Nofollow is sure shorter and a more simple way (for me) to do this rather than using onclick. Especially for the newbie blogger or those who really don't know more stuff on web development.
Chino Yray
· 2 years ago
Nofollow is sure shorter and a more simple way (for me) to do this rather than using onclick. Especially for the newbie blogger or those who really don't know more stuff on web development.
Johnny Ancich
· 2 years ago
Wow, I just saw a posting where a user said they got 500 surprise visitors to their webpage as a result of YOU stumbling upon it. PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE STUMBLE MY blog.
Nice. Let us know if this "technique" works for you.
Johnny Ancich
· 2 years ago
Wow, I just saw a posting where a user said they got 500 surprise visitors to their webpage as a result of YOU stumbling upon it. PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE STUMBLE MY blog.
Nice. Let us know if this "technique" works for you.
robwatts
· 2 years ago
I think the noscript tag would be a nice addition in making sure a vanilla html link was there for reference purposes or for those with script disabled. (We all know the power of the noscript tag from the recent SEJ outing of that payday loan hitcounter site).
For me, the best implentation would be some kind of 301 implementation that would be seen by all.
Something like click.php?url=url which when followed returned
header('HTTP/1.1 301 Moved Permanently'); header('Location: '.$url);
Although such an approach would need a little additional work to check if it was a verified url at the backend.
Provided its not a 302, then personally, I dont see any problem.
In terms of their existing set up, my gut feeling is that the onclick event is irrelevant and that PR is passed simply because its contained within a straight a href container. Its not produced via document.write (which would be an issue as the bots wouldnt see it)
Anyhow, this can be tested can it not? Just gotta stick up a secret page or domain and point a page from BC to the page that nobody knows about, and voila, end of hypothesis. :D
AndyBeard
· 2 years ago
A test would only prove some, just like the example screenshot is proof that some do pass juice.
I also found some blogs whilst testing that didn't have any backlinks being reported at all, from anywhere!
A redirect isn't obvious for users, and is ugly just like the Alexa redirects - there are a few other options but the current one in use is the cleanest.
If you use other methods, you are hiding the fact you are redirecting.
There really should be a published best practice or clearer indication.
robwatts
· 2 years ago
I think the noscript tag would be a nice addition in making sure a vanilla html link was there for reference purposes or for those with script disabled. (We all know the power of the noscript tag from the recent SEJ outing of that payday loan hitcounter site).
For me, the best implentation would be some kind of 301 implementation that would be seen by all.
Something like click.php?url=url which when followed returned
header('HTTP/1.1 301 Moved Permanently'); header('Location: '.$url);
Although such an approach would need a little additional work to check if it was a verified url at the backend.
Provided its not a 302, then personally, I dont see any problem.
In terms of their existing set up, my gut feeling is that the onclick event is irrelevant and that PR is passed simply because its contained within a straight a href container. Its not produced via document.write (which would be an issue as the bots wouldnt see it)
Anyhow, this can be tested can it not? Just gotta stick up a secret page or domain and point a page from BC to the page that nobody knows about, and voila, end of hypothesis. :D
AndyBeard
· 2 years ago
A test would only prove some, just like the example screenshot is proof that some do pass juice.
I also found some blogs whilst testing that didn't have any backlinks being reported at all, from anywhere!
A redirect isn't obvious for users, and is ugly just like the Alexa redirects - there are a few other options but the current one in use is the cleanest.
If you use other methods, you are hiding the fact you are redirecting.
There really should be a published best practice or clearer indication.
Rose
· 2 years ago
I recommend that you read this.
It may look like a clear link, but is not pass PR.
It is a discussion where Matt Cutts and Brian White both from Google mention that OnClick Java does not pass PageRank.
If you look closely you will see that it is very similar to what BC uses.
PS: Andy I don't like you. Everyone else may think that you are all that and a bag of chips, but I think you are nothing more than a BC Guppy with a brown nose.
AndyBeard
· 2 years ago
Rose
In my article there is a whole section extensively quoting from that page on SEW, and also from other references.
There is a good chance that if I have extensively quoted from it, then I also read it.
But then I am also showing proof that as far as is possible with the tools Google is providing currently, that links with onclick can be counted by Google.
I didn't make up that SERP, it shows a link coming from a blog profile page on Blogcatalog.
I don't know why Google decide to show that link, any more than I don't know why Google choose not to show me links from so many other sites.
The Washinton Post situation was totally different, they were selling off-topic links which was obvious with a human inspection.
The links from Blogcatalog are provided free, and lead from a blog profile that is within a category, and point to the related blog.
Blog Strokes
· 2 years ago
The problem here isn't that Andy didn't see this thread, but that you read into it what you hoped to find there. Neither Brian nor Matt come out and say what you claim they do, and interestingly, when Danny questions the point directly, neither choses to respond.
And that develops as a pattern. Every time any one asks a question that could only be answered by resting this issue, Google's representatives remain silent.
The simple fact is that this does not close your case as you hope, and even the least rational reflection will bring you to that understanding. Does it prove you are wrong? No. It doesn't do that either, but a lot of other information leads to the invertible conclusion that either your view is incorrect to some extent or that Google is becoming fractured and schizoid.
Rose
· 2 years ago
I recommend that you read this.
It may look like a clear link, but is not pass PR.
It is a discussion where Matt Cutts and Brian White both from Google mention that OnClick Java does not pass PageRank.
If you look closely you will see that it is very similar to what BC uses.
PS: Andy I don't like you. Everyone else may think that you are all that and a bag of chips, but I think you are nothing more than a BC Guppy with a brown nose.
AndyBeard
· 2 years ago
Rose
In my article there is a whole section extensively quoting from that page on SEW, and also from other references.
There is a good chance that if I have extensively quoted from it, then I also read it.
But then I am also showing proof that as far as is possible with the tools Google is providing currently, that links with onclick can be counted by Google.
I didn't make up that SERP, it shows a link coming from a blog profile page on Blogcatalog.
I don't know why Google decide to show that link, any more than I don't know why Google choose not to show me links from so many other sites.
The Washinton Post situation was totally different, they were selling off-topic links which was obvious with a human inspection.
The links from Blogcatalog are provided free, and lead from a blog profile that is within a category, and point to the related blog.
Blog Strokes
· 2 years ago
The problem here isn't that Andy didn't see this thread, but that you read into it what you hoped to find there. Neither Brian nor Matt come out and say what you claim they do, and interestingly, when Danny questions the point directly, neither choses to respond.
And that develops as a pattern. Every time any one asks a question that could only be answered by resting this issue, Google's representatives remain silent.
The simple fact is that this does not close your case as you hope, and even the least rational reflection will bring you to that understanding. Does it prove you are wrong? No. It doesn't do that either, but a lot of other information leads to the invertible conclusion that either your view is incorrect to some extent or that Google is becoming fractured and schizoid.
Rob
· 2 years ago
The link command being all squiffy doesn't help matters either, add a 'just because it shows as a backlink doesn't necesarilly mean that it passes pagerank' theory into the mix and, well, suffice to say we could all end up in circuitous go nowhere very fast cul-de-sacs.
I get why BC want to track such things sure, info is power etc. Maybe they aren't interested in users who surf Javascript disabled who knows, a detractor might say that this was a way of restricting link juice but...heck that would be so pre 2007 it just wouldn't be funny. PR leakage I'm sure is the least of a site like BC's worries. We all can all name a good few domains that link out like crazy yet rank for many things. Where is the PR leakage theory there I ask?
