The most authoritative site on the web....

[brooksmoses]

Google's searches now allow wildcards. If you enter * *, you get the most general possible search, given that a single * returns nothing.

Now, in theory, what this should give is everything on the internet, in order of raw PageRank. At the top of the list, according to the theory, will be the most popular, most highly-regarded, most authoritative sites in existence.

Currently, the most authoritative site in existence is the MBTA Subway schedule.

Comments

[kiya]

keshwyn says "Further confirmation that we are, in fact, the Hub of the Universe."

[beaq.livejournal.com]

MBTA, Keio Uni, and PBS! A trifecta!

[redbird]

I like a search that returns, in order, MBTA schedules, the Keio University Health Centre, and PBS.

[leback.livejournal.com]

So having just poked a bit at this wildcard search thing, I am not impressed. Isn't a wildcard supposed to represent any string of characters including the empty string? Throwing one into the middle of a two-word quoted phrase just failed to find me any hits containing the phrase itself, which seems not-right...

[dendrophilous.livejournal.com]

As you add asterisks you get the American Association for the Advancement of Science and Fermilab. But PBS outranks them.

[worthyadvisor.livejournal.com]

Well, it just confirms for me that getting around on the T is not as easy as it could be. :D

Love it!

[pariyal.livejournal.com]

When I tried I got the MBTA fourth; first and second are Wikipedia, third is the Keio University Health Center of all places. I wonder if results are different for different continents or something (anybody in Australia, Africa or Asia who can comment?)

[brooksmoses]

The particular order does seem to change frequently -- now, for me, it's Wikipedia, MBTA, Keio, and PBS. (And the MBTA site is "projects underway" rather than "subway schedules". Big Dig, anyone?)

[brooksmoses]

Well, arguably, Google's system is based on doing word-by-word searches. And so an asterisk matches a word made up of any string, including the empty string -- except that Google doesn't get any empty-string words in its database, so that's match made only in theory.

Or so I was going to claim, until I actually tried it. It seems that a search on "this * that" finds phrases with one word, two words, three words, and so forth, between the words. And it does definitely seem wrong if they're doing that to not include the zero case. (Except that the pattern seems to be * meaning one or more, * * meaning two or more, and so forth, and so I suspect it would be confusing to try to support that and also allow the zero case.) Bah!

And they've taken out the old way of doing wildcards, which matched exact wordcounts -- it used to be that it ignored common words like "at", so placing "at" in a phrase amounted to an exactly-one-word wildcard. But they've stopped doing that now; a Google search on "a" returns results. (Specifically, aside from the stock with "A" as the ticker symbol, www.apple.com is the first result.)

[bungo.livejournal.com]

At Fri Sep 30 09:05:40 UTC 2005 in Berlin, Germany:
http://www.google.de/search?hl=en&q=*+*&btnG=Google+Search
just gave me Wikipedia:Copyrights, Wikipedia:Stub, Keio, and MBTA.
So the same as you.

[sraun]

My understanding is that Google has several never-quite-identical databases - they aren't doing real-time data replication. You get a random server, which hits a random database, which may or may not return identical hits to someone else. This looks like it may be one of the few cases where you can actually tell the difference!

[sraun]

It's more random than that - see my reply to brooksmoses

[micheinnz.livejournal.com]

New Zealander here. Wikipedia first, MBTA second, Keio University Health Centre third, PBS fourth.

Data may be subject to change.