Coffeehouse Thread

18 posts

Forum Read Only

This forum has been made read only by the site admins. No new threads or comments can be added.

Google: Bing is copying our search results

Back to Forum: Coffeehouse
  • User profile image
    Bass

    Google went public today with some evidence that Microsoft is copying search results directly from their search engine. Apparently some Google engineers did some experiments with adding fake search results originating from largely random strings and found the same results mysteriously pop in Bing.

    http://searchengineland.com/google-bing-is-cheating-copying-our-search-results-62914

  • User profile image
    Cream​Filling512

    That's not what they found at all.  They discovered that the Bing toolbar let's the user opt-in to send clickthrough data to "improve search results".  And so if you click on search results on Google, the clickthrough data gets sent to Bing and is incorporated into the relevance rating.

  • User profile image
    Blue Ink

    I see a minor flaw with the experiment... if I understood the article correctly:

    Google makes a bogus search term to return a list of results the first of which is an unrelated page (the "honeypot"). Engineers start clicking on that link and in a couple of weeks Bing also starts reporting that link as the top result for the bogus search. Bing is exploiting Google's rank, QED.

    Here's the rub... what would have happened if Google had made the honeypot appear *last* on the result page and had its engineers click on that anyway? I suspect that Bing would have still reported that link as the top result. The ranking would be totally different, and it would be really hard to contend that Bing is exploiting Google's rank.

    My wild (really wild) guess is that Bing is just collecting the URLs users navigate to and feed those to their indexer, exploiting user's judgment to find obscure relations between search terms and web sites (Google's redirection would make this pretty efficient). The fact that Google is the most popular search engine by a mile, and the fact that the top search result gets an inordinate amount of click love, would make Bing's results more similar to Google's than one might expect, but if my guess is correct, it would be just be a welcome (but not necessarily intentional, or malicious) side effect.

  • User profile image
    contextfree`

    OK, out of curiosity I actually tried installing the Bing bar (on IE9 beta, which, btw, makes you manually activate add-ins after they're installed. The installer for the addin itself is pretty upfront about it sending your click data and stuff - it's right there on the one and only options page, next to one of three checkboxes - though the box is checked by default which I think is dubious). I haven't been able to influence the Bing search results (no surprise there, since I've only spent a few minutes on it and not weeks like the Google folks) but one thing I did find very interesting was that if you search for something on another site, the Bing bar actually lights up and populates its own search field with your query so that with another click you can search for it on Bing. As far as I can tell, the bar doesn't seem to be using any heuristic to tell what is a search query but just has a list of sites/URL patterns it knows about. Besides Google and Bing itself, these include Wikipedia, Yahoo, Ask.com, Amazon, Facebook, eBay, YouTube, MSDN and IMDB, but not Twitter or DuckDuckGo. This doesn't prove anything about how it's feeding search results of course but it does at least suggest that the Bing bar is very interested specifically in search queries, though not only on Google.

  • User profile image
    cynthia399

    Also, goggle had about 100 terms that they were using but only 7-9 of them turned up in bings results.  So microsoft is obviously doing something else with the results from the toolbar data beside putting it directly into the results at bing.

     

    This also could be spun as goggle tring to figure out how bings algorithm works.

  • User profile image
    magicalclick

    I don't think so. I searched Halo, and they are different orders. Bing has the correct official site and wiki up-front. Google puts the developer's site who no longer owns Halo franchise up-front.

    Leaving WM on 5/2018 if no apps, no dedicated billboards where I drive, no Store name.
    Last modified
  • User profile image
    Dovella

    , Bass wrote

    Google went public today with some evidence that Microsoft is copying search results directly from their search engine. Apparently some Google engineers did some experiments with adding fake search results originating from largely random strings and found the same results mysteriously pop in Bing.

    http://searchengineland.com/google-bing-is-cheating-copying-our-search-results-62914

     

    http://www.bing.com/community/site_blogs/b/search/archive/2011/02/01/thoughts-on-search-quality.aspx

  • User profile image
    Dovella

    Google's Childish Response To Microsoft Using Google To Increase Bing Relevance 

  • User profile image
    fanbaby

    One troubling thing (for Bing users, that is) is that they took the info from the user-toolbar interaction, and incorporated it as is into bing's results. WITHOUT EVEN CHECKING THE PAGE! In their test, google did this:

    They artificially made the search for the word "akjsdlfjalsdjf" (for example, i dont have the actual words) return the page xxx.yyy.com. The page itself has no mention or relevance to the search word.