Im sure too that those search dudes play around in the space and plant all manner of anti seo landmines all over the shop simply to cause confusion and uncertainty. They'd be silly not to. Why they still give us guys tools to deconstruct them is very odd and yes, makes you think twice about what we are being shown anyways!
The onclick example in that SEW thread is slightly different to this too, as in the BC onclick event example, it calls a function and not a full url. The example cited at SEW cited too separate full urls. The one in the href and the one in the onclick call, a marginal difference but a difference nontheless.
I'm surpised that MC and BW even commented in that thread as really, there shouldn't be any problem with a company tracking what their users do. Where was the harm in washington post doing what they did? Who got killed even. Why should a little bit of tracking effect the ultimate destination? As long as the user ends up at what they see in the status bar then it really shouldn't matter.
That said, as the onclick event would only work for users with JS enabled, then perhaps it would be beter for them to simply document.write the output of the onclick aspect of the code, that way, the bot wouldn't see it anyway, so it becomes a moot point.
Ultimately the intent is to give users a link and track user behaviour. The SE's shouldn't get all bent out of shape over that.
Dane Morgan
· 2 years ago
The harm, and the thing that Brian seems to have been hitting on was that these links were in a paid "blogroll". Even if one believes that onclick tracking automatically means no passing, and even if there were no onclick tracking in these links Google would have sought to blog PR passing from them.
It's okay to sell links, but only for those on Google's short list, and it would seem that, at least a year ago, the Post wasn't on that list.
robwatts
· 2 years ago
Absolutely Dane, I too alluded to this in my Sphinn comment.
Sometimes those guys have a habit of creating FUD. It isn't the 1st time something will be left open to interpretation and wont be the last either!
Rob
· 2 years ago
The link command being all squiffy doesn't help matters either, add a 'just because it shows as a backlink doesn't necesarilly mean that it passes pagerank' theory into the mix and, well, suffice to say we could all end up in circuitous go nowhere very fast cul-de-sacs.
I get why BC want to track such things sure, info is power etc. Maybe they aren't interested in users who surf Javascript disabled who knows, a detractor might say that this was a way of restricting link juice but...heck that would be so pre 2007 it just wouldn't be funny. PR leakage I'm sure is the least of a site like BC's worries. We all can all name a good few domains that link out like crazy yet rank for many things. Where is the PR leakage theory there I ask?
Im sure too that those search dudes play around in the space and plant all manner of anti seo landmines all over the shop simply to cause confusion and uncertainty. They'd be silly not to. Why they still give us guys tools to deconstruct them is very odd and yes, makes you think twice about what we are being shown anyways!
The onclick example in that SEW thread is slightly different to this too, as in the BC onclick event example, it calls a function and not a full url. The example cited at SEW cited too separate full urls. The one in the href and the one in the onclick call, a marginal difference but a difference nontheless.
I'm surpised that MC and BW even commented in that thread as really, there shouldn't be any problem with a company tracking what their users do. Where was the harm in washington post doing what they did? Who got killed even. Why should a little bit of tracking effect the ultimate destination? As long as the user ends up at what they see in the status bar then it really shouldn't matter.
That said, as the onclick event would only work for users with JS enabled, then perhaps it would be beter for them to simply document.write the output of the onclick aspect of the code, that way, the bot wouldn't see it anyway, so it becomes a moot point.
Ultimately the intent is to give users a link and track user behaviour. The SE's shouldn't get all bent out of shape over that.
Dane Morgan
· 2 years ago
The harm, and the thing that Brian seems to have been hitting on was that these links were in a paid "blogroll". Even if one believes that onclick tracking automatically means no passing, and even if there were no onclick tracking in these links Google would have sought to blog PR passing from them.
It's okay to sell links, but only for those on Google's short list, and it would seem that, at least a year ago, the Post wasn't on that list.
robwatts
· 2 years ago
Absolutely Dane, I too alluded to this in my Sphinn comment.
Sometimes those guys have a habit of creating FUD. It isn't the 1st time something will be left open to interpretation and wont be the last either!
Mat
· 2 years ago
Hi Andy
Great post. I find the whole area so confusing
It's also annoying the way Google so consistently choose not to answer questions clearly! They know what issues are being had, why don't they clarify this for people more often. The amount of times in the past 2 weeks I have seen people asking questions and getting vague answers is high. Too high.
Overall they're a great company but the great stone wall of silence makes it difficult for the lay man to get by.
Mat
· 2 years ago
Hi Andy
Great post. I find the whole area so confusing
It's also annoying the way Google so consistently choose not to answer questions clearly! They know what issues are being had, why don't they clarify this for people more often. The amount of times in the past 2 weeks I have seen people asking questions and getting vague answers is high. Too high.
Overall they're a great company but the great stone wall of silence makes it difficult for the lay man to get by.
Dane Morgan
· 2 years ago
"Do not assume just because you see a backlink that it’s carrying weight."
When he says this he is speaking about the webmaster tool, not the link: operator. When asked if this same advice applies to the link: operator he does not reply.
So while it is *possible* that this could relate to the link: operator, such a conclusion can not be drawn from this data.
Since the webmaster tool includes nofollowed links where the link: operator does not we do know there are differences and it is not unreasonable to consider that this might be one of them.
Halfdeck
· 2 years ago
Here's a confirmation from Matt Cutts, Dane:
"That's just not true. Or rather, if most SEOs take it that way, then most SEOs are wrong. :) Halfdeck actually bothered to hunt down where I've said it before. I'm happy just to re-state: if you assume links for a link: query automatically carry weight, then you've made a faulty assumption."
Dane Morgan
· 2 years ago
Can you link to that?
Dane Morgan
· 2 years ago
Actually, I just read it at the sphinn discussion.
Am I entirely alone, or does the fact that he took the time to make that post and STILL did not address the topic of the discussion say something?
Dane Morgan
· 2 years ago
"Do not assume just because you see a backlink that it’s carrying weight."
When he says this he is speaking about the webmaster tool, not the link: operator. When asked if this same advice applies to the link: operator he does not reply.
So while it is *possible* that this could relate to the link: operator, such a conclusion can not be drawn from this data.
Since the webmaster tool includes nofollowed links where the link: operator does not we do know there are differences and it is not unreasonable to consider that this might be one of them.
Halfdeck
· 2 years ago
Here's a confirmation from Matt Cutts, Dane:
"That's just not true. Or rather, if most SEOs take it that way, then most SEOs are wrong. :) Halfdeck actually bothered to hunt down where I've said it before. I'm happy just to re-state: if you assume links for a link: query automatically carry weight, then you've made a faulty assumption."
Dane Morgan
· 2 years ago
Can you link to that?
Dane Morgan
· 2 years ago
Actually, I just read it at the sphinn discussion.
Am I entirely alone, or does the fact that he took the time to make that post and STILL did not address the topic of the discussion say something?
John Mortgage Broker
· 2 years ago
Interesting stuff; I found your blog a while back from John Chow I think and I just thought of it today and thought I'd stop by. I've honestly never used onclick in my entire life :-) I do follow on my blog, however. I'm not huge on code, I deal with it because I have to I guess...I think I just need an assistant.
