Google's Three Times Bigger!

CNet reports that Google has removed the index size from their homepage. The really interesting part, they will supply a specialized query for verifying index size. Sounds like we'll be able to go beyond the 1000 result cap at some point? That's really good news. But this test CNet did cracked me up:

In a random test, using the search terms "Joe Schmoe" and "pickles," Google returned 451 results, Yahoo returned 62 results, MSN returned 60 results and Ask Jeeves returned 54 results.

When I did the search, Google returns 182 non-duplicate results, Yahoo! 49. Just a quick spot check ... the last result Google returns doesn't contain either "Joe Schmoe" or the word "pickles", it's a "Supplemental Result". Clicking on the cached version of the page, you understand what that means exactly:

These terms only appear in links pointing to this page: joe schmoe pickles.

The earliest the terms appear are in result 161, and large numbers of "Supplemental Results" appear before that as well. I guess you could call this a nice feature ... for users, but not for estimating index size. Even when I go into Google's advanced search options and specifically set the occurrences to "anywhere in the page" I still get supplemental results. All of which is to say, the waters are muddier than they appear.

1. CNet article.
2. John Battelle's Searchblog, takes a crack at this.
3. How Big Is Your Index, a little thing I wrote last time.