Google Not (always) Indexing the Proper URL?

[Azri]

The following is the normal URL: http://example.com/default.asp?catid={50}&details_type=1&itemid={3}

For some reason Google is indexing the wrong version without default.asp which is: http://example.com/?catid={50}&itemid={3}

In both of these 2 url's you have a canonical tag leading leading to the proper version (with default.asp)

What's weird us that according to google analytics I see organic traffic to these wrong without default.asp URL's only from around mid October 2013, before that all the organic traffic was to the proper url's

Also, most of the url's (thousands) are indexed correctly (with default.asp), and this wrong indexing happens only to small number of url's

What wrong here?

What prevents Google from respecting the canonical tag?

Best,
Azri

[edited by: goodroi at 11:13 am (utc) on Oct 23, 2014]
[edit reason] Examplified [/edit]

[WebEnthusiast]

Hello,

it might be that someone have created a link to your site with
"http://example.com/?catid={50}&itemid={3}"

That is why google index it.

[Azri]

Hi,

What you say sound reasonable.
Indeed I found a links to the non default URL "http://example.com/?catid={50}&itemid={3}"

Do you think this is a problem in terms of SEO?

[aakk9999]

In theory the canonical link element should take care of consolidating all page duplication.

In practice this is not always the case. From some time ago Google has decided not to always index URL you prefer, but may use shorter version of URL if both URLs contain the same content. We had a discussion on this here:

Google now chooses my shortest url possible, ignoring 301/canonical
Dec 5, 2013
http://www.webmasterworld.com/google/4627937.htm [webmasterworld.com]