No index, follow vs. canonical url

[grayloon]

We have a site that consists almost entirely as a directory of videos.

Example here: http://realtree.tv/channels/realtreeoutdoorsclassics

We're trying to figure out the best way to handle pagination and utility features such as sort for most recent, most viewed, etc.  We've been reading countless articles on this topic, but so far have been unable to determine what might be considered the industry standard.

Two solutions seem to stand out...

Using the canonical url on all the sorted and paginated pages.  However, after reading many blog posts, it seems that you should NEVER use the canonical url to solve the issue of paginated, and thus duplicated content because the search bots will never crawl past the first page leaving many results not in the index.  (We are considering ruling this method out.)

Another solution seems to be using the meta tag for noindex, follow so that a search engine like Google will crawl your directory pages but not add them to the index themselves.  All links are followed so content is crawled and any passing link juice remains unchanged.  However, I did see a few articles skeptical of this solution as well saying that there are always better alternatives, or that there is no verification that search engines obey this meta tag.  This has placed some doubt in our minds.

I was hoping to get some expert advice on these methods as it would pertain to our site.

Thank you.

[john4math]

Here's a similar Q&A post: http://www.seomoz.org/q/canonical-pagination-content.  The answer there suggests adding a view all link, and then setting rel=canonical to all of your paginated search result pages to the view all page.  They also suggest this on the search engine roundtable blog here.  A good point is that since these pages have different search results, if you try to rel=canonical these pages, there's a good chance it'll be ignored.

[grayloon]

Thank you for that response.  I wanted to follow up a little on it.

The article you link to sounds good.  My concern, however, is that if I were to have a "view all" page then we could be talking about 1,000+ nodes on a single page.  This page would take much longer to load, which itself would become an SEO hit.  Though I do get why that's a suggestion, I have to wonder if that's the best solution for sites with large directories.

The other article/post seems to mention both the "view all" method and the "noindex, follow" method which gives a little more validity to that option in my opinion.  I'm still trying to discern what the "best" method is, and it's starting to become clear that maybe there isn't exactly one industry standard for this.