BING BUSTED! Copying Google Results.

[Automata]

Even if they did it unintentionally (via the toolbar, for example), it is still stealing.

We can't argue intentions since we have no idea if they meant to do it or not. The fact that an obscure search term gives the same results, however, is quite obvious that it was copied, regardless of the means.

 

[DaveHCYJ]

I guess I don't view it as stealing because its User provided data, although I could see how some people come to that conclusion.

I think the reason its such a non issue for me is that: Yeah sure if you pick some extremely obscure term that only returns 1 results obviously Bing will do the same after a user provides that data to them. If there is absolutely nothing else for Bing to base their decision/search rank on then its obvious they will regurgitate the only data they've collected.
You would have to look at a more common term like "cars" or "intel" or something to be able to tell if their is foul play. Like if google dramatically changed their search rankings for "cars" and all of a sudden so did Bing.

At the end of the day for me the headline is just "Bing uses your opt in data in its search results" not "Google accuses Bing of stealing"

 

[Automata]

But they did a change, exactly like that. They just used an obscure term that isn't in either system. That is a lot easier to track than general terms. Especially since they could brush it off with some bs reason.

 

[bigben6]

. What are the chances that out of 24 Billion searches, not one single person mashed the keyboard and came up with one of Google's honeypot terms? Or rather 9 of their honeypot terms. Eitherway, Statistically its looking pretty plausible to me.

Here is th problem, the odds of some one mashing and randomly getting that are asurd... assuming number ARE NOT included as possibilities in a random keyboard smash there are 141167095653376 possible combinations of letter 10 digits in length, thats 141 TRILLION possibilities... if you add in the 10 numbers at the top of most key boards the possible iterations are 3656158440062976 thats over 3 quadrillion possible permutations.... to put it in perspective that means that in 24 billion searches the chances are 1.7^-4th power, thats a fraction of a tenth of a percent...

This was not chance, like i sadi above it was not directly attacking google, but it was not freak occurring, there is no way those % work out to a 12% correlation in 2 days... no way...

 

[DaveHCYJ]

Right. I guess I just don't see what the big deal is.

Going back to the cooks example:
If Cook#1 puts mud in his cookies and Cook#2 surveys 20 billion consumers (general consumers, not just Cook#1's consumers, and not just cookie consumers) and 1 of them happens to mention he likes mud in his cookies, I would not find it strange that if someone later asked Cook#2 "Hey what foods are good with mud in them" if the only answer is "cookies" and not "pasta", I don't find that strange at all. I just think google is upset that MS is doing a good job of surveying consumers. Which again points to the issue being MS using opt in data, not data theft.

Or how about a news example:
If Fox News runs a story no other network is running, and that causes a bunch of protests, and some protester uploads an iReport to CNN, which CNN then posts. Did CNN just "steal" Fox New's work? I mean no other new agency was running the story but now CNN is, so therefor CNN must be engaging in outrageous and unethical theft?

 

[Automata]

The problem is, the protester didn't create or own the information. In this case, Google created and owned it.

 

[bigben6]

They used specifically created google data and listed it as their own, intentionally or not, its sketchy...

 

[DaveHCYJ]

They used specifically created google data and listed it as their own, intentionally or not, its sketchy...

You make it sound like MS specifically targeted google. There is no way to know this. MS probably uses User opt in data in thousands of different ways.

I guess it comes down to motives and intentions. These articles and google "accusing MS of stealing" are implying MS's intentions were bad. A search engine really is nothing more than a massive phone book for the internet. I don't see the harm or ill intention of including these results for completeness sake. I mean yes obviously data ended up in Bing that somehow originated on Google and who knows how that actually happened.

But if the argument is that search engines should never include data originating on other sites in their results then you are sort of arguing that search engines themselves should not exist.