Halfdeck
· 2 years ago
I have to say Matt's statement in that thread is open to interpretation. I don't have much faith in what the link: command returns either though. Still, the entire debate brings up an interesting question - easily provable by running a test:
1. Install a sitewide onclick link pointing at a new page. 2. Track crawler activity and indexing.
Sitewide links will direct enough PageRank to a new page to get it indexed in a few days - assuming onclick doesn't block link juice / bot crawl.
Anyone want to bet money on what happens? :)
AndyBeard
· 2 years ago
Halfdeck, think about it, a site wide link from my blog probably carries as much juice as a single link from John Battelle (PR8)
I found blogs on Blogcatalog that had been established for a while, at least a few weeks, and Google's link commend wasn't showing up any links at all.
This is getting into Vanessa Fox "SEO is Art" territory.
I linked in the article to test results stating that Google does crawl the links, so it is not the same as "nofollow"
JavaScript onclicks do not pass PageRank, or the PageRank they do pass is so small that it is not seen. It was verified that Google does follow the links and index the pages. However, from a pure PageRank perspective, no “juice†is being passed.
I know Rose read that too, but without a lot more details, it is impossible to say whether those tests were conclusive.
I am intending to do some additional testing to see if I can Gogglebomb a page to rank for something weird, but that won't be conclusive because the onclick might modify the juice given.
The 2 best terms to describe SEO are YMMV and FWIW
Halfdeck
· 2 years ago
"a site wide link from my blog probably carries as much juice as a single link from John Battelle (PR8)"
Yeah, it should be more than enough to get a page indexed.
"It was verified that Google does follow the links and index the pages. However, from a pure PageRank perspective, no “juice†is being passed."
Ok, now this makes me wonder: does Google index URLs then run PageRank calculations, dropping URLs with very low PageRank after indexing? Or does Google "guess" the PageRanks of newly discovered URLs and prevent indexing of URLs that probably have very low PageRank?
Guessing the PageRanks of every newly discovered URL would be more involved, I imagine, than just dropping low PageRank URLs from the index during the PageRank iteration phase.
So, his statement doesn't make sense if I assume that Google guestimates the PageRanks of a newly found URL and index or not index depending on that PageRank because:
If no PageRank is being passed, a URL shouldn't be indexed. If very little PageRank is being passed (but enough to force indexing), then the URL should be in the supplemental results.
Therefore, if he set up his test correctly and his page got indexed, then his claim that no juice is being passed is false - assuming Googlebot is proactive about not crawling or indexing URLs with very low PageRank.
If instead I assume Google will crawl first and drop URLs later, then I'd ask: "did the page stay indexed? Or was it dropped from the index a month or two later?"
Also keep in mind a page that Google includes in its main index will never have zero PageRank; the minimum PageRank of a page is (1-d) or .15 if we use the original PageRank formula with d = .85. So "or the PageRank they do pass is so small that it is not seen." isn't convincing. Was the PageRank passed or was it "created" by the page itself?
Finally, his post was written back in Aug 2006 when even on WMW the line between indexing/crawling and PageRank was somewhat unclear. He also doesn't say how he measured PageRank (my guess is the toolbar).
We all love to hear a simple "yes/no" answer to things like this, but Google is a program, and when we ask questions about a program with thousands of lines of code, the answers can sometimes be pretty complicated.
AndyBeard
· 2 years ago
Halfdeck, there has been a response from Matt suggesting that your link to his article about links in the webmaster console was correct. http://sphinn.com/story/5310#c7990
That also doesn't make sense as I have responded.
Webmaster console is generally a lot more up to date Webmaster console includes nofollow links
No answer on the onclick though
John Mortgage Broker
· 2 years ago
Interesting stuff; I found your blog a while back from John Chow I think and I just thought of it today and thought I'd stop by. I've honestly never used onclick in my entire life :-) I do follow on my blog, however. I'm not huge on code, I deal with it because I have to I guess...I think I just need an assistant.
Halfdeck
· 2 years ago
I have to say Matt's statement in that thread is open to interpretation. I don't have much faith in what the link: command returns either though. Still, the entire debate brings up an interesting question - easily provable by running a test:
1. Install a sitewide onclick link pointing at a new page. 2. Track crawler activity and indexing.
Sitewide links will direct enough PageRank to a new page to get it indexed in a few days - assuming onclick doesn't block link juice / bot crawl.
Anyone want to bet money on what happens? :)
AndyBeard
· 2 years ago
Halfdeck, think about it, a site wide link from my blog probably carries as much juice as a single link from John Battelle (PR8)
I found blogs on Blogcatalog that had been established for a while, at least a few weeks, and Google's link commend wasn't showing up any links at all.
This is getting into Vanessa Fox "SEO is Art" territory.
I linked in the article to test results stating that Google does crawl the links, so it is not the same as "nofollow"
JavaScript onclicks do not pass PageRank, or the PageRank they do pass is so small that it is not seen. It was verified that Google does follow the links and index the pages. However, from a pure PageRank perspective, no “juice†is being passed.
I know Rose read that too, but without a lot more details, it is impossible to say whether those tests were conclusive.
I am intending to do some additional testing to see if I can Gogglebomb a page to rank for something weird, but that won't be conclusive because the onclick might modify the juice given.
The 2 best terms to describe SEO are YMMV and FWIW
Halfdeck
· 2 years ago
"a site wide link from my blog probably carries as much juice as a single link from John Battelle (PR8)"
Yeah, it should be more than enough to get a page indexed.
"It was verified that Google does follow the links and index the pages. However, from a pure PageRank perspective, no “juice†is being passed."
Ok, now this makes me wonder: does Google index URLs then run PageRank calculations, dropping URLs with very low PageRank after indexing? Or does Google "guess" the PageRanks of newly discovered URLs and prevent indexing of URLs that probably have very low PageRank?
Guessing the PageRanks of every newly discovered URL would be more involved, I imagine, than just dropping low PageRank URLs from the index during the PageRank iteration phase.
So, his statement doesn't make sense if I assume that Google guestimates the PageRanks of a newly found URL and index or not index depending on that PageRank because:
If no PageRank is being passed, a URL shouldn't be indexed. If very little PageRank is being passed (but enough to force indexing), then the URL should be in the supplemental results.
Therefore, if he set up his test correctly and his page got indexed, then his claim that no juice is being passed is false - assuming Googlebot is proactive about not crawling or indexing URLs with very low PageRank.
If instead I assume Google will crawl first and drop URLs later, then I'd ask: "did the page stay indexed? Or was it dropped from the index a month or two later?"
Also keep in mind a page that Google includes in its main index will never have zero PageRank; the minimum PageRank of a page is (1-d) or .15 if we use the original PageRank formula with d = .85. So "or the PageRank they do pass is so small that it is not seen." isn't convincing. Was the PageRank passed or was it "created" by the page itself?
Finally, his post was written back in Aug 2006 when even on WMW the line between indexing/crawling and PageRank was somewhat unclear. He also doesn't say how he measured PageRank (my guess is the toolbar).
We all love to hear a simple "yes/no" answer to things like this, but Google is a program, and when we ask questions about a program with thousands of lines of code, the answers can sometimes be pretty complicated.
AndyBeard
· 2 years ago
Halfdeck, there has been a response from Matt suggesting that your link to his article about links in the webmaster console was correct. http://sphinn.com/story/5310#c7990
That also doesn't make sense as I have responded.
