decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books
Your contributions keep Groklaw going.
To donate to Groklaw 2.0:

Groklaw Gear

Click here to send an email to the editor of this weblog.


To read comments to this article, go here
A Search Engine Mystery Solved
Monday, November 17 2003 @ 10:17 AM EST

Slashdot has a followup on the story we first broke about peculiar results on MSN compared with Google. I read through all the comments carefully, and picking through the trolls and the shills, I found that some think the phenomenon may be the result of MSN having first paid-for listings, followed by all the rest, the real results. However, it turns out there are no real results. You can't escape paid results anywhere on MSN. Moreover, it isn't results by computer algorithm alone; human editors are involved in filtering the results you get. Here is what I found out and how.

First, I noted that one person on Slashdot reported running a search for "George W. Bush" and got some odd results:

"Try George W. Bush. I just did. It'll say something on the order of 301 hits. Scroll through the pages to the last hit. Suddenly, the number jumps through the roof. I find it hard to believe that the first 300 hits are all sponsored links. I think something else is going on here: MSN has not only sponsored links, but some kind of edited directory scheme going here, and it doesn't care to let you know that the first number it quoted is of those links which are sponsored or added editorial, and the second number is a raw search result."

I did the same search and I agreed it couldn't be all paid results, unless the White House pays MSN for putting an official bio of the First Lady high on MSN's list. I don't think they'd use our tax dollars to do that. But sure enough, when you scroll through the 305 results they first promised to show you, if you continue by clicking "Next" again, it jumps to 1,153,228 results, with no explanation on how 305 just became a million plus.

I next did a search on MSN for "search engines" (with and without the quotations) and search and invariably, MSN comes up first. I never found Google at all. I stopped looking for it after the 300th result for search engines and 200 for search. Google was simply not findable in any reasonable way on MSN. Maybe you can find it, but I tried twice and I couldn't. Results don't seem to be identical when you run searches at different times.

They listed things like "Internet Public Library" and "Recipes search engines" and "Nerdworld", "Looksmart", "LinkMe.com", "Napster", and "Korean Search Engines Dot Com", and even a dead link to an old 1996 CNET article called "Can You Trust Your Search Engine?" -- but no Google. What possible algorithm could make that happen, without human intervention? There has to be something wrong when you can't find Google in a search for "search engines". They want to buy it, but they don't want you to find it?

I then went to Google, and I ran the same searches. You can find MSN on Google just fine, on page 2. It doesn't list itself first, either. They are number 4, with Yahoo and Alta Vista ahead of it when you do a search for "search." When you search for "search engines" you get helpful things like Search Engine Watch, number one on the list. By now, I'm thinking maybe MSN just isn't a good search engine if you are looking for actual information, as opposed to what MSN will let you find. Was it true, though, that after the first few hundred paid search results, you could reach the rest of the unskewed results?

Rather than assume, why not ask Microsoft, I thought? Surely they know how they built their search engine, no?

If you go here and then click on "About MSN Search results" on the top of the list, you get to their page describing the results you may get on an MSN search:

"Depending on what you search for, the following categories of results may appear:

"Popular Topics
"Featured Sites
"Sponsored Sites
"Web Directory Sites
"Web Pages
"Broaden Your Search

"If any of the above categories don't appear in your results, it simply means that no Web sites in that category were relevant to your search."

Each item is clickable to more information explaining what it means. Microsoft has it set up so that you can't link there directly, natch, but if you follow the links, here is what you will find for these items:

First, Popular Topics:

"Popular Topics results help you refine your search by suggesting related topics. Clicking one will start a new search and display a new results page. The most relevant or popular topics will be displayed first.

"Note
"Popular Topics results appear at the top of the first search results page, but won't appear for all searches."

Next, Featured Sites:

"Featured Sites are links that MSN Search editors believe are likely to be particularly relevant and useful. These sites are chosen from ones published by MSN affiliates, partners, sponsors, and advertisers, as well as other sites proven to be especially popular among our users. Featured Sites that best match your search words are drawn from:

"The top sites for news in entertainment, sports, business, and politics.
"The most popular musical artist sites for biographies and song samples.
"MSN Encarta for encyclopedia information.
"MSN content.
"MSN content partners.
"MSN advertising partners. (Microsoft accepts payment for listings from these.) "

Next, Sponsored Sites:

"Sponsored Sites are paid links provided to MSN Search and other Web search engines by a third party. The third party ranks the sites based upon bids received from advertisers, as well as their relevance to search words and phrases.

"To highlight their special nature, MSN Search labels sponsored sites as such.

"Sponsored Sites that best match your search words appear:
"Only when you perform a basic search.
"On the first page of results.
"On subsequent result pages if additional Sponsored Sites are available.
"When your search words are terms that Web sites have bid on."

Next, Web Directory Results:

"Web Directory results contain Web sites within the MSN Web Directory that best match your search words.

"Note
"Within Web Directory results, there may also be links where the Web site owners have paid for the expedited review of their site or for clicks to their site. These sites are ranked using the normal algorithm applied to all links within each section, with no change in rank due to payment."

Next, Web Pages:

"Web Page results include all other Internet-wide Web sites that best match your search words.

"Note
"Within Web Page results, there may be links where the Web site owners have paid for either expedited review of their site or paid for clicks to their site. These sites are ranked using the normal algorithm applied to all links within each section, with no change in rank due to payment."

So it seems there is no way to escape paid-for results on MSN, no matter how hard you try. That isn't the most alarming part. The scariest on the entire list to me is the Featured Sites explanation, about the "links that MSN Search editors believe are likely to be particularly relevant and useful". There's the human intervention. Now you're talking scarey. This is, after all, Microsoft.

Useful to whom? To Microsoft or to me? If I run a search for "search" I probably do want to know about Google.

If I am looking for info on GNU/Linux, I probably don't want MS editors deciding for me what is most useful. And if I am looking for facts about the government or whatever, I especially don't want humans with an agenda, any agenda, filtering for me. I have a brain that I trust to do that filtering. It's one thing for a company to want to conrol a market; it's another when it tries to control what you know.

So what, you may say? Just don't use it, if you don't want paid results. Trust me, I don't and I won't. This was strictly in the line of duty. I will never use it again, and I will explain to everyone I know what I found out about MSN Search. But if they bought Google? Then what?

So, now you know what Microsoft thinks a search engine should be: just another way to use customers to get a competitive advantage. They have no concept of the public interest, I discern, from the design of their search engine. It's all about Microsoft and their friends. That same blind spot is likely what keeps them from understanding the value of the GPL and the freedoms it affords users.

And that, ladies and gentlemen, is why I use Google instead of MSN and will throw up if Microsoft buys Google. Then I will stop using Google. After that, some genius or other will just write another search algorithm and I'll use that search engine instead. I hope you are working on it now, actually, whoever you are. Release it under the GPL, will you, so Microsoft et al can't buy it and ruin it? That's the thing about freedom. Humans just can't stop wanting it. We're wired that way.

If Microsoft were not a monopoly, and if they didn't have MSN set as the default search engine, maybe none of this would matter much. It would be repulsive, but it wouldn't matter. But they are and they do.


  View Printable Version


Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )