How bloggers game Google

[Daniel Brandt]

1. Start with a juvenile Google bomb prank:

March 13, 2003. A small-time, pro-war blogger puts up a page that resembles a Google page. You find it by looking for "French military victories" in a Google search and hitting "I'm feeling lucky." The trick works because his page turns up number one for these keywords, and the "I'm feeling lucky" button always takes you directly to the top page of the search results. This blogger seeds links to this page by inserting comments with links behind them in other blogs. He gets his 15 minutes of fame in the media, but everyone expects that the joke will fade soon because it's dependent on Google's freshbot results, which are insanely generous to bloggers. After the monthly deep crawl, the prank page is expected to be lower in the search results. So it's ha-ha, very funny. Script kiddies everywhere are laughing, while Iraqi civilians are dying. Screen shot of the prank page.

2. A big-time blogger links to it:

This blogger is running a scam to inflate his Google juice. Not only does the prank page survive the next monthly update, but it ends up with a respectable PageRank of 6, and 138 backlinks. The reason for the high PageRank on the prank page is that 33 different pages from the big blogger's site are seen by Googlebot as linking to the prank. Each of these pages on big blogger's site has a PageRank of 5. The prank survives the update as number one for "French military victories." Screen shot of Google's backlinks to the prank page. If you scrolled down and counted all the links from www.shellen.com, you would see that all 33 of his archive pages repeat the link to the prank.

3. Jason Shellen's game:

How does Jason do it? When you fetch any of his archive pages, there's a table on the left margin that also shows current postings and links. You can see the start of this table on this screen shot, taken on April 30. If you were to scroll down, you would see that he has postings from April 18 to April 29 in this table. But the page itself, as indicated by the URL and the right two-thirds of the screen, is from July 31, 2002.

4. What the Googlebot found:

When Google did the regular deepbot crawl in March, it found the prank link on each of the 33 archive pages on Jason's site, since those were current at that time. Here's a screen shot of the top half of Google's cache copy. Note that the table on the left side has now switched to the right side. This is a minor glitch and doesn't affect our analysis. Google injects an extra table on top to brand their cache copy, and all these tables together confuse our browser.

5. Here's the prank link:

This screen shot is the bottom half of the previous page. On the lower right, you can see the words "French Military Victories?" What you cannot see is that the link behind these words leads to the prank page. Remember, all 33 of Jason's archive pages, one for each month since July 2000, have exactly the same link on them. Each of these 33 pages is a PageRank of 5, because Jason is a big-time blogger. Google tries to detect duplicate pages, and accepts only one (or none) of them when found. But only about one-third of the page is duplicated in this case, so Google thinks they're all worth indexing.

Will this juvenile prank stop working after the next update? Let's hope so. Meanwhile, Jason's Google juice is the biggest scam of all. But we're thankful that so far we haven't found a blogroll on his site.

http://www.google-watch.org/gaming.html

[Derek McMillan]

Thanks Daniel I had wondered about that.