March 5th, 2007
Google Search: Is PageRank reliable?
How reliable is Google's beloved PageRank based search algorithm formula?
Google blogger Matt Cutts points to a fellow Googler blog post, that of Peter Norvig, Director of Research at Google.
Cutts’ Google shout out warns of self-selection bias in toolbar driven metrics.
One of the first commenters at Cutts’ blog, Michael Martinez, perceptively put forth:
"The Toolbar Wars are over.”You’ve won the war, Google. No need to keep fighting the battles.
Phillip Lessen countered, however:
Michael, I don’t think Matt is fighting a toolbar battle here, he’s fighting the false widespread impression that Alexa stats are reliable.
Alexa stats are not reliable, but are the results of Google’s $140 billion dollar market cap touted “perfect search engine” reliable?
Perhaps it is time to fight a false widespread impression that Google SERPs are reliable!
I have dissected at length previous Cutts’ pronouncements and the Google PageRank formula, SEE “Google search PageRank excludes relevant Websites and Google PageRank: Biased and fundamentally flawed?.”
I concluded PageRank does not support the Google assertions that it supports development of “the perfect search engine.”
Adam Bosworth, Vice President of Engineering, Google, would undoubtedly not disagree that Google co-founder Larry Page’s desire to create a search engine that "understands exactly what you mean and gives you back exactly what you want" has not been fully realized.
I recently spent an evening at the NYC Googleplex, chatted with Bosworth, and enjoyed his presentation: “Physics, Speed, and Imprecision: What Works and What Doesn't in Software, and Why."
(See “Google on Search: ‘Natural Language works when it isn’t’” and “Google’s Adam Bosworth to NYC technologists: Speed rules”)
Before an audience of about 200 New York City technologists, Bosworth expressed confidence that fuzzy problem resolution “works really well” in search because “choices are expected anyway and the magic is just ranking.” Search ought not be treated as a natural language problem, because there would be no room for error, Bosworth conveyed.
Imprecision is better than nothing and searchers "don't know if we are right," Bosworth put forth. Google makes educated guesses and people are pretty forgiving, he indicated.
Fuzzy problem, imprecision, educated guesses…do not seem to support a “perfect search engine” qualification.
Should Google searchers be “forgiving,” or should they expect higher standards from the world’s largest search engine, Google?
ALSO: Google warns of Microsoft, Yahoo competition
