Can PR Be Negative?

[kgun]

Originally Posted by Peter (IMC)

Finaly somebody that responds to that post,. and be nice kgun.

You should know that most Scandinavians use direct language, and I am among the worst

1. So in the worst case scenario, I could have your site excluded from the Google index by throwing a bad link pattern at your site?
2. Do you also mean that there is a negative PageRank, like the RepRank here at WPW?
3. If your answer is yes I will personally add a negative RepPoint to your answer.

To be more explicite: 2 excludes the possibility of loosing PageRank from a positive level, because of boughts links etc. That is the page's PR is reduced from n to k where k < n, because n was bought.

[crankydave]

Originally Posted by kgun

You should know that most Scandinavians use direct language, and I am among the worst

1. So in the worst case scenario, I could have your site excluded from the Google index by throwing a bad link pattern at your site?
2. Do you also mean that there is a negative PageRank, like the RepRank here at WPW?
3. If you answer yes I will personally add a negative RepPoint to your answer.

Direct or not kgun, be nice please.

As far as negative PageRank, it can be reduced by hand as a negative. Google has done this with sites that they believe sell links.

Dave

[kgun]

Originally Posted by Peter (IMC)

Link analysis is not about the pages that contain the links, but about the links them selves.

So now we have a new null hypothesis:

H0: Link analysis is not about the pages that contain the links, but about the links them selves.

that has to be tested by scientific data?

Any scientific proof with the data published, so the conclusion can be retested?

[kgun]

Originally Posted by crankydave

Direct or not kgun, be nice please.

As far as negative PageRank, it can be reduced by hand as a negative. Google has done this with sites that they believe sell links.

Dave

Yes, but isn't that what I call zero out. Can the net effect be negative?

[crankydave]

kgun...

I've asked you to please start a new thread if you want one to discuss your theory. I've asked you to not try and change the topic of this thread.

I'll not ask again.

Thank you

Dave

[Peter (IMC)]

Originally Posted by kgun

You should know that most Scandinavians use direct language, and I am among the worst

1. So in the worst case scenario, I could have your site excluded from the Google index by throwing a bad link pattern at your site?
2. Do you also mean that there is a negative PageRank, like the RepRank here at WPW?
3. If your answer is yes I will personally add a negative RepPoint to your answer.

To be more explicite: 2 excludes the possibility of loosing PageRank from a positive level, because of boughts links etc. That is the page's PR is reduced from n to k where k < n, because n was bought.

Well, I believe that a bad link profile can have a negative impact on your rankings. Being excluded from the Google index requires a lot more bad stuff. You need to make it really bad in order to get excluded from the index.

A bad link profile can not result in negative PR. A link can never transfer a negative PR value. At least it wouldn't make much sense to work like that. The impact of the link profile is not related to the PR it self, I guess it's more related to how much Google trusts a page/site.

You can give me all the negative RepPoints you want. I get more than enough positive RepPoints to not be bothered by it. And that's pretty much how Google looks at it too. In order to really have a negative impact of links (reppoints) you'd have to organize a lot of people to agree with you and they have to care enough to actually help. That's difficult especially if very few agree with you. And here too you run the risk of obtaining the opposite result, simply because you´re giving so much attention to it.

Same thing with links in Google.

[Peter (IMC)]

Originally Posted by kgun

So now we have a new null hypothesis:

H0: Link analysis is not about the pages that contain the links, but about the links them selves.

that has to be tested by scientific data?

Any scientific proof with the data published, so the conclusion can be retested?

The weather has been observed for a couple of 100 years now and many scientific tests have been done, but still the best they can do is make a prediction and hope they get close enough to reality for people to say that the prediction was correct.

SERP behaviour is very similar to Weather behaviour.

[kgun]

Originally Posted by crankydave

kgun...

I've asked you to please start a new thread if you want one to discuss your theory. I've asked you to not try and change the topic of this thread.

I'll not ask again.

