Welcome to the Q&A Forum

Looking at Pagerank or DA is not a good way of determining a bad link.

It's usually the ones you created yourself or ones created by somebody you hired.  And it often resides in those dark corners of the internet like:

• 100% of the so called article directories
• 99.99% of the general link directories [there's only like 6 - 12 worth submitting to ]
• social bookmark mass submissions
• "Web2.0", Wiki, or .edu pages created for the sole purpose of passing link juice and page rank.
• spammed to death blog comment pages
• blog networks or groups of sites created for the same reason as above
• Unnatural Forum Profiles that are never used and created for the same reason as above
• Forum Posts with anchor-text rich signatures
• over-optimized press releases on websites that no human reads

Unnatural links might also include:

• poorly executed "guest posts" that border on blog network type schemes.
• widgets that are distributed with targeted anchor text
• paid links
• When you interlink sites you own in a way that's only meant to manipulate search rankings.
• Any other links with unnatural anchor text [people seldom link to you naturally with the exact keywords you want]
• Obvious Link exchanges
• and more...

The "Not Selected" isn't cumulative.   The "Ever Crawled" is though.

I have a large Wordpress content site.  It was hit by Panda on a very same day that my "not selected" multiplied by 8.  I don't think it was a coincidence, and I didn't make any large changes to the site besides the regular addition of about 10 posts per week.

I've been able to effect a downward movement on the not selected count by removing/redirecting things like "replytocom" variable URLs in the comments section;reworking print and email versions of each article, etc.   It very slow though, only reducing by an average of 100 per week.

Needless to say, I think the not selected metric means quite alot.

This is not something you want to leave up to automation.  You need human input to decide whether a link is bad.  Disavowing innocent links will do more harm than good.

So you have to go through your link profile and focus on the domains that link to you the most, then move onto the ones that link to you with non-branded anchor text.

I corrected several thousand 404s.  A further 10,000 404s were turned into 410s since they aren't coming back,  I can't redirect them, and I want google to stop crawling them to conserve crawl budget.

But none of the changes were affected the main content of the site.

I should mention that the site was hit in the May Penguin update.

I lost "megasitelinks" as well recently.  I think it may be due to changes or tweaks they are making.

Terms I had sitelinks for before were:

domain.com;
domain

Now I only have sitelinks for:

www.domain.com;
http://www.domain.com

Your site shows similar behavior.  I expect the megasitelinks will reappear soon as they make further changes.

It really does not matter.  But if given the choice, I would go with either the .net or .org.  And keep an eye on the .com in case it drops.

Or better yet, I'd try to pick a brandable domain where I can register all the TLDs to secure it.  Having all the keywords in your domain isn't worth much these days.  Once users start to associate your unique name to "super domain", the benefits will be multi-fold.

I would avoid the .info and the OPT1 for the same reason.  Marketers abuse them.

First, you cannot prevent link juice from being distributed or spent by nofollowing a link. The link juice still flows, it just doesn't go anywhere ..it just evaporates.

Second, nofollow is like saying you don't trust this page or you can't vouch for it. Why would you tell Google that you don't trust your own pages ? For that reason, I never use nofollow on internal links.

Third, I don't think you even need to worry about those pages getting juice. It's natural. It might be even a quality signal in Google's algos if they can see that you have a contact page, or privacy policy, or details like an address or telephone number on the about page. Bad sites tend to not have those things.

Well I think you should focus less on making it seem "more natural", and actually allow it to be natural.  Once you do that, there's no mixing and measuring anchor text as if you're a chemist.   No worrying about ratios to dilute and all that nonsense.

But that's just me.

In categories that have a dedicated editor, their name and link to their profile is usually shown where it currently says "volunteer to edit this category".

Yes, it's still worth it.  Got listed a month ago.  In addition to the listing, I've now got links from 50+ other sites that scrape dmoz.  I'm sure those 50 don't do much, but I did get a noticeable bump.

I submitted it and then 5 days later I contacted the editor and let him know that two sites that were listed in the category were now dead.  He removed them and added mine.

Sites that lift content from today's article directories are not sites worth getting a link from.

If you want to distribute content, find sites and building relationships with people who are willing to accept your content directly.

Years worth of free material here:
http://www.seomoz.org/blog/category/4

Most of it is all about earning your links and position in SERPs.

As for the Q1.

The majority of items in that list should be avoided.  With things like Guest Posts, Press Releases and Directory submissions, it's very easy to go overboard.   Personally,

I avoid "guest posting" on poor sites, or sites that don't have traffic, or sites which seem to only publish guest posts.

With press releases, your goal isn't to get a link from prweb.com or prnewswire.  Those links are worthless.  Yes you read right.  Your goal is to raise awareness and get people to take action on the content of your release.  That rules out all the free or cheap press release sites since humans don't read them.

There are less than a dozen general directories that are worth submitting to.  Definitely avoid anything that even smells like mass submission.   In addition to those few like Yahoo, DMOZ, maybe business.com, botw.org ..look at smaller, specialized niche directories.

Avoid article directories, period.  There were a bad idea from the start, they're even worse now.

Infographics ? I'd try not to overdo it.  Definitely avoid trying to get anchor text in there.

"Do the subdomains of a network contribute any increased value to the Root Domain?"

Short answer, yes.

Google published this today actually.

http://www.youtube.com/watch?v=_MswMYk05tk&feature=g-all-u

Pagerank just measures your link popularity.  Rankings are determined by it and hundreds of other factors.  That's why Google and good SEOs will tell you not to focus on PageRank.

You can be PR7 and not rank for anything worthwhile.

So far I've only seen a beta tester of the tool post results.  Looks promising.

http://www.davidnaylor.co.uk/google-disavow-tool-released-we-tested-it.html

http://www.davidnaylor.co.uk/disavow-tool-beta-test-recovery-laid-bare.html

You know they are bad ?  Try to get as many as possible removed.  And THEN Disavow the rest.  I stress the importance of actually getting some removed.  I don't think you'll be successful in just disavowing all those thousands of links alone.

So yeah, those links may have been helping the site to rank before ..but that's not going to happen again.  You need to get rid of them, while also attracting new links at the same time.

I've been seeing lawyers, doctors, and dentists doing it with authorship directly to their profile pages.  It stands out, so why not.  Google+ Pages don't display however.   But you should link that with rel=publisher.  [you can do both]

Also remember that blocking in robots.txt doesn't prevent Google from indexing those URLs.  If the URLs are already indexed or if they are linked to, either internally or externally they may still in appear in the index with limited snippet information.  If so, you'll need to add a noindex meta tag to those pages.

http://support.google.com/webmasters/bin/answer.py?hl=en&answer=156449

Head down to the pattern matching section.

I think

User-agent: *
Disallow: /default.aspx?z=review*

should do the trick though.

I've never seen anything like that.  I don't think they punish or think its some kind of black hat shenanigans.

More likely, loss of traffic may have to do with 301 redirect ?