Brainstorming my Panda problem

[numnum]

For your consideration, here are three (marginally) plausible reasons I've been pandalized:

* I don't have my own domain name. Instead, I'm at www.isp.net/~username. Perhaps the algo now looks a bit askance at me solely for this reason, despite the age of the site. I'm referring to the algo -- not to search-engine users, many of whom I'm sure bypass my links when they see my quaint-looking, 15-year-old URLs on the SERPs. (I never established a domain because I didn't want to risk the possible consequences of a 301 redirect. And my site is not commercial. Besides, I'm old school.)

* I provide many pairs of near-duplicate pages, down parallel branches of my site's page hierarchy. I've included a META robots noindex instruction in the source code for each and every one of those pages, in order to avoid a decision as to which should be crawled/indexed and because I don't want to facilitate entry to the site through those pages. Here's the rub: the higher-level pages linking to those duplicate pairs include some of my top entry pages for visitors coming from the SERPs. That is exactly what I wanted. But has the algo now decided that those popular entry pages are suspect because they link to pages that I've directed to not be indexed? (I'm not sure whether it's just those pages that are ranking lower in the SERPs post-Panda, and my site logs aren't providing a definitive answer.)

* My customized 404 page is an exact duplicate of my home page, which is static HTML in its entirety. (I'm on an Apache server, and my .htaccess file contains a proper 404 command.) Is it remotely possible that the algo now takes issue with my 404 page replicating my home page? This would make no sense, since proper 404-page content isn't crawled or indexed in any event, right?

Grasping at straws, or not?

[tedster]

As long as there is a proper 404 status being returned in the HTTP header, there won't be any problem from that direction.

[goodroi]

Overall it sounds like your site has usability/branding issues and that can lead to poor quality signals.

www.isp.net/~username does not look as good in serps as widgets.com. I would seriously think about getting a new domain name for your site. I would expect people are less likely to click on you in the serps and that lowers your CTR.

IMHO it is less likely that people would view you as an authority since you do not even have your own domain name so that is probably reducing the number of backlinks you could have pointing to your site and we know backlinks matter.

Having a large amount of near duplicate content can create problems. It can confuse users leading to less lower time on site and fewer page views which Google might be using as quality signals.

Also you need to hope that all of the meta robots is setup perfectly otherwise you can have headaches with Google crawling your site. Worst case Google thinks you are trying to stuff the Google index with poor quality.

Having a custom 404 that looks near identical to your homepage is a usability mistake. Let us assume that the custom 404 is properly setup and returning a 404 (which is not always the case). The problem is that your 404 needs to be very clear to users that it is an error page. Otherwise users may not realize it is an error page and think your site is broken just going in a circular link loop. If users think your site is broken they tend to leave sooner and do not return any time soon. That would not be good quality signals for Google.