I have done extensive and controlled test on this myself.
There is no doubt in my mind that the patent, or in particular the section I referenced IS being used in whole or in part.
If you have read the patent, and understood the exact circumstances that this would apply and why that is important to
SEO, then you have enough information to do a controlled study. If not, then any test or analysis will just lead to more myth spreading.
othellotech, have you actually read the patent? Do you understand the implications and the specific circumstances that matter? If not, then you have probably been testing the wrong thing.
MSN and Inktomi are irrelevant to a Google Algo.
FWIW I know people involved in the current MSN search engine algo (poor though it is). I have been doing this for at least as long as you, started
SEO before it had a name and analysed and optimised for AltaVista before they burnt their database and gave Google the opportunity to step in. I am qualified in statistics and can perform a controlled study with the best of them. I am sure you will claim something similar - so lets drop the rank-pulling and debate the issue at hand.
There is only one debate that matters here - the Google Patent is 100% clear as to their intent.
Quote:
|
Re-ranking component 124 next compares all pairs of documents in B(y) for any pair in which IP3(first document of the pair)=IP3(second document of the pair), and removes the document of the pair from B(y) that has the lower OldScore value. (Acts 303-306). In other words, if there are multiple documents in B(y) for the same (or similar or affiliated) host IP address, only the document most relevant to the user's search query, as determined by the document's OldScore, is kept in B(y). Documents are removed from B(y) in this manner to prevent any single author of web content from having too much of an impact on the ranking value.
|
..direct quote from the patent.
Of course, it needs to be read in context - and I urge everyone to do so if they have the time.
The question: is the algo (popularly known as 'LocalRank') described in that patent being applied in whole, in part or not at all.
My tests say it is being applied at least in part - you seem to think differently.
I strongly urge you to read it, or at least the relevant section, and then test for those circumstances.
If you have already done this, then tell us - I am happy to hear details of tests which contradict my own - maybe we can then discuss it in more detail in private (or public if you wish), come up with a new test that removes any doubt and publish the results back here. I have no axe to grind other than that of 'unknowns' being presented as 'certainties' and 'possiblities' being dismissed as 'myth'.
Regardless of different shady techniques for getting top ranking, and the morals and ethics thereof, if you have a few sites that cover a similar theme, and which are crosslinked, then you really do want Google to see them as on very different IP addresses IF the algo detailed in this patent is being applied.
For those who have no wish to delve into the murky world of
SEO and algo analysis, perhaps this as at least clarified things a little. Google issued a patent that showed their intent re:same and similar IP addresses, there is debate over whether this is being used or not. Until somebody presents the results of a carefully controlled, repeatable, study there will continue to be some reasonable debate.
Until then, IMO, dismissing the theory as myth shows a lack of judgement.
[added]
in retrospect, there may be some debate over the exact meaning of:
Quote:
|
for the same (or similar or affiliated) host IP address
|
... but the overall 'drift' is fairly clear, I'd say.
Mulit c-class IP needed
RIPE won't let you..