Webmaster console is generally a lot more up to date Webmaster console includes nofollow links
No answer on the onclick though
AndyBeard
· 2 years ago
Half, I think almost all TLA links are in that format, or affiliate links ;)
It is not just SEJ, and there is no proof that money changed hands
Maybe Pat Gavin buys more drinks at SES than Matt Cutts
AndyBeard
· 2 years ago
Half, I think almost all TLA links are in that format, or affiliate links ;)
It is not just SEJ, and there is no proof that money changed hands
Maybe Pat Gavin buys more drinks at SES than Matt Cutts
AndyBeard
· 2 years ago
I certainly didn't ask Dane to write anything, he must have decided on his own to use the power of the masses to bring in some more data.
I didn't do that for my original research because I wouldn't want to be a scaremonger. There would be too many people on Blogcatalog who wouldn't understand the reason why they were doing the search.
Dane's further research if anything corroborates this post. A large number of people are finding links.
I can only see 2.5% of my own links using the link command, many from extremely notable blogs.
Dane Morgan
· 2 years ago
As evidenced by the number of times I've had to assure people that there is nothing "wrong" if their search does not return a result.
And It is my research, not Andy's. My methods are mine.
AndyBeard
· 2 years ago
I certainly didn't ask Dane to write anything, he must have decided on his own to use the power of the masses to bring in some more data.
I didn't do that for my original research because I wouldn't want to be a scaremonger. There would be too many people on Blogcatalog who wouldn't understand the reason why they were doing the search.
Dane's further research if anything corroborates this post. A large number of people are finding links.
I can only see 2.5% of my own links using the link command, many from extremely notable blogs.
Dane Morgan
· 2 years ago
As evidenced by the number of times I've had to assure people that there is nothing "wrong" if their search does not return a result.
And It is my research, not Andy's. My methods are mine.
Halfdeck
· 2 years ago
"and there is no proof that money changed hands"
Nope. Still, it would be too easy if I can check the effectiveness of a paid link by running a link: command.
Do those paid links count? According to the link: command, they do ... if you assume that all URLs returned by the link: command pass juice.
I'm not saying they don't pass juice. My point is if link: command gives away that information, its revealing way too much info; I don't believe Google is dumb enough to let SEOs know which of their backlinks are actually counting.
Halfdeck
· 2 years ago
"and there is no proof that money changed hands"
Nope. Still, it would be too easy if I can check the effectiveness of a paid link by running a link: command.
Do those paid links count? According to the link: command, they do ... if you assume that all URLs returned by the link: command pass juice.
I'm not saying they don't pass juice. My point is if link: command gives away that information, its revealing way too much info; I don't believe Google is dumb enough to let SEOs know which of their backlinks are actually counting.
AndyBeard
· 2 years ago
I do agree with that to a certain extent.
The claim of the initial post was
"BlogCatalog - Blog Catalog Doesn't Pass Pagerank"
You would hope that of my 28600 (though Yahoo has reported up to 40000 in the past) links that Google only report as 863 and then only let me see 189, that there would be some value to them.
Maybe it shows the links that would have been counted if they hadn't been detected as piad (FUD)
And it could well be that links using onclick are being counted or lumped in with paid links, but there is no proof, and the statements that Google have made are as always too ambiguous to act as proof.
AndyBeard
· 2 years ago
I do agree with that to a certain extent.
The claim of the initial post was
"BlogCatalog - Blog Catalog Doesn't Pass Pagerank"
You would hope that of my 28600 (though Yahoo has reported up to 40000 in the past) links that Google only report as 863 and then only let me see 189, that there would be some value to them.
Maybe it shows the links that would have been counted if they hadn't been detected as piad (FUD)
And it could well be that links using onclick are being counted or lumped in with paid links, but there is no proof, and the statements that Google have made are as always too ambiguous to act as proof.
Richard Hearne
· 2 years ago
Dan Crow mentioned that Google were working on psuedo-pagerank.
That should cause further brain swelling for you Half ;-)
Excuse my dumbness here, but if Andy you are saying that an anchor with an onclick event doesn't pass pagerank I'm ever so slightly confused:
That being the case then Urchin events added to a link would inhibit the flow of PR. I've never tested this, but I'd be mighty shocked if Google advised people to use their tracking software without mentioning that doing so may actually affect the juice of that link. Maybe I haven't read enough of the posts and comments, but this couldn't be a blanket change in behaviour applied to ANY anchor containing an onclick attribute?
Rgds Richard
AndyBeard
· 2 years ago
Richard whichever code you wrote with a link to this side didn't work.
You need to use square brackets for the code here
The links are interesting reading if you haven't seen them before, and you should probably read the 2 threads on Sphinn as well, both are on the front page currently.
One of the sites linked though to was a useful tutorial on Urchin.
I am saying it should still pass pagerank, Rose is saying it isn't based purely on 2 threads.
Matt Cutts is also now saying the link: commend doesn't effectively mean anything at all as far as which links pass value.
Maybe every site should just use nofollow on every single link, start a national nofollow day
Richard Hearne
· 2 years ago
Dan Crow mentioned that Google were working on psuedo-pagerank.
That should cause further brain swelling for you Half ;-)
Excuse my dumbness here, but if Andy you are saying that an anchor with an onclick event doesn't pass pagerank I'm ever so slightly confused:
That being the case then Urchin events added to a link would inhibit the flow of PR. I've never tested this, but I'd be mighty shocked if Google advised people to use their tracking software without mentioning that doing so may actually affect the juice of that link. Maybe I haven't read enough of the posts and comments, but this couldn't be a blanket change in behaviour applied to ANY anchor containing an onclick attribute?
Rgds Richard
AndyBeard
· 2 years ago
Richard whichever code you wrote with a link to this side didn't work.
You need to use square brackets for the code here
The links are interesting reading if you haven't seen them before, and you should probably read the 2 threads on Sphinn as well, both are on the front page currently.
One of the sites linked though to was a useful tutorial on Urchin.
I am saying it should still pass pagerank, Rose is saying it isn't based purely on 2 threads.
Matt Cutts is also now saying the link: commend doesn't effectively mean anything at all as far as which links pass value.
Maybe every site should just use nofollow on every single link, start a national nofollow day
Richard Hearne
· 2 years ago
Sorry - it was simply the code you published at the head of your post.
We had a guy over on Google Groups recently proposing something similar - nofollowing every link on your site in case Google penalises you... *sigh*
So just that I'm clear about what you're saying Andy - my reading of the above gives the impression that your testing of an anchor with an onclick attribute set passes little or no juice?
I see Half chiming in with some interesting thoughts on whether a page would be indexed without pagerank (and he's one guy who knows his PR).
AndyBeard
· 2 years ago
Here is what I wrote in one of the threads on Sphinn
Whilst I might be able to set up an effective example to try to prove whether a weighting is given for anchor text with an onclick component, using some mumbo jumbo phrase such as your recent meta keywords experiment, that still wouldn't prove anything because Google might decide to apply a penalty for the onclick at a later date.
Whatever test you set up for this, it is never going to be conclusive.
I have always maintained that Google could possibly be doing some kind of 2 pass system for link allocation
1. Decide which links they should use 2. Decide which weighting to give them
Sitewide links in the sidebar people suggest have less value than links in content, and the same is possibly true of comment links if detected.
At what stage in the calculations is that being discounted? How does the juice deducted get allocated?
If all you have is sidebar links on your front page, then it doesn't matter what percentage they are allocated, it will all be in proportion 0.15:0.15 is the same as 1:1
Matt is effectively saying that just because links show up, that doesn't mean they have value, thus even the links I have found to sites from Blogcatalog could be totally meaningless.
