separation between people,
      there is 19% websites, without using google to go directly to that website?

"One Architecture, One OS" also translates as "One Egg, One Basket".

Working...