What is in Google's algorithm

[AussieWebmaster]

Everyone has wondered, attempted to reverse engineer etc. but I think the better approach may be to list what ways we would produce the best possible results.

In another post a newbie asked if a link from a high PR page was better than one from a very relevant page.

I went with the relevancy and thought that it would be a good criteria for helping to improve the relevancy of search results. Given that it becomes a factor....

let's build on them here and see where it goes.... and possibly develop a proactive method of getting deeper into the algorithm.

Add ideas here

[guinanie]

Honestly I did not understand the algorithm of Google. Although off-page is the important factor but sometimes they also depends on the on-page factor especially on the content. Just like this July, there is some changes on google algorithm that most of the links from our site is omitted or deleted. Do they filter the unrelated links and duplicate content?

[Dan01]

I think the number of links trumps relevancy and rank. Other than advertisers who does the NY Times link to? Their articles are good, but are they the most informative. Wikipedia shows up in the top two nearly every time and most high ranking sites do not link to them.

Those are my thoughts on it. But now that I think about it a little more... LOL

[Misscj]

I know a fair few people doing post-docs on this kind of thing, it's pretty tricky. There are lots of ways of treating links, to identify neighbourhoods for example, rather than authority sites. Link strength is always under debate, I mean how do we really assess this accurately? There are a ridiculous number of components in the Google algorithm, it's hard to pick on just one, but interesting nevertheless. Precision and recall are the 2 main evalutaion metrics in IR, but Google is far from being simply an IR system now. It actively interacts with the content it retrives and will continue to do so, that's a good thing. It does do this to improve precision but I think there are other evalutation methods at play. Wikipedia is a pretty important site all round, everyone I know goes there for info. It would seem right that it comes up first for a fair few things. The more complex your query the more centered the results are, to a degree.

[seomax]

r google and yahoo pr countring techniques equal

[KSew]

quality content + relevant high PR links + quality content + relevant high PR links + quality content + relevant high PR links + quality content + relevant high PR links .....

top rank in Google. Yahoooooooo!

[j0nyDzine]

You know though, when you talk sites like the New York Times, you have to take into account that they probably enjoy a trusted site status to Google. I don't buy that G looks at all sites the same...
However, I still will always think of Google as "Backscratch", which is why I always have a slanted view of Google in that while as much as it touts the importance of "relevance", the link issue is of much higher consequence than many of us would like...