Thank you

Dave

Originally Posted by galide

How to report paid links
"We’ve always tried very hard to prevent site A from hurting site B. That’s why these reports aren’t being fed directly into algorithms, and are being used as the starting point rather than being used directly. You might also want to review the policy mentioned in my 2005 post (individual links can be discounted and sellers can lose their ability to pass on PageRank/anchortext/etc., which doesn’t allow site A to hurt site B)."

That bolding is in line with my theory. To me that implies that the only effect can (in theory) be zero. That does not imply that sites can not have got an undeserved penality. If you have read my posts, you will have understood that algorithmes are not perfect.

To clear my standpoint:

1. This discussion is not about selling links, since that topic has been discussed extensively elswhere. In bound bought links can bring down your PageRank from n to k.
2. The same with IBL's that is identified as irrellevant.

To repeat: "That does not imply that sites can not have got an undeserved penality".

If that is the subject, I have nothing more to add.

[kgun]

Originally Posted by Peter (IMC)

Well, I believe that a bad link profile can have a negative impact on your rankings. Being excluded from the Google index requires a lot more bad stuff. You need to make it really bad in order to get excluded from the index.

My bolding. A priori or a posteriori?
In my view:

1. A priori the effect can only be zero or positive.
2. A posteriori (after an undeserved positive ranking is acquired) the zeroing out effect can be negative.

Originally Posted by Peter (IMC)

A bad link profile can not result in negative PR. A link can never transfer a negative PR value. At least it wouldn't make much sense to work like that. The impact of the link profile is not related to the PR it self, I guess it's more related to how much Google trusts a page/site.

Good answer IMO.

[kgun]

Originally Posted by Peter (IMC)

The weather has been observed for a couple of 100 years now and many scientific tests have been done, but still the best they can do is make a prediction and hope they get close enough to reality for people to say that the prediction was correct.

SERP behaviour is very similar to Weather behaviour.

No that is wrong. Predictions are much better today than earlier. In finance weather predicitons are big business.

Most probably the weather lives on a high dimensional chaotic strange attractor. That implies that it is not possible to predict the weather for a periode longer than the lenght of the attractor.

And if it is deterministically chaotic, the prediction error increases exponentially with time.

[kgun]

A thread cut and pasted out of context and started by crankydave.

Private copy: kgun

[crankydave]

Originally Posted by kgun

A thread cut and pasted out of context and started by crankydave.

Private copy: kgun

The topic and discussion was started by you in another thread kgun. Yes, I moved posts that were a side discussion in another thread. I made it a topic and thread of its own rather than removing the posts for an admin review.

Dave

[kgun]

Originally Posted by crankydave

...for an admin review.

Hm.

That answer was too short so I had to add this ...

[AVC]

When Google busted all the interlinked directory networks due to the same owners promoting "bidding directory link farms" they deindexed entire networks of spam directories and took their PR to ZERO since many of them were artificially moving PR up by interlinking with their buddies.

So yes, if you make the wrong moves or are interlinked with networks like this, you will certainly see your PR going negative for gaming Google and working scams.

[Peter (IMC)]

Originally Posted by kgun

No that is wrong. Predictions are much better today than earlier. In finance weather predicitons are big business.

Most probably the weather lives on a high dimensional chaotic strange attractor. That implies that it is not possible to predict the weather for a periode longer than the lenght of the attractor.

And if it is deterministically chaotic, the prediction error increases exponentially with time.

Yes,... kind of like the behaviour of the SERP's, wouldn't you say?

[kgun]

Originally Posted by Peter (IMC)

Yes,... kind of like the behaviour of the SERP's, wouldn't you say?

No, I don't think that the SERP's are the output of a chaotic sytem, even if the dimension is high.

I think linear algebra combined with statistical methods are important.

Informally I have called Google's model:

The adaptive inverted link matrix model of the web.

Advanced SQL queries on the index / register should also be an important part of the model. So the model may be fairly complex, but the kernel is in my opinion the inverted link matrix model of the web.