Google Algorithm

About google search results

[Experienced]

Hi to all,

I am new to this forums so please don't mind if this is not the right place to ask this question.

I just find the keyword on google <keyword1 keyword2>but when the results came which site appears on top does not contain searched keyword.

First number site does not contains any search called <keyword1>.
Second site contains only <keyword2> but domain is keyword1keyword2.com, means domain preference.
third site contains neither <keyword1> nor <keyword2> but on the forth number site contains <keyword1 keyword2> term. then why this site is on the forth number.

If google prefer the domain name then second site shd come first, and fifth site shd be on second.

else second number site shd be somewhere on 15 - 16 number.

If site does not contains search string or keyword then why it is on the top.

Please tell me about these unexpected results.

thanks and regards

Exp...

[edited by: WebGuerrilla at 6:07 pm (utc) on Oct. 19, 2002]
[edit reason] removed specifics [/edit]

[chiyo]

Hello Exp.

1. Google says they use more than 100 criteria to rank web sites. If you know each one and exactly how they are applied, you will have predictable results. (depending what machine and area you connect from, what data centre you are connecting to, temporary boosts for new content and everflux issues. etc etc )

2. In this case the top site is an official government toursim site. I would expect they have a lot of incoming links, from very authoirtative sites too, themselves with good PR. I also would expect that at least some these, given the nature of the site, use "Visit India" as part of their linked text - as an example of just one of the hundred criteria that people "think" determine ranking.

(remember the weightings and number of these criteria change all the time, probably monthly. )

IMHO, trying to work out Google's algo (beyond simple fairly well established broad guidelines such as document and link structure, and incoming linking) is in 99.9% of cases a waste of time. Even if it worked last month, it may not work next month.

It is in Google's interest to have RELEVANT results for sure, but not to have PREDICTABLE results, both from an end-users and SEO view of the data. The more predictable the easier to spam (that is why Y! directory and to varying but much lesser extent all PPI influenced engines, degraded so badly) and the less reason for users to visit regularly with the same queries.

[rfgdxm1]

Good point, chiyo. While Google wants relevancy, a predictable algo would make it possible for spammers to figure them out, and that would make them less relevant. Thus, better to keep modifying the algo enough to frustrate spammers, while keeping the results suitably relevant.

[cminblues]

Even if it worked last month, it may not work next month.

Exactly, and more.. :)
I think Google is already beyond this.
I continuosly see, exactly in the same moment, different algo-element weight, per different local zones, & areas.

cminblues