Robert_Charlton - 1:01 am on Aug 15, 2011 (gmt 0)
Noindex/nofollow thin content (just for googlebot)
Several comments on this point. Google's intent with Panda is to improve the satisfaction of Google users with results delivered by Google search. That involves, among other things, returning not only relevant pages, but also sites that engage, inform, sell, entertain... whatever. The site needs to hold the visitor's attention.
I see noindexing as a short term band-aid for a site, to be used only on a limited number of pages. It's a good way, eg, to keep a page not intended for general users from showing up inappropriately in the serps. You definitely don't want to have very many searchers landing on a page and then backing out of it and returning to Google.
But Google is looking at user engagement sitewide, not just for pages that show up in the serps... and if you have a lot of thin content, the site overall is ultimately not going to satisfy your visitors until you fix that content. I'd be very wary about noindexing a lot of pages as an approach to site improvement.
Regarding nofollow, I wouldn't use that at all. In the robots metatag, the nofollow attribute stops the circulation of PageRank through the page it's applied to. It effectively creates a PageRank black hole, much as nofollowed links do. In a Panda triage situation, such as you're describing, I see nothing to be gained, and lots to be lost, from using the nofollow attribute together with noindex.