Canonical tag help

I would like to add canonical tags to my site to eliminate the duplicate listings that google webmaster tools reports.

The site has a product page that serves 3 purposes.

1) When viewed without a query string it displays a list of categories for the visitor to choose from. i.e. (website)/widgets.asp

2) When the visitor clicks a category it displays a list of products in that category using the same page but with an added query string. i.e. (website)/widgets.asp?cat=1

3) When the visitor clicks a product it displays the product detail, again using the same page with a different query string. i.e. (website)/widgets.asp?prodcode=WIDGET1

Most of the duplicates are for the product detail page with links that have old query parameters, or referring site parameters like:
(website)/widgets.asp?prodcode=WIDGET1&style=blue&refid=someaffiliate

Using the same page seemed like a great idea at the time, but is causing a few problems. Like how to add the canonical tag. If I do it to follow the above example it would look like:

1) choose a category page
<link rel='canonical' href='(website)/widgets.asp' />

2) choose a product page
<link rel='canonical' href='(website)/widgets.asp?cat=1' />

3) product detail page
<link rel='canonical' href='(website)/widgets.asp?prodcode=WIDGET1' />

The problem I am concerned about is that they all point to the same page and the 'choose a category page' canonical tag does not have a query string. So does that tell google, and others, that any time it sees this page with a query string, like the cat= or prodcode=, to ignore the query string, essentially nullifying the product listing and product detail pages canonical tag? Or does it confuse and possibly create some infinite loop issue.

A redesign of the site is in order to correct the problem eventually and use different page names/folders. But for now I would like to get rid of the duplicate webmaster listings.

Any advice, thanks in advance!

Using the canonical tag is to 'help' eliminate duplicate content issues, but not all search engine have implemented it.

The only definite resolution to your duplicate content issues is going to be through using the 301s when the request uri does not match what your script has determined to be this pages canonical url.

The urls with the old query parameters: you can use googles web tools to instruct google to ignore those vars/values (like you would with a session id), but this is just for google. The issue still exists for other engines. 301 em...

Vinny, thanks for the reply. Many of the url's with the additional query parms are valid, so a 301 doesnt apply here. They are additional parms like referrer id from another site and in some cases even from my site. After reading on canonical tags it seems to be the perfect solution to let bots know the true url, it even seems to be what they were designed for.

widgets.asp?prodcode=ABC&color=red is a valid url but it doesnt really change the content of the page except to say the widget color is red. So widgets.asp?prodcode=ABC would be the correct canonical tag?

My biggest question is how the bots will handle a canonical to the same page where one does not have any query string and others do. In other words does "canonical=widgets.asp" override or nullify "canonical=widgets.asp?prodcode=ABC"?

Yes, that should be fine.

in the example of ref/affiliate ids: you absolutely want to 301 those requests! stash the ref id in a cookie or a session value, then 301 to the pages standard url. otherwise, you rish engines indexing the page with the affilaites id, and you're paying them for your organic traffic!

Again, adding that ref id in gwt can't hurt either. You're telling google to remove these values from indexed urls.

I can understand using canonical tags and redirects where you really have to. But in many cases they can be avoided by good and early planning.