That effectively makes the link: command about as useful as "Are you feeling lucky"
Richard Hearne
· 2 years ago
Sorry - it was simply the code you published at the head of your post.
We had a guy over on Google Groups recently proposing something similar - nofollowing every link on your site in case Google penalises you... *sigh*
So just that I'm clear about what you're saying Andy - my reading of the above gives the impression that your testing of an anchor with an onclick attribute set passes little or no juice?
I see Half chiming in with some interesting thoughts on whether a page would be indexed without pagerank (and he's one guy who knows his PR).
AndyBeard
· 2 years ago
Here is what I wrote in one of the threads on Sphinn
Whilst I might be able to set up an effective example to try to prove whether a weighting is given for anchor text with an onclick component, using some mumbo jumbo phrase such as your recent meta keywords experiment, that still wouldn't prove anything because Google might decide to apply a penalty for the onclick at a later date.
Whatever test you set up for this, it is never going to be conclusive.
I have always maintained that Google could possibly be doing some kind of 2 pass system for link allocation
1. Decide which links they should use 2. Decide which weighting to give them
Sitewide links in the sidebar people suggest have less value than links in content, and the same is possibly true of comment links if detected.
At what stage in the calculations is that being discounted? How does the juice deducted get allocated?
If all you have is sidebar links on your front page, then it doesn't matter what percentage they are allocated, it will all be in proportion 0.15:0.15 is the same as 1:1
Matt is effectively saying that just because links show up, that doesn't mean they have value, thus even the links I have found to sites from Blogcatalog could be totally meaningless.
That effectively makes the link: command about as useful as "Are you feeling lucky"
Halfdeck
· 2 years ago
Rose, you don't win any brownie points for bringing up off-topic issues like MyBlogLog. Bottom line: we have no conclusive evidence that onclick links do or do not pass juice. You're jumping to the conclusion that it doesn't. Making personal attacks only weakens your position.
AndyBeard
· 2 years ago
Halfdeck, that was actually Rose repeating one of my comments on Sphinn, which is a throwback to a previous attemp by Rose to discredit Blogcatalog, around the same time she was banned from the service.
It is actually relevant to this conversation, because Rose seems to think the only directory which might have bad signals is Blogcatalog, whereas I would personally be more concerned with the links on MBL if I was worried about link credit.
Halfdeck
· 2 years ago
Rose, you don't win any brownie points for bringing up off-topic issues like MyBlogLog. Bottom line: we have no conclusive evidence that onclick links do or do not pass juice. You're jumping to the conclusion that it doesn't. Making personal attacks only weakens your position.
AndyBeard
· 2 years ago
Halfdeck, that was actually Rose repeating one of my comments on Sphinn, which is a throwback to a previous attemp by Rose to discredit Blogcatalog, around the same time she was banned from the service.
It is actually relevant to this conversation, because Rose seems to think the only directory which might have bad signals is Blogcatalog, whereas I would personally be more concerned with the links on MBL if I was worried about link credit.
Halfdeck
· 2 years ago
Oh ok, so its you who brought it up... nevermind :)
Halfdeck
· 2 years ago
Oh ok, so its you who brought it up... nevermind :)
Dane Morgan
· 2 years ago
Have you received any emails from your own directory members about the redirect links you use that pass no PR or link pop?
Dane Morgan
· 2 years ago
Have you received any emails from your own directory members about the redirect links you use that pass no PR or link pop?
Dane Morgan
· 2 years ago
Also, while the links do not prove that pagerank is being passed, this is not the same as the links do not prove anything.
Dane Morgan
· 2 years ago
Also, while the links do not prove that pagerank is being passed, this is not the same as the links do not prove anything.
Rose
· 2 years ago
Halfdeck , you said I did not win any brownie points for bringing up off-topic issues like MyBlogLog. Well as Andy said it was him that brought it up and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?
Got to love the SEO circle... I can't call Andy here a Guppy yet users can call me a dumb a**.
I can't mention BC but Andy can.
Andy if you think this all has to do with my being removed for pointing out their lack of terms of service and that they were a clone- you better think again. The funny thing is NOW Daniel is making changes.
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
You just don't get it and you never will Andy.
Here let me dumb it down.
It was Mr Doubts who first found this June 7, 2007.
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments to wake them up.
Credit goes to Mr. Doubts though.
Rose
· 2 years ago
Halfdeck , you said I did not win any brownie points for bringing up off-topic issues like MyBlogLog. Well as Andy said it was him that brought it up and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?
Got to love the SEO circle... I can't call Andy here a Guppy yet users can call me a dumb a**.
I can't mention BC but Andy can.
Andy if you think this all has to do with my being removed for pointing out their lack of terms of service and that they were a clone- you better think again. The funny thing is NOW Daniel is making changes.
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
You just don't get it and you never will Andy.
Here let me dumb it down.
It was Mr Doubts who first found this June 7, 2007.
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
Credit goes to Mr. Doubts for first finding this though.
Andy this was not personal. It had already been pointed out back in June and I was only stating that it should be fixed.
Halfdeck why it ok for Andy here to bring it up yet not me? I was only quoting what Andy said to Shawn. That certainly did not WIN brownie points in the debate.
Rose
· 2 years ago
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
He was replying to Sebastian.
It was Mr Doubts who first found this June 7, 2007.
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
Credit goes to Mr. Doubts for first finding this though.
Andy this was not personal. It had already been pointed out back in June and I was only stating that it should be fixed.
Halfdeck why it ok for Andy here to bring it up yet not me? I was only quoting what Andy said to Shawn. That certainly did not WIN brownie points in the debate.
Rose
· 2 years ago
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
He was replying to Sebastian.
Rose
· 2 years ago
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
He was replying to Sebastian.
RoseDesRochers
· 2 years ago
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
RoseDesRochers
· 2 years ago
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
Halfdeck
· 2 years ago
"Well as Andy said it was him that brought it up"
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a dumb a**."
Look, Rose, I know its not fair. The only justification for Greg Boser's behavior is "she started it" which doesn't fly. Like I posted over on Sphinn, his idea is dumb. I don't think anyone else in that thread objected as strongly and bluntly as I did to Greg's post. But you did cast the first stone, and what happened next isn't all that unexpected.
It's Friday night - make yourself a drink and relax (I'm way ahead of you).
Halfdeck
· 2 years ago
"Well as Andy said it was him that brought it up"
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a dumb a**."
Look, Rose, I know its not fair. The only justification for Greg Boser's behavior is "she started it" which doesn't fly. Like I posted over on Sphinn, his idea is dumb. I don't think anyone else in that thread objected as strongly and bluntly as I did to Greg's post. But you did cast the first stone, and what happened next isn't all that unexpected.
It's Friday night - make yourself a drink and relax (I'm way ahead of you).
Halfdeck
· 2 years ago
[Andy, I hit a spam filter so I'm resubmitting this again, in case my other comment gets lost in the spam box]
"Well as Andy said it was him that brought it up"
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a ..."
Rose, I don't think anyone else objected as bluntly as I did to Greg's post. But you did cast the first stone, so what happened next shouldn't have surprised you, even if it was completely uncalled for and unjustifiable.
It's Friday - make yourself a drink and relax (I'm way ahead of you).
Halfdeck
· 2 years ago
[Andy, I hit a spam filter so I'm resubmitting this again, in case my other comment gets lost in the spam box]
"Well as Andy said it was him that brought it up"
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a ..."
