How does Google know?

So nobody knows?

I'm going to take a guess at this one and say that it's probably because they can tell how long visitors stay on your website. If visitors stay longer it means your website is giving them what they are looking for and therefore likely to be of high quality or value.

I have heard, their artificial intelligence is getting better at at least learning the context of an article, though sites using made for search engine, "spun" articles probably get penalized most heavily after a human review. It's said the new generation of filters are trained to use the same criteria that their human reviewers are found to employ.
I don't know the they can determine who contributed an article with out an author credit, Google maintains a cached copy of your site so they can compare the revised version with the previous version of a post. Google recognizes syndicated or duplicated content that appears on more than one location on the web, generally the site where a particular article was first found will be considered the originator and may be given higher credit than duplicate copies.
Some have suggested in the light of the higher standards the Panda and Penguin filters may enforce that you create 800 word articles on blogs instead of the conventional 450-500 word articles.
The latest rash of "Unnatural link pattern" warnings seemed to be for sites with numerous links bought from private blog networks that had been recently de-indexed in a manual identification process outside of their normal algorithms.
Google keeps their warnings vague to avoid giving details to the gamers playbooks. They probably analyze the backlink patterns of a site, irregular rates of links, like 1000 links one week and no new links the next week, or an above average percentage of links with anchor text, particularly when one particular keyword appears in too many backlinks are all patterns that may raise a red flag, they previously made concessions for the possibility of competitors sabotaging a site with bad backlinks, simply discrediting individual suspect links, Google seems to be playing it a little tougher now, it's reps have mentioned that some innocent sites may be effected.

DeePower,

I am not sure if the link below is a "bull's-eye" answer to your questions. Please check.

I just Google'd these. You better do the same to get more information.
How Does Google Know if It’s a Paid or Natural Link?
Google Says: Your Blog Has Low-Quality Content And Here Is How to Fix It


I hope these help.

Honestly Google tracks everything their databases of information paint a clear picture of what sites are link farms and what are well known and trusted sites. While you might get away with blasting your sites with backlinks for a few weeks they system will catch you and either sandbox your site or give you a nice big penalty.

In any case honestly stay with whitehat methods and be a honest person your rewards will take a little long but your income will last. An empire isn't built in a day

Google has hundreds of the smartest, probably genius-IQ-level programmers and analysts to sort through a massive amount of data (gathered by Google services and especially Google Analytics).

They can't know "for sure" if something should rank highly or not, but they can analyze all of the data they have and make a very educated guess about it.

They also employ some website analysts who manually go through search results, look at websites, and decide if those websites should rank higher or lower.

Geniuses used to be recruited by the government.
Now, they all work for Google.

Paid for: I don't think they have any way of knowing.

Guest Post: Same deal if it is a unique post I would think.

Syndicated Post: The linkback would be a giveaway, and the same content being syndicated in other places (and almost always in an article directory).

The million dollar question. I won't pretend to have enough insight to speculate.

I'd assume the algorithm has a mechanism in place for that. Again, speculation.

Don't know.

Yep Google knows all so be careful. Stay away from blog networks. Bad news!

You got some good answers there Dee. I will just add that it is not all about the algo; they use humans too.

Will

Easy: They "know" because someone programs them to know.

Basically, you have a team of programmers/aka 'engineers' who regularly scour forums such as this, who regularly look for the "tricks" that are being used to game the search engines.

They also have a fairly extensive database (aka 'profiling'), to know when something doesn't match. Again, no magic -- just something programmers put together.

As more and more updates are made, it contributes better and better knowledge overall, making it harder and harder for a single individual to game the search engine.

So with respect to paid links... Figuring if it is paid or not would go through a series of 'tests'. If it failed 'too many' tests, then it would be automatically flagged as a paid link. So something like this:

a) Does outgoing link match domain type? (I.e., golf & say poker?)
b) Does outgoing link make extensive use of optimization (i.e., "Buy Golf Clubs in Atlanta", vs "Click here to get golf clubs", vs etc, etc).
c) Does the website that the site is hosted on have a bunch of similar styled (overly optimized) links?
d) Are inbound links to website spammy? What is the domain age of inbound links? What is the 'velocity' of links? (I.e., did you just get a backlink package of 5,000 links overnight? Or did it take 2 years to get to that point?)
e) What is the bounce rate on the site? (I.e., how long to visitors 'stay' on the website)

Etc, etc.

The above list is by no means exhaustive, but that is a set of criteria that google goes through when analyzing links.

If it "fails" too many tests, then google makes an "educated" guess that it is a paid link (rather, a program does, and assigns it a number).

And that is how "google" (a program developed by thousands of workers), determines if a link is paid or not, and acts accordingly.

Same here I am continuously learning.

Haven't you heard about "The Google Toilet"?
They track everything.

If we knew how googles algorithm works, we would all have pages in the #1 spot. If you find out let me know too

Statistics Stats and Stats,

They probably have tons of thousands of recorded pattern data for different type of sites so that they can easily differentiate .

Going to assume that google has algorithms in place that can detect the different types of websites and can boil it down to what is what.

Genius's I wouldn't be surprised.

I think they use bot with algorithm rules to define and crawl over our website . I don't think those crawl involve huge amount of human due to numbers of website is scary

They have sophisticated softwares that can tell these things. They are 30 years ahead of us.