"site" and "site+inurl" commands showing different results

[searcherati]

To check saturation of our listings at Google I've been using the "site:www.mysite.com inurl:www.mysite.com" command.

Today I noticed that the while the above mentioned command was yielding about 1 million pages (of a 500,000 page website!) the "site:www.mysite.com" was yielding only 400,000 pages)


So, three questions:

1. Which is more accurate of the two commands to determine Google saturation?

2. Is there a better command to do this?

3. Why are there 1 mill. results for a website with half the number of pages??

P.S. We don't use Session IDs. We do use tracking URLs for our paid search campaigns and others, but definitely not for every single page on our website!

[softplus]

Google's pushing junk into the index to inflate it, imho. It's just a matter of time and then those sites that Google "helped" will be penalized for publishing duplicate content. :mad: :mad:

Search for "sessionid" or whatever you want, Google used to filter quite well, now it indexes so much junk, that it's hard to keep your serps clean. I know a site that has approx 1000 real URLs, Google has pushed the count up to 7000! All with sessionids (found who knows where [perhaps the googlebar? hard to tell], Google doesn't get them served) and URLs with a strange mix of parameters that would hardly be found on the site or in the wild (i.e. wrong order, missing parameters in between, etc.). It's just a matter of "when" not "if" and the site will probably be pushed into the sandbox... :mad:

Also, Google is having a field trip with IIS servers, pushing all case-variations of the URLs it finds into the index (it's great to see "index.asp" and "Index.asp" in the serps, but you can bet Google will "clean it up" and possibly take out both). Even my GSiteCrawler can check if a server is known for case-insensitive URLs, but I guess that's a lot to expect for a group of high-class search-engine scientists (yes I know, URLs are by definition case-sensitive, so they're more or less following the rules there, but it still bothers me :-)).

oops, enough rant :D

Another variant to try is "inurl:www.mysite.com com" (not in quotes, note the extra "com" in the end, the TLD; this again gives a different count... perhaps the same as "inurl:www.mysite.com -asdasdfasdfafsasd"?)

Cheers!