The task of determining the affiliation of two sites in the Yandex has always been one of the most important in the SEO-analytics. there were a few working ways to solve it, until recently, which was based on one fundamental principle: it was necessary in a certain way form the search results, so you can make sure of the following.
If the search results are both scanned documents on affiliation sites, and at the same time there is a group of issuance of sites, these sites are not affiliated.
If, however, in the absence of groups of sites, to issue the documents present at both sites, and the presence of groups - from only one, it checks websites affiliated.
This principle imposes significant limitations on the choice of method for determining the affiliation of the two sites.
We need to narrow the sites ungrouped issue so that it included documents from both sites. And then to be able to group it to check whether it is stored presence documents from both sites, or one of them disappears. Or demonstrate in a bundle on sites SERP presence of documents with both.
At various times I have published three ways to determine the affiliation site in the Yandex, based on this principle.
The first method was the most simple and elegant. Issuance narrowed to the minimum - two documents (one at each site) by an operator url:. But this statement, as well as the majority of Document operators present official rather meager language Yandex requests (ie, site:, host: and rhost:, the only exception is the operator domain:, but later on this), there is one feature - they ungroup issuance domain. Therefore, after the restriction was necessary to built the issuance grouped by domain, to check for the affiliation. And it was done pretty simple, but interesting way - by adding to demand any document-operator (typically used by operators url: or site:) with an arbitrary value (of course, other than the operators url:, used in the base query) after the negation operator ~ ~. This bundle operator of negation and document the operator causes the opposite effect of the use of document-operator - it Grouped issue.
But, unfortunately, after the periodic castrations language queries Yandex negation operator ~~ lost status documented, though, and continued to work correctly in many cases. But the most unpleasant, that a few months ago, disappeared a wonderful effect the reverse grouping of sites inherent in its conjunction with document-operators. Minus one way, alas.
The second method was based on the use to narrow the issue of exact quotations from the pages scanned for affiliation sites. But over time, the query language Yandex logic has changed dramatically, and now he does not want to apply correctly the operator | ( "Logical OR") for the exact phrase (enclosed in the operator "quotes"). As a result, the search results for queries generated by the application of the operator | for an exact phrase, it turns out is sheer nonsense, not associated with the logic of the request. Minus two ways, alas.
The stock remains a third way , based on a single documented terms of document (sorry for the involuntary pun) operator domain:, which, unlike the rest of Document operators not ungroup the issuance of the sites. But some time ago he began to throw out trick.
The fact that a sufficiently large number of affiliated domains known to me pair (and not only couples, but also triples, quadruples, etc.) verification using this method has suddenly ceased to show their affiliation. For example, take a couple of sites, used me as an example of affiliated in the last article about this method. Now they thrive on the same page in issue formed using the operator domain:
That is, they are, from the point of view voiced by me above the principle of determining the affiliates are not considered affiliated. Moreover, there is a fourth method, which I have not posted. It was based on the use to narrow the issue once became undocumented operator inurl:, which is similar to the operator domain: not ungroup the search results. So, he too now passes on all the cases known to me pair of affiliated sites, including on the above:
However, as my practical observations and opinions of other SEO-specialists expressed in private conversations, professional, speak for the fact that in itself affiliate filter is still here. But very similar to what he mimicry, changing its nature with tight grouping of affiliated websites in the search results to something else, such as a discount (post-penalty) to the value of relevance to the pages of one of the affiliates who considered ranking algorithm less relevant, than the other.
If so, then the ability to adequately test the two sites to affiliation, to my great regret, we can put a cross. But still I want to and do not exclude the possibility of some other reasons for the observed effect, for example, you never know, a large-scale amnesty affiliates, under which were all known to me pair. So I would be very grateful if readers will send me-mail an e email@example.com their examples formerly affiliated couples sites to replenish my test sample.