Rose, I don't think anyone else objected as bluntly as I did to Greg's post. But you did cast the first stone, so what happened next shouldn't have surprised you, even if it was completely uncalled for and unjustifiable.
It's Friday - make yourself a drink and relax (I'm way ahead of you).
Dane Morgan
· 2 years ago
Hey Andy, are you blocking co comments on your blog somehow? I don't get the tracking tool, and it's not picking up the comments here after I commented.
Dane Morgan
· 2 years ago
Hey Andy, are you blocking co comments on your blog somehow? I don't get the tracking tool, and it's not picking up the comments here after I commented.
Daniel
· 2 years ago
Just wanted to update everyone: We have changed our click tracking this evening to a 'cleaner' version. While there was no real evidence that our previous links weren't passing PR, and a lot of the same claims can be made about the new links, we figured it couldn't hurt to take the safer route.
A big thanks for sebastian on sphinn for digging up a solution.
Daniel
· 2 years ago
Just wanted to update everyone: We have changed our click tracking this evening to a 'cleaner' version. While there was no real evidence that our previous links weren't passing PR, and a lot of the same claims can be made about the new links, we figured it couldn't hurt to take the safer route.
A big thanks for sebastian on sphinn for digging up a solution.
All Comments
It's my own cartoons, so you might even be entertained!
Keep up the good work.
Johnny
PS: If you like the cartoons, I could provide you a "widget" to display them in some obsure corner of your site...for FREE!
It's my own cartoons, so you might even be entertained!
Keep up the good work.
Johnny
PS: If you like the cartoons, I could provide you a "widget" to display them in some obsure corner of your site...for FREE!
For me, the best implentation would be some kind of 301 implementation that would be seen by all.
Something like click.php?url=url which when followed returned
header('HTTP/1.1 301 Moved Permanently');
header('Location: '.$url);
Although such an approach would need a little additional work to check if it was a verified url at the backend.
Provided its not a 302, then personally, I dont see any problem.
In terms of their existing set up, my gut feeling is that the onclick event is irrelevant and that PR is passed simply because its contained within a straight a href container. Its not produced via document.write (which would be an issue as the bots wouldnt see it)
Anyhow, this can be tested can it not? Just gotta stick up a secret page or domain and point a page from BC to the page that nobody knows about, and voila, end of hypothesis. :D
I also found some blogs whilst testing that didn't have any backlinks being reported at all, from anywhere!
A redirect isn't obvious for users, and is ugly just like the Alexa redirects - there are a few other options but the current one in use is the cleanest.
If you use other methods, you are hiding the fact you are redirecting.
There really should be a published best practice or clearer indication.
For me, the best implentation would be some kind of 301 implementation that would be seen by all.
Something like click.php?url=url which when followed returned
header('HTTP/1.1 301 Moved Permanently');
header('Location: '.$url);
Although such an approach would need a little additional work to check if it was a verified url at the backend.
Provided its not a 302, then personally, I dont see any problem.
In terms of their existing set up, my gut feeling is that the onclick event is irrelevant and that PR is passed simply because its contained within a straight a href container. Its not produced via document.write (which would be an issue as the bots wouldnt see it)
Anyhow, this can be tested can it not? Just gotta stick up a secret page or domain and point a page from BC to the page that nobody knows about, and voila, end of hypothesis. :D
I also found some blogs whilst testing that didn't have any backlinks being reported at all, from anywhere!
A redirect isn't obvious for users, and is ugly just like the Alexa redirects - there are a few other options but the current one in use is the cleanest.
If you use other methods, you are hiding the fact you are redirecting.
There really should be a published best practice or clearer indication.
It may look like a clear link, but is not pass PR.
http://forums.searchenginewatch.com/showthread....
It is a discussion where Matt Cutts and Brian White both from Google mention that OnClick Java does not pass PageRank.
If you look closely you will see that it is very similar to what BC uses.
PS: Andy I don't like you. Everyone else may think that you are all that and a bag of chips, but I think you are nothing more than a BC Guppy with a brown nose.
In my article there is a whole section extensively quoting from that page on SEW, and also from other references.
There is a good chance that if I have extensively quoted from it, then I also read it.
But then I am also showing proof that as far as is possible with the tools Google is providing currently, that links with onclick can be counted by Google.
I didn't make up that SERP, it shows a link coming from a blog profile page on Blogcatalog.
I don't know why Google decide to show that link, any more than I don't know why Google choose not to show me links from so many other sites.
The Washinton Post situation was totally different, they were selling off-topic links which was obvious with a human inspection.
The links from Blogcatalog are provided free, and lead from a blog profile that is within a category, and point to the related blog.
And that develops as a pattern. Every time any one asks a question that could only be answered by resting this issue, Google's representatives remain silent.
The simple fact is that this does not close your case as you hope, and even the least rational reflection will bring you to that understanding. Does it prove you are wrong? No. It doesn't do that either, but a lot of other information leads to the invertible conclusion that either your view is incorrect to some extent or that Google is becoming fractured and schizoid.
It may look like a clear link, but is not pass PR.
http://forums.searchenginewatch.com/showthread....
It is a discussion where Matt Cutts and Brian White both from Google mention that OnClick Java does not pass PageRank.
If you look closely you will see that it is very similar to what BC uses.
PS: Andy I don't like you. Everyone else may think that you are all that and a bag of chips, but I think you are nothing more than a BC Guppy with a brown nose.
In my article there is a whole section extensively quoting from that page on SEW, and also from other references.
There is a good chance that if I have extensively quoted from it, then I also read it.
But then I am also showing proof that as far as is possible with the tools Google is providing currently, that links with onclick can be counted by Google.
I didn't make up that SERP, it shows a link coming from a blog profile page on Blogcatalog.
I don't know why Google decide to show that link, any more than I don't know why Google choose not to show me links from so many other sites.
The Washinton Post situation was totally different, they were selling off-topic links which was obvious with a human inspection.
The links from Blogcatalog are provided free, and lead from a blog profile that is within a category, and point to the related blog.
And that develops as a pattern. Every time any one asks a question that could only be answered by resting this issue, Google's representatives remain silent.
The simple fact is that this does not close your case as you hope, and even the least rational reflection will bring you to that understanding. Does it prove you are wrong? No. It doesn't do that either, but a lot of other information leads to the invertible conclusion that either your view is incorrect to some extent or that Google is becoming fractured and schizoid.
I get why BC want to track such things sure, info is power etc. Maybe they aren't interested in users who surf Javascript disabled who knows, a detractor might say that this was a way of restricting link juice but...heck that would be so pre 2007 it just wouldn't be funny. PR leakage I'm sure is the least of a site like BC's worries. We all can all name a good few domains that link out like crazy yet rank for many things. Where is the PR leakage theory there I ask?
Im sure too that those search dudes play around in the space and plant all manner of anti seo landmines all over the shop simply to cause confusion and uncertainty. They'd be silly not to. Why they still give us guys tools to deconstruct them is very odd and yes, makes you think twice about what we are being shown anyways!
The onclick example in that SEW thread is slightly different to this too, as in the BC onclick event example, it calls a function and not a full url. The example cited at SEW cited too separate full urls. The one in the href and the one in the onclick call, a marginal difference but a difference nontheless.
