Hard questions for the pro's

Can google recognise transparent or small images.
Also how do I check of my content is to similar to another website, How similar is to similar.
If I have taken over a site and its getting nowhere because it is linked to by all the same sites another site my company designed is linked to with to similar content what do I do.
Also whats the password to mossads computer database anyone.
Actually scratch the last one,I dont speak hebrew.

If anyone can help me I would appreciate.

Google doesn't see images - at least, the search engine doesn't. http://images.google.com will.

Well, presumably if the content has been created by yourself from scratch, it won't be like anybody else's. There are some websites that will allow you to enter several URLs and see the similarity as a %, although I dont' know of any of hte URLs off-hand.

Get new content, come up with a new layout. Some sites, the layout is 50-80% of the actual source of the page. Just changing the layout and having the actual page body content the same (an article or whatever), can be enough.

The password to what? huh?

"Can google recognise transparent or small images. " Google will look at images that link to something...not just graphics on the page.

Axe: "Some sites, the layout is 50-80% of the actual source of the page. Just changing the layout and having the actual page body content the same (an article or whatever), can be enough."

I thought it was the other way around...otherwise if your site uses similar page construction with different content none of the interior page would get listed. The dupe filter is looking at content not html...right?

"Also how do I check if my content is to similar to another website, How similar is to similar. "

Search using phrases from your site enclosed in quotes...that will show other pages that have identical content.

to a degree, yeah, but exactly what % Google determines as "the same" I'm not sure, but if you've got a large site with a super-table-heavy layout, lots of <font> tags, and other such things that could easily be knocked out with some CSS, or simply a completely different huge chunk of tables code & formatting tags.

Just think of all the electronics store websites that all carry the exact same product reviews on their pages as the main bulk of their human-readable content.

Often-times, the only thing that's different in an example like this, is the layout and the price. The description and the reviews are all identical, and there could be 50 or more sites showing this. You may see 15 of them in the top 20 SERPS for a particular search phrase though ("make model compare prices" type search, without the quotes).

"Just think of all the electronics store websites that all carry the exact same product reviews on their pages as the main bulk of their human-readable content. "

A lot of those site got thier pages dropped two updates ago when the dupe filter first went into effect...since I think they have lighten up the filter to allow some back in....

I think you guys are giving Google too much credit here ... I doubt they can pick out 2 pages from their database and determine they are 99% the same.



I think you meant the affiliate sites got knocked out ... those would be easy to detect by all the outgoing links and of course the affiliate numbers in all those links.

Other than that the duplicated page problem generally exists on the SAME SITE ... and again that would be easily found when two or more pages end up at the same place in their ranking system.

The only other form of duplication that might be detected is when two domains point to the same IP address.

"I think you guys are giving Google too much credit here ... I doubt they can pick out 2 pages from their database and determine they are 99% the same."

Guess again...read this: http://www.webworkshop.net/seoforum/vie ... php?t=1010

Cool link ...

Guess I have to take it back. Google CAN fingerprint pages, and remove them from search results if they are too similar.

It does not matter if they are different domain names or different servers.
If the content is duped they can find it.

I am still not clear where they cut off duplication. Some seemed to report whole pages .... some seemed concerned with key phrases.

How low can they go?

how low?

will duplicated phrases/paras/pages be affected?

"will duplicated phrases/paras/pages be affected?"

phrases - -I don't think so...otherwise any site that used a common slogan on every page would get zapped. Search this string in quote on google "dedicated to equal opportunity in both housing and employment" it turns up over 2000 sites all still in the index.

paras - I have a few pages that use exact paragraphs but they are only about 20% of the total text on the page and the site is about 1000 pages....I work with a few real estate sites that use a common disclaimer paragraph on the bottom of listings pages and they have not been affected.

This is impossible to check on google since it limits the search to ten words...but if you grab the beginning of a paragraph and search it in quotes you should come up with some very similar results if dupes exist. There are a lot of spammy sites that write an optimized paragrah and then insert local or regional terms in to it to make it rank for those terms ie: Hi my web site is about "insert town" real estate we sell homes and condos in "insert town or state"...etc...this is what the algo should be targeting.

Dupe pages - nearly everytime one or the other will get dropped from the index....worst thing is you can't control which one...I have heard the higher PR gets indexed...I have also heard that the oldest page remains in the index and the new dupe gets dropped....so I can't really say...not an experiment I want to take on
Do a search on "Page Jacking" ..it has to do with people who copy ("jack") a high ranking page and then cloak off it to rank for those terms....

How low? ...I don't know...WilliamC, who started the post linked to above, comes here occassionally...WilliamC you out there? May be he can shed some light on how low is low.

Yeah, the similarity has to be over a certain perccentage for it to raise any flags with Google.

I would say the % is fairly high though. But mostly it's to help stop the spammers.

It's not difficult to setup 50 Geocities accounts, or some other free hosting company, upload the same exact page to all of them, same links on all of them, just in an attempt to boost PR, and get those search results up. So, Google seems to mostly watches out for these kinds of tactics.

Thanks for all the info.
I am building a site for a company that uses a lot of content from old site.
I was wondering if there is a way of splitting up the text using something invisible like an underscore in em tags set to bgcolor of table with external style sheets protected by robots .txt

PS Mossad is Israel's secret service.
Apparently they have the most secure computer system ever.
(mainly because no one knows where it is.)