Google Drops The Home Page Count

[dannysullivan]

Google's decided to drop the long-time count of pages it searches from its home page. It's the latest move in the Yahoo-Google dispute over size. Google says it is now three times bigger than its competitors (that's Yahoo) and is the most comprehensive. But it's not citing figures feeling you no longer can do apples-to-apples comparisions. I agree with that, basically feeling the dropped home page count may defuse things. Ultimately, both Google and Yahoo are leaving it to the users to decide. I wrote up much more about this in today's SearchDay article, End Of Size Wars? Google Says Most Comprehensive But Drops Home Page Count

[Jeff Martin]

Is this like when the kid who brought the ball to the game is being beat and he grabs his ball and says "I dont want to play anymore"?

Google made this more of an issue than it needed to be and now Yahoo! is trying to play their numbers game. I agree that the numbers arent the definitive answer to quality but to Joe User (not an SEO/Mer) that number staring them in the face may have been important.

[hiero]

Ya, but how many Big Macs have they sold?

[orion]

For those interested in a little inside scoop, check here http://forums.searchenginewatch.com/...2253#post62253

That relevancy is more important than size or the lose notion of comprehensiveness is a no brainer. Regarding the comprehensive part, that will start another volume war.

Even if search results are more comprehensive that does not mean they are more relevant. The real war is for relevancy.


Orion

[samd]

Google today put up a challenge. It went something like this:

Our Index is biggest, try a query anywhere else and then with us and we'll give you more answers.

This might well be the case. But when testing it out with MSN search I noticed a little hitch. On the first results page the total number of results on MSN is always MUCH higher than the real number of results available.

For example I searched for Oranges. MSN says it has 2,009,389 results.

I flick to page 17. It stills gives a total of 2,009,389. (see image).

But then on page 21 suddenly the number it has in the index is 209! (see image).

It seems MSN want to make their index look large but they don't actually have the results to back it up. On page 21 it had to admitt there weren't several million results because the results ran out on that page.

I have tried this with several other queries and the search engine consistently indicates a larger amount of results than it has on the first page.

I can understand a but of leyway (in Google the figure is approx) but quoting 10,000 times bigger!

That is just lying.

[glengara]

Very American, size matters, quality doesn't, ho-hummm......

[Gurtie]

that size and relevancy are different things is obvious, but you can't escape the fact that the larger your index size the more likely it is to contain the most relevant document - perhaps not so much for the commercial searches but certainly for some of the obscure ones.

If there was a foolproof way of pulling the most relevant pages then size really would matter. There are searches now where I know the most relevant pages aren't returned by some SE's because they aren't indexed, so an increase in the size can only improve those searches.

[glengara]

AFAIK G never shows results past 1000 if that, so who knows/cares how many they may actually have?

[Gurtie]

my point is if you never bothered to index the walrus cleaning kit page on cleaningseaboundmammals.com then you can't return the page.

If you indexed it and someone searches for "walrus polishing and cleaning equipment" then you may find that page and return it within the top 1000 of the 27,000 results. Therefore you may return the most relevant result. Otherwise your users walrus will never be completely clean.

It's a niche result yes, but it happens to be the official page for that niche. Large indexes must by definition be best provided that you can also search them effectively.

[glengara]

Well I'm sceptical, when you start obsessing about size, you're more likely to include any old rubbish you might otherwise have excluded ;-)

[Gurtie]

oh agreed; total index size is irrelevant for any individual search. So is any page which is simply a dupe for whatever reason, scraper pages of any type, etc etc etc. But by that criteria 90% of the pages already indexed are pretty useless.

but to return the most relevant results for any one query you need to be able to say which pages on the web are most relevant for the query and we already know that number of inbound links, what sites point to the pages and length of time since publication aren't especially good indicators of how good a page is - so you can't rely on only indexing some pages and assuming you have all the quality ones, you have to index them all and thenonly show the quality ones

[redclaw]

http://www.google.co.uk/search?hl=en...e+Search&meta=

I know this is trolling a bit, but do you think that in that tottal count, they are including error pages and those that they have, themselves, scanned in from Yahoo's sites?