I'm surpised that MC and BW even commented in that thread as really, there shouldn't be any problem with a company tracking what their users do. Where was the harm in washington post doing what they did? Who got killed even. Why should a little bit of tracking effect the ultimate destination? As long as the user ends up at what they see in the status bar then it really shouldn't matter.
That said, as the onclick event would only work for users with JS enabled, then perhaps it would be beter for them to simply document.write the output of the onclick aspect of the code, that way, the bot wouldn't see it anyway, so it becomes a moot point.
Ultimately the intent is to give users a link and track user behaviour. The SE's shouldn't get all bent out of shape over that.
It's okay to sell links, but only for those on Google's short list, and it would seem that, at least a year ago, the Post wasn't on that list.
Sometimes those guys have a habit of creating FUD. It isn't the 1st time something will be left open to interpretation and wont be the last either!
I get why BC want to track such things sure, info is power etc. Maybe they aren't interested in users who surf Javascript disabled who knows, a detractor might say that this was a way of restricting link juice but...heck that would be so pre 2007 it just wouldn't be funny. PR leakage I'm sure is the least of a site like BC's worries. We all can all name a good few domains that link out like crazy yet rank for many things. Where is the PR leakage theory there I ask?
Im sure too that those search dudes play around in the space and plant all manner of anti seo landmines all over the shop simply to cause confusion and uncertainty. They'd be silly not to. Why they still give us guys tools to deconstruct them is very odd and yes, makes you think twice about what we are being shown anyways!
The onclick example in that SEW thread is slightly different to this too, as in the BC onclick event example, it calls a function and not a full url. The example cited at SEW cited too separate full urls. The one in the href and the one in the onclick call, a marginal difference but a difference nontheless.
I'm surpised that MC and BW even commented in that thread as really, there shouldn't be any problem with a company tracking what their users do. Where was the harm in washington post doing what they did? Who got killed even. Why should a little bit of tracking effect the ultimate destination? As long as the user ends up at what they see in the status bar then it really shouldn't matter.
That said, as the onclick event would only work for users with JS enabled, then perhaps it would be beter for them to simply document.write the output of the onclick aspect of the code, that way, the bot wouldn't see it anyway, so it becomes a moot point.
Ultimately the intent is to give users a link and track user behaviour. The SE's shouldn't get all bent out of shape over that.
It's okay to sell links, but only for those on Google's short list, and it would seem that, at least a year ago, the Post wasn't on that list.
Sometimes those guys have a habit of creating FUD. It isn't the 1st time something will be left open to interpretation and wont be the last either!
Great post. I find the whole area so confusing
It's also annoying the way Google so consistently choose not to answer questions clearly! They know what issues are being had, why don't they clarify this for people more often. The amount of times in the past 2 weeks I have seen people asking questions and getting vague answers is high. Too high.
Overall they're a great company but the great stone wall of silence makes it difficult for the lay man to get by.
Great post. I find the whole area so confusing
It's also annoying the way Google so consistently choose not to answer questions clearly! They know what issues are being had, why don't they clarify this for people more often. The amount of times in the past 2 weeks I have seen people asking questions and getting vague answers is high. Too high.
Overall they're a great company but the great stone wall of silence makes it difficult for the lay man to get by.
When he says this he is speaking about the webmaster tool, not the link: operator. When asked if this same advice applies to the link: operator he does not reply.
So while it is *possible* that this could relate to the link: operator, such a conclusion can not be drawn from this data.
Since the webmaster tool includes nofollowed links where the link: operator does not we do know there are differences and it is not unreasonable to consider that this might be one of them.
"That's just not true. Or rather, if most SEOs take it that way, then most SEOs are wrong. :) Halfdeck actually bothered to hunt down where I've said it before. I'm happy just to re-state: if you assume links for a link: query automatically carry weight, then you've made a faulty assumption."
Am I entirely alone, or does the fact that he took the time to make that post and STILL did not address the topic of the discussion say something?
When he says this he is speaking about the webmaster tool, not the link: operator. When asked if this same advice applies to the link: operator he does not reply.
So while it is *possible* that this could relate to the link: operator, such a conclusion can not be drawn from this data.
Since the webmaster tool includes nofollowed links where the link: operator does not we do know there are differences and it is not unreasonable to consider that this might be one of them.
"That's just not true. Or rather, if most SEOs take it that way, then most SEOs are wrong. :) Halfdeck actually bothered to hunt down where I've said it before. I'm happy just to re-state: if you assume links for a link: query automatically carry weight, then you've made a faulty assumption."
Am I entirely alone, or does the fact that he took the time to make that post and STILL did not address the topic of the discussion say something?
1. Install a sitewide onclick link pointing at a new page.
2. Track crawler activity and indexing.
Sitewide links will direct enough PageRank to a new page to get it indexed in a few days - assuming onclick doesn't block link juice / bot crawl.
Anyone want to bet money on what happens? :)
I found blogs on Blogcatalog that had been established for a while, at least a few weeks, and Google's link commend wasn't showing up any links at all.
This is getting into Vanessa Fox "SEO is Art" territory.
I linked in the article to test results stating that Google does crawl the links, so it is not the same as "nofollow"
I know Rose read that too, but without a lot more details, it is impossible to say whether those tests were conclusive.
I am intending to do some additional testing to see if I can Gogglebomb a page to rank for something weird, but that won't be conclusive because the onclick might modify the juice given.
The 2 best terms to describe SEO are YMMV and FWIW
Yeah, it should be more than enough to get a page indexed.
"It was verified that Google does follow the links and index the pages. However, from a pure PageRank perspective, no “juice†is being passed."
Ok, now this makes me wonder: does Google index URLs then run PageRank calculations, dropping URLs with very low PageRank after indexing? Or does Google "guess" the PageRanks of newly discovered URLs and prevent indexing of URLs that probably have very low PageRank?
Guessing the PageRanks of every newly discovered URL would be more involved, I imagine, than just dropping low PageRank URLs from the index during the PageRank iteration phase.
So, his statement doesn't make sense if I assume that Google guestimates the PageRanks of a newly found URL and index or not index depending on that PageRank because:
If no PageRank is being passed, a URL shouldn't be indexed. If very little PageRank is being passed (but enough to force indexing), then the URL should be in the supplemental results.
Therefore, if he set up his test correctly and his page got indexed, then his claim that no juice is being passed is false - assuming Googlebot is proactive about not crawling or indexing URLs with very low PageRank.
If instead I assume Google will crawl first and drop URLs later, then I'd ask: "did the page stay indexed? Or was it dropped from the index a month or two later?"
Also keep in mind a page that Google includes in its main index will never have zero PageRank; the minimum PageRank of a page is (1-d) or .15 if we use the original PageRank formula with d = .85. So "or the PageRank they do pass is so small that it is not seen." isn't convincing. Was the PageRank passed or was it "created" by the page itself?
Finally, his post was written back in Aug 2006 when even on WMW the line between indexing/crawling and PageRank was somewhat unclear. He also doesn't say how he measured PageRank (my guess is the toolbar).
We all love to hear a simple "yes/no" answer to things like this, but Google is a program, and when we ask questions about a program with thousands of lines of code, the answers can sometimes be pretty complicated.
http://sphinn.com/story/5310#c7990
That also doesn't make sense as I have responded.
Webmaster console is generally a lot more up to date
Webmaster console includes nofollow links
No answer on the onclick though
1. Install a sitewide onclick link pointing at a new page.
2. Track crawler activity and indexing.
Sitewide links will direct enough PageRank to a new page to get it indexed in a few days - assuming onclick doesn't block link juice / bot crawl.