    After a few weeks it apeared in bing!!! WTF?

    On a different note, to me it seems that if Google finds an obsecure page, Bing would learn about it from the user-toolbar interaction. kinda cheap in my book.

  • User profile image
    Ray7

    , fanbaby wrote

    One troubling thing (for Bing users, that is) is that they took the info from the user-toolbar interaction, and incorporated it as is into bing's results. WITHOUT EVEN CHECKING THE PAGE! 

    Not really that surprising at all since Bing is not actually hijacking pages from Google. The search bar looks at the search term and then looks at the item that the user clicked on. The results are then posted back to the Microsoft. This is what every search bar does, so I find it really hard to get excited about it, to be honest. 

    If Microsoft did check the page then add it to Bing then that would be more of a problem, because then it would go beyond the user data mining that Google has been performing across its product range for years and years.

    And if I were a company that had been accused of writing a whole operating system using someone else's code, then I'd be a little less likely to start throwing stones.

     

  • User profile image
    Ray7

    What I also find bizarre is Engadget's reporting on this.

    The accusation by Google gets an article.

    The response from Microsoft does not

    Google's response to Microsoft's response - quel supris - gets another separate article.

    Folk need to remember that blogging doesn't necessarily equal journalism.

     

  • User profile image
    magicalclick

    @Ray7: like it ganna make a difference. Fanboy don't read it anyway, just like someone already stated the reason in the first reply, and they still don't read it. It is futile.

    Leaving WM on 5/2018 if no apps, no dedicated billboards where I drive, no Store name.
    Last modified
  • User profile image
    Bas

    @Bass: I love that you only post about the initial claim and conveniently neglect to mention the entire fistfight that came after it, which puts everything in a totally different light. But hey, keep on trollin'.

  • User profile image
    W3bbo

    A few moments ago, Microsoft just announced the next Bing redesign phase. Here's the new logo.

  • User profile image
    brian.​shapiro

    If Microsoft really combed through 8 billion links on Google's site it would be evident through heavy site traffic from MS' servers there would be no need for experiments.

    What Microsoft is saying sounds right, it just got the results through user click data.

    Anyway, anyone see this:

    http://www.neowin.net/news/bing-engineer-accuses-google-of-profiting-from-search-spam

  • User profile image
    cbae

    , Blue Ink wrote

    I see a minor flaw with the experiment... if I understood the article correctly:

    Google makes a bogus search term to return a list of results the first of which is an unrelated page (the "honeypot"). Engineers start clicking on that link and in a couple of weeks Bing also starts reporting that link as the top result for the bogus search. Bing is exploiting Google's rank, QED.

    Here's the rub... what would have happened if Google had made the honeypot appear *last* on the result page and had its engineers click on that anyway? I suspect that Bing would have still reported that link as the top result. The ranking would be totally different, and it would be really hard to contend that Bing is exploiting Google's rank.

    My wild (really wild) guess is that Bing is just collecting the URLs users navigate to and feed those to their indexer, exploiting user's judgment to find obscure relations between search terms and web sites (Google's redirection would make this pretty efficient). The fact that Google is the most popular search engine by a mile, and the fact that the top search result gets an inordinate amount of click love, would make Bing's results more similar to Google's than one might expect, but if my guess is correct, it would be just be a welcome (but not necessarily intentional, or malicious) side effect.

    The biggest flaw in the experiment is that Google set up one "honeypot" page for each bogus search. Therefore, Google's experimental search results each yielded one page that was both the "top ranked" and "bottom ranked". There WAS NO OTHER LINK to click. Therefore, when Bing's search turned up the one link, the only thing Google engineers proved is that Bing's optimization methodology worked.

    There are two important aspects of search results: the number results returned AND the rankings of each result. Google engineers took the trivial case and declared a correlation.

    These guys must have gotten Fs on all the labs they did in their undergraduate work. 

     

     

     

  • User profile image
    kettch

    @cbae:Whether the methodology is valid or not, they've still accomplished their goals by poisoning the internet with this FUD.

  • User profile image
    cbae

     

    , kettch wrote

    @cbae:Whether the methodology is valid or not, they've still accomplished their goals by poisoning the internet with this FUD.

    That FUD was quickly killed by a number of sites, and this news has already died off. The only people still grasping at "Microsoft is stealing our search results!" are ABMs.

Conversation locked

This conversation has been locked by the site admins. No new comments can be made.