I mean if I posted on OCForums something like "hey guys check it out my madeup_space_alien_fake_name_pet_dinasour_thingy in fact has a strong correlation with catapiller dump trucks" and that somehow causes the word processing algorithms of google/bing/whoever to list http://www.cat.com/cmms/307?x=7 as the only search result if you type in madeup_space_alien_fake_name_pet_dinasour_thingy, should I be outraged because google/bing/whoever clearly stole my analysis?
No, its just a search engine doing what search engines do -- they agregate data.

 

[bigben6]

You make it sound like MS specifically targeted google. There is no way to know this.

like i sadi above it was not directly attacking google, but it was not freak occurring

I guess it comes down to motives and intentions. These articles and google "accusing MS of stealing" are implying MS's intentions were bad. A search engine really is nothing more than a massive phone book for the internet. I don't see the harm or ill intention of including these results for completeness sake. I mean yes obviously data ended up in Bing that somehow originated on Google and who knows how that actually happened.

It happened because intentional or not bing process lead to google created data being use in their search, who cares if it is hidden behind the guise of user opted data sharing...

But if the argument is that search engines should never include data originating on other sites in their results then you are sort of arguing that search engines themselves should not exist.

This is not just a site it is a search engine, other sites WANT to be found via internet searches, google does not WANT to provide results for bing...

I mean if I posted on OCForums something like "hey guys check it out my madeup_space_alien_fake_name_pet_dinasour_thingy in fact has a strong correlation with catapiller dump trucks" and that somehow causes the word processing algorithms of google/bing/whoever to list http://www.cat.com/cmms/307?x=7 as the only search result if you type in madeup_space_alien_fake_name_pet_dinasour_thingy, should I be outraged because google/bing/whoever clearly stole my analysis?
No, its just a search engine doing what search engines do -- they agregate data.

Just cause an end user provides the data does not mean bing is in the right here, if ford accepts toyota engines from consumers and the produces an identical design does that make it right just cause ford didnt them selves steal the data, it was an opt in program?!?!?

If you put an article online and a search engine finds it no its not an issue, if you want it private you can make it password protected, if someone heard about your article from a friend without ever reading it yourself and then posted the conclusion as their own, yes thats a problem... and thats the net result, BING is effectively using googles algorithm and search results and hiding behind the guise of user opt ins as a cover.... intentional or not its not right, and accidents do happen, i doubt microsoft set out to copy google, but in the end they are and it should stop...

 

[DaveHCYJ]

BING is effectively using googles algorithm and search results and hiding behind the guise of user opt ins as a cover

Bing is not using google's Analysis/Algorithm at all. If you come up with an example so extreme that it only produces a single result it obviously is not demonstrating how the algorithm works. If there is only a single input and output Every algorithm will give the same result.

Its important to see the difference between data and analysis. Search engine results are analysis, its the ranking algorithm that is so important. The input data is not valueable, its splattered all over the internet. In this case google got the data because they invented it. Bing got the data because a user gave it to them.

If google did an experiement where they significantly changed their search rankings for a meaningful word like cars, pizza, intel, etc. and then Bing all of a sudden had significantly different rankings also -- that would be a meaninful test of wether or not Bing is actively copying Google's algorithm.

The fact that google posted some data to the internet and bing later produced results based on that data does not imply bing is copying google's algorith or results -- the only thing you can infer from that is Bing is somehow ending up with data (not rankings or anaylsis or important stuff, just data) that google had posted.

 

[Automata]

The difference is, why would a user search for "mbzrxpgjys" in either search engine? The chances of Bing suddenly showing the same as Google within a few days is incredibly low, when it comes to pure dumb luck. I understand that is one example, but it literally does not make sense otherwise. Now, if for example, they had "pizza" show "Best Buy" as the top result, then I could see where that could throw off Bing since it uses data seen by the user. That is something a user would actually search for.

So, unless someone using a Bing toolbar searched for "mbzrxpgjys" and clicked the only link inside, how did they get that data? Not only that, it happened on 12+ ones that Google did. So you are telling me that a user searched for these random phrases, clicked on them, had the Bing toolbar and the information was collected, sent, processed and used by the Bing search engine? I just find that many steps too ridiculous to believe.

