In order for an indexed link to be of use to a Search Engine, it must have a target webpage that is also indexed.

If the target webpage is already indexed at the time the link in question is being evaluated by the indexing engine, the engine has all that it needs.

If not, it can dispatch an order to its crawler to fetch a copy of the target webpage, so as to be able to complete indexing of the link. I.e., it can use the link for purposes of discovering a webpage previously unknown to it.

If it does not use "nofollow" links for such discovery, it will not command a crawler to go fetch the target webpage, with the result that it has no target in its indices for the link to point to, so that it does not index the link either.

In the cited Google tests, it was found that, if the target webpage were previously indexed, then a "nofollow" link pointing to it was indexed; otherwise, it was not, thus proving that "nofollow" links:

(a) are indexed only if the target webpage is already indexed; and,
(b) are not used for discovery purposes.

A new, fresh and empty website was set up. It is this very webpage that you are now reading, but it was then empty.

Various nofollow links were aimed from forum posts at the target webpage, with the anchor text that we shall refer to as test-keyphrase-1. (Of course, we cannot say here on this webpage what that anchor text was, or that would obscure any serps power of nofollow links!). The webpages that contained the OBL (out-bound-links) were indexed by Google, but this webpage was not. So far so good, and according to theory.

From other forums, dofollow links were aimed at this webpage with a different anchor text, that we will refer to as test-keyphrase-2. (We appreciate that some purists would prefer us to say, "not nofollow", rather than "dofollow", but the language is what it is, and "dofollow" makes it easier to follow a line of argument.)

This should have resulted in Google indexing this webpage. Some six months ago, in another test, I drove a blank webpage to the top of serps (on an obscure keyphrase, of course) by anchor text power alone.

This was when the Panda delivered a roundhouse kick to my test.

No, the webpage was not indexed. Was that because the webpage was blank? Is the Panda too clever to allow blank webpages to be indexed any longer?

I have now added some worthy and erudite text, and will continue throwing links at the page!

Two days elapse and Google has not indexed my site, despite indexing quite a few pages with OBL (both dofollow and nofollow) to my site.

Is this because the Panda is now too clever to allow Google Bombing, like the famous "Miserable failure" exploit? Will it not follow links to a website with no connection semantically to the IBL's? Perhaps we'll leave that to another test.

For the moment, I am manually submitting my site to Google.

The next day, Google has indexed my site. However, and this is A BIG HOWEVER, this site is NOT showing on serps for the anchor text on the IBL's! It does indeed appear that Google Bombing is history. But that was not the point of this test.

Four days pass, and we see useful results. This site is now indexed by Google under both the keyphrases anecdote anecdata disinformation and also misinformation commander.

The utility BackLinkWatch.com says that it can find 230 links to the site. 2 were mistakes (bplaced.net and hostcell.net) , and an analysis of the 228 useful links were:

We also note that the Google search command 'link:serpsmaster.info' shows no results at all. Google has obviously made a policy decision not to help us see our competitors' IBL's. We read at googlewebmastercentral.blogspot.com/2007/02/discover-your-links.html roughly as follows:

You asked us, and we listened: We've extended our support for querying links to your website to much beyond the link: operator that you might have used in the past. Now you can use webmaster tools to view a much larger sample of links to webpages on your site that we found on the web. Unlike the link: operator, this data is much more comprehensive and it can be classified, filtered, and downloaded. All you need to do is first verify site ownership to see this information.

The 228 links reported by BackLinkWatch.com were clearly gleaned from Bing and Yahoo, and not from Google. If  'link:serpsmaster.info' won't work for me, then their utility can hardly do better.

A Google search for 'anecdote anecdata disinformation' cheerfully says, "About 34 results", but actually shows 26 results:

Why did the BlueBackLinks.com pages beat me in serps? Well, they did have the keyphrase repeated 173 times! The pages from BackLinksForum.com that beat me had respectively 4 mentions of the keyphrase (from my signature link, because I had made that many posts on that thread), 3 instances, and 3 instances. (At the time of that test, this webpage had the keyphrase mentioned about 3 times.) Reviewing the webpages that ranked below me in serps, they all seem to have 3 or less mentions of the keyphrase. The serps results seem to be ranked in order of keyword density. So much for keyword stuffing being discounted by Google, one might wonder! One can also speculate that, as all the links were nofollow, the PR (Page Rank) of the sites owning the pages was ignored. Nonetheless, the only pages which were targetted with the anchor text did not rise to the top of serps. This seems to show that nofollow links do not help with positioning in serps. Owing to other unexpected results, this current website cannot be used to test whether nofollow links will result in a previously-unindexed webpage being indexed by Google, which may therefore be the subject of a future test.

A Google search for 'misinformation commander' says, "About 155 results", and actually shows 124 as follows:

The page from the top of serps at docstoc.com seems to contain a pdf document for download which is actually internally titled, "Disinformation Commander" in the parameters to a javascript call. Perhaps it may have backlinks to it from other websites under that name. Otherwise it might have been thought that docstoc.com's PR6 authority lets it come top of serps, until one notices that jamestown.org is also PR6 and ranks below this site.

Now let us go to Google WMT (WebMaster Tools) and look to see the links. We ask to see all links, and are shown 15. 5 are from BackLinksForum, and 10 are from WebProWorld. Can this really be true? Yahoo and Bing have created 228 backlinks from the OBL's on forum pages, and Google says it has created only 15. Some people say that Google does not show all links, and is therefore deliberately lying when WMT offers to "Download all links".

The alternative is that Google is really choosy when deciding to create backlinks. Even though Google has indexed 114 WebProWorld webpages, each with an attempted link to this website, perhaps only 15 links were actually created. Having looked at them, and this does involve some value judgement, it does seem that the links that were approved and established had a semantic fit between the originator page and the target page.

This is an important subject, so it seems to be time for another test at: I test a theory on keyphrases highly specialized, valiant endeavor, loyal follow, tricky subject, beautiful corpse