Shunir Shah at the SocialText blog marvels at
the prominent presence of Wikipdia pages in Google search results, and claims that Wikipedia is the highest ranked site
in the index. Surely, he doesn't mean that Wikipedia (or any of its pages) has the highest PageRank.To illustrate, here
are some PRs for Wiki pages:
Wikipedia home page: 8
Wikipedia English home page: 3
Featured article page, October 2005: 4
Featured article page, September 2005: 4
Wiki entry blog: 7
But even making the distinction between PageRank and page position (since relevance to the keyword string helps
determine page position), his conclusion is hard to accept—even if Steve Rubel
buys into it.
Shah's conclusion is based on a search string (*+*) that presumably spills the entire Google Web index. That search
query puts the Wikipedia Copyright page to the number one position. (That page has a PageRank of 8.) Wiki takes the
first four positions, with two obscure entries occupying the third and fourth slots. We can believe this as a
reflection of superlative ranking only if we also believe that the fifth slot, taken by the health center of Keio
University, enjoys better ranking (and better relevance to "*+*") than Google, Yahoo!, MSN, CNN, and dozens of other
likely candidates. Shah attributes Wiki's fine placement to its dense intralinkage and social structure—attributes for
which Keio University, so far as I know, is not renowned.
No, I think something must be wrong with Shah's method. I was intrigued by the test search string of two asterisks
separated by a plus sign (the AND operator). Playing with it, I easily arrived at a search string that delivered more
results, undermining Shah's premise. Here are the results of a few experiments:
*+*: Shah's string yields 10.98-billion results; Wiki on top.
**: 10.98B; Wiki.
*: 0 results.
***: 10.99B; PBSkids.org on top.
****: 11.77B; PBS.org.
*+*+*: 10.99B; PBSkids.org.
*.*: 11.76B; Wiki.
There's no question that Wikipedia is doing well in Google. But any magic formula used to explain its success must also be applied to Keio University and the Public Broadcasting Service. Like Einstein trying to crack the unified field theory, Shah has more work to do.
1. Another interesting result set is the search: (* and *) which returns 8.46 billion pages. I've long suspected that Google doesn't really ignore "and" as it tries to imply.
Posted at 4:42AM on Dec 19th 2005 by Glenn Fannick