 

[bigben6]

Did you ignore the parts of the articles about how there was already odd correlations in place BEFORE they input the test site linked to RIM?

From the first link

Search Engine Land said Google first noticed something suspicious in October 2010, when "Bing was showing a much greater overlap with Google's top 10 results than in preceding months." There was also an increase in the number of times that Google and Bing had the same page as their top spot.

This is not an isolated incident, the specifics of the set up just verified that bing is using user opt ins to effectively mimick google when bing fails to be as good as google....


An it was not JUST data, thats like saying copying a car is ok as long as oyu cannot compy the machines that made it?!?!?

This set up wit hthe RIM/10 digit random input was a result that ONLY googles search engine was capable of returning via its own algorithm, if bing is not capable of returning the result on its own, but does anyways, how is that not an end around using googles algorithm...

you dont need the algorithm for "cars" or "pizza" or "house" if you just steal the result via your users, you may be 2-3 weeks behind compiling the results, but in th end your theifing them...

 

[DaveHCYJ]

Did you ignore the parts of the articles about how there was already odd correlations in place BEFORE they input the test site linked to RIM?

From the first link

This is not an isolated incident, the specifics of the set up just verified that bing is using user opt ins to effectively mimick google when bing fails to be as good as google....

Correlation does not imply causation.

It should not be surprising that multiple search engines come to the same conclusion. Its like asking two people what the answer to 1+1 is and then being shocked when they both say 2. On top of that there is only some overlap, not complete overlap. But maybe Bing is doing that on purpose as part of their consipiracy?

An it was not JUST data, thats like saying copying a car is ok as long as oyu cannot compy the machines that made it?!?!?

This set up wit hthe RIM/10 digit random input was a result that ONLY googles search engine was capable of returning via its own algorithm, if bing is not capable of returning the result on its own, but does anyways, how is that not an end around using googles algorithm...

Maybe I'm not understanding the article. I thought that the only data that the user was providing to Bing was "I was interested in X then clicked on Y" not "I searched for X on googles website and it told me all the results in order were 1. 2. 3. etc." Because honestly bing doesn't need a user for the later it can do its own brute force search of google.

Given the first case Bing's algorithm "knows" that users interested in mbzrxpgjys are also intersted in RIM (because the user via bing toolbar told them so). And given the obscurity of mbzrxpgjys it should be obvious that bings is only going to return 1 result and that result happens to be the same as the one hard coded by google. It in no way implys Bing stole google's algorithm.

 

[DaveHCYJ]

So, unless someone using a Bing toolbar searched for "mbzrxpgjys" and clicked the only link inside, how did they get that data? Not only that, it happened on 12+ ones that Google did. So you are telling me that a user searched for these random phrases, clicked on them, had the Bing toolbar and the information was collected, sent, processed and used by the Bing search engine? I just find that many steps too ridiculous to believe.

I guess my question is: if one of the billions of user searches didn't provide the data then how else did Bing get it? Is Bing brute force searching Google for random letter combinations? Doesn't sound likely to me mainly because what is the value of that? Search engines make money based on clicks on advertisements. If its so implausible that a user would search for mbzrxpgjys, then what is Bing's motivation for having that term in their index, they surely won't make any money off it.

Wouldn't they be better off profit wise just spending their time copying the search results of the top money making words? Or even better yet not copying them but making them even better by observing on Bing's website and Google's and Yahoo's and every other site which links users are most likely to click based on their interests and then incorperating those statistics into the Bing algorithm?

 

[bigben6]

You obviously did not read the article...

These dummy search results were added to Google on December 17 and by December 31, about 9 of the 100 tests added to Google were showing up on Bing. Google told Search Engine Land that it doesn't know why all the search results didn't show up on Bing, but even a half dozen was enough to convince the search giant that some copying was going on.

Then it created its own "honeypot page" with Research in Motion at the top of the page. Within a couple of weeks Research in Motion began appearing at the top of Bing searches.

