I am currently writing an SEO booklet for newbies. To ensure all the fact and claims made in the booklet which is titled " SEO In a Nutshell " a few SEO experiments has been conducted, old threads in UK Webmaster World forum with relevance to the page I am working on were dug out for reference. On such thread was a thread on writing the perfect meta tags. One UK WW member who seem to missed 'the perfect meta tags' thread when it was originally created many months maybe even as far back as one year ago was Miha Jepan. Coincidentally, he was writing a blog about what should be in and what should not be in a meta tags. This prompted him to create an experiment where he tries to find out the part or parts of a meta tag Google uses to rank a site. I will not go into the experiment in this post, I think its better that you read Mihai Jepan's finding directly from his contribution to the 'perfect meta tags' thread which can be found here.