Google was hesitant to tell Search Engine Land that Microsoft's move was illegal, and Google did not explicity address that inquiry when PCMag asked the same question.

No bing does not have googles algorithm, but again, if you just steal the result, the NET EFFECT is that they are using it... this is NOT coincidence... its sketchy and it is theft disguised by user opt in reporting...

 

[Saaby900t]

The difference is, why would a user search for "mbzrxpgjys" in either search engine? The chances of Bing suddenly showing the same as Google within a few days is incredibly low, when it comes to pure dumb luck. I understand that is one example, but it literally does not make sense otherwise. Now, if for example, they had "pizza" show "Best Buy" as the top result, then I could see where that could throw off Bing since it uses data seen by the user. That is something a user would actually search for.

So, unless someone using a Bing toolbar searched for "mbzrxpgjys" and clicked the only link inside, how did they get that data? Not only that, it happened on 12+ ones that Google did. So you are telling me that a user searched for these random phrases, clicked on them, had the Bing toolbar and the information was collected, sent, processed and used by the Bing search engine? I just find that many steps too ridiculous to believe.

So your saying google never goes to bing.com types in a word/phrase and see what it comes up with compared to google.
and then figure out what the better results are and adjust there algorithm to come up with the same data. ( I bet every engine has people that are paid to just check other sites data vs. there's )

 

[Automata]

So your saying google never goes to bing.com types in a word/phrase and see what it comes up with compared to google.
and then figure out what the better results are and adjust there algorithm to come up with the same data. ( I bet every engine has people that are paid to just check other sites data vs. there's )

I didn't say that at all and your comment is completely different. If Google went to Bing, compared their results and used them, then yes, it would be the exact same scenario. The "using" portion is extremely important to my argument. Comparing is fine, just like looking at/browsing products within a store. Once you steal the item, walk out and claim it is yours, then there is an issue.

 

[bigben6]

I didn't say that at all and your comment is completely different. If Google went to Bing, compared their results and used them, then yes, it would be the exact same scenario. The "using" portion is extremely important to my argument. Comparing is fine, just like looking at/browsing products within a store. Once you steal the item, walk out and claim it is yours, then there is an issue.

This

I think you guys ignore the fact that googles PRODUCT is their search result, the algorithm may be the nuts and bolts but if it did not produce the results it would be uselss, so who cares HOW bing gets the results, they are produced by GOOGLE and bing is passing them off as their own....

And no correlation is not causation, but if you look at the scope and timing of everything, it is not JUST a correlation, SOMETHING changed and now bing has results that ONLY googles algorithm is capable of producing, and it is not just the mxbjlkems scenario, it was in top 10's and other areas, the funky word just solidified their theory...

 

[DaveHCYJ]

No bing does not have googles algorithm, but again, if you just steal the result, the NET EFFECT is that they are using it... this is NOT coincidence... its sketchy and it is theft disguised by user opt in reporting...

Maybe we're defining "stealing the results" differently. I would define stealing the results as Bing litterally having the same set of results as google. In the world of search engines the ranking is very important, so showing an example where two search engines only have a single result does nothing to show how they are ranking multiple pages relative to each other.

You're right I only read the first article and skimmed the second, did I miss an example where Bing stole a top 10 list or something like that?

Maybe I'm just not into conspiracy theories enough?

 

[bigben6]

I'm not claiming it was malicous, but yes there was alot overlap:

I posted this quote once already...

Search Engine Land said Google first noticed something suspicious in October 2010, when "Bing was showing a much greater overlap with Google's top 10 results than in preceding months." There was also an increase in the number of times that Google and Bing had the same page as their top spot.

My impression is that bing produced a software with good intentions of making their product better without breaking any laws, but the result is that they have crossed a line, and it should stop, I by no means am implying that bing set out to just use google and regurgitate their results... and remember bing's process would apply to MSN, Yahoo, Ask jeeves, ANY search query on a browser with their toolbar...