Anyone want to bet money on what happens? :)
I found blogs on Blogcatalog that had been established for a while, at least a few weeks, and Google's link commend wasn't showing up any links at all.
This is getting into Vanessa Fox "SEO is Art" territory.
I linked in the article to test results stating that Google does crawl the links, so it is not the same as "nofollow"
I know Rose read that too, but without a lot more details, it is impossible to say whether those tests were conclusive.
I am intending to do some additional testing to see if I can Gogglebomb a page to rank for something weird, but that won't be conclusive because the onclick might modify the juice given.
The 2 best terms to describe SEO are YMMV and FWIW
Yeah, it should be more than enough to get a page indexed.
"It was verified that Google does follow the links and index the pages. However, from a pure PageRank perspective, no “juice†is being passed."
Ok, now this makes me wonder: does Google index URLs then run PageRank calculations, dropping URLs with very low PageRank after indexing? Or does Google "guess" the PageRanks of newly discovered URLs and prevent indexing of URLs that probably have very low PageRank?
Guessing the PageRanks of every newly discovered URL would be more involved, I imagine, than just dropping low PageRank URLs from the index during the PageRank iteration phase.
So, his statement doesn't make sense if I assume that Google guestimates the PageRanks of a newly found URL and index or not index depending on that PageRank because:
If no PageRank is being passed, a URL shouldn't be indexed. If very little PageRank is being passed (but enough to force indexing), then the URL should be in the supplemental results.
Therefore, if he set up his test correctly and his page got indexed, then his claim that no juice is being passed is false - assuming Googlebot is proactive about not crawling or indexing URLs with very low PageRank.
If instead I assume Google will crawl first and drop URLs later, then I'd ask: "did the page stay indexed? Or was it dropped from the index a month or two later?"
Also keep in mind a page that Google includes in its main index will never have zero PageRank; the minimum PageRank of a page is (1-d) or .15 if we use the original PageRank formula with d = .85. So "or the PageRank they do pass is so small that it is not seen." isn't convincing. Was the PageRank passed or was it "created" by the page itself?
Finally, his post was written back in Aug 2006 when even on WMW the line between indexing/crawling and PageRank was somewhat unclear. He also doesn't say how he measured PageRank (my guess is the toolbar).
We all love to hear a simple "yes/no" answer to things like this, but Google is a program, and when we ask questions about a program with thousands of lines of code, the answers can sometimes be pretty complicated.
http://sphinn.com/story/5310#c7990
That also doesn't make sense as I have responded.
Webmaster console is generally a lot more up to date
Webmaster console includes nofollow links
No answer on the onclick though
It is not just SEJ, and there is no proof that money changed hands
Maybe Pat Gavin buys more drinks at SES than Matt Cutts
It is not just SEJ, and there is no proof that money changed hands
Maybe Pat Gavin buys more drinks at SES than Matt Cutts
I didn't do that for my original research because I wouldn't want to be a scaremonger. There would be too many people on Blogcatalog who wouldn't understand the reason why they were doing the search.
Dane's further research if anything corroborates this post. A large number of people are finding links.
I can only see 2.5% of my own links using the link command, many from extremely notable blogs.
And It is my research, not Andy's. My methods are mine.
I didn't do that for my original research because I wouldn't want to be a scaremonger. There would be too many people on Blogcatalog who wouldn't understand the reason why they were doing the search.
Dane's further research if anything corroborates this post. A large number of people are finding links.
I can only see 2.5% of my own links using the link command, many from extremely notable blogs.
And It is my research, not Andy's. My methods are mine.
Nope. Still, it would be too easy if I can check the effectiveness of a paid link by running a link: command.
For example, link:customermagnetism.com brings up (ok, no question these are paid links):
seroundtable.com
unofficialseoblog.com
v7n.com
seo-scoop.com
arcadeshack.com (arcade site linking to an SEO site? yeah.. doesn't pass the smell test)
seo4fun.com (yep, I'm selling them a link too)
videogamesblogger.com
thebostonchannel.com
Do those paid links count? According to the link: command, they do ... if you assume that all URLs returned by the link: command pass juice.
I'm not saying they don't pass juice. My point is if link: command gives away that information, its revealing way too much info; I don't believe Google is dumb enough to let SEOs know which of their backlinks are actually counting.
Nope. Still, it would be too easy if I can check the effectiveness of a paid link by running a link: command.
For example, link:customermagnetism.com brings up (ok, no question these are paid links):
seroundtable.com
unofficialseoblog.com
v7n.com
seo-scoop.com
arcadeshack.com (arcade site linking to an SEO site? yeah.. doesn't pass the smell test)
seo4fun.com (yep, I'm selling them a link too)
videogamesblogger.com
thebostonchannel.com
Do those paid links count? According to the link: command, they do ... if you assume that all URLs returned by the link: command pass juice.
I'm not saying they don't pass juice. My point is if link: command gives away that information, its revealing way too much info; I don't believe Google is dumb enough to let SEOs know which of their backlinks are actually counting.
The claim of the initial post was
"BlogCatalog - Blog Catalog Doesn't Pass Pagerank"
You would hope that of my 28600 (though Yahoo has reported up to 40000 in the past) links that Google only report as 863 and then only let me see 189, that there would be some value to them.
Maybe it shows the links that would have been counted if they hadn't been detected as piad (FUD)
And it could well be that links using onclick are being counted or lumped in with paid links, but there is no proof, and the statements that Google have made are as always too ambiguous to act as proof.
The claim of the initial post was
"BlogCatalog - Blog Catalog Doesn't Pass Pagerank"
You would hope that of my 28600 (though Yahoo has reported up to 40000 in the past) links that Google only report as 863 and then only let me see 189, that there would be some value to them.
Maybe it shows the links that would have been counted if they hadn't been detected as piad (FUD)
And it could well be that links using onclick are being counted or lumped in with paid links, but there is no proof, and the statements that Google have made are as always too ambiguous to act as proof.
That should cause further brain swelling for you Half ;-)
Excuse my dumbness here, but if Andy you are saying that an anchor with an onclick event doesn't pass pagerank I'm ever so slightly confused:
That being the case then Urchin events added to a link would inhibit the flow of PR. I've never tested this, but I'd be mighty shocked if Google advised people to use their tracking software without mentioning that doing so may actually affect the juice of that link. Maybe I haven't read enough of the posts and comments, but this couldn't be a blanket change in behaviour applied to ANY anchor containing an onclick attribute?
Rgds
Richard
You need to use square brackets for the code here
The links are interesting reading if you haven't seen them before, and you should probably read the 2 threads on Sphinn as well, both are on the front page currently.
One of the sites linked though to was a useful tutorial on Urchin.
I am saying it should still pass pagerank, Rose is saying it isn't based purely on 2 threads.
Matt Cutts is also now saying the link: commend doesn't effectively mean anything at all as far as which links pass value.
Maybe every site should just use nofollow on every single link, start a national nofollow day
That should cause further brain swelling for you Half ;-)
Excuse my dumbness here, but if Andy you are saying that an anchor with an onclick event doesn't pass pagerank I'm ever so slightly confused:
That being the case then Urchin events added to a link would inhibit the flow of PR. I've never tested this, but I'd be mighty shocked if Google advised people to use their tracking software without mentioning that doing so may actually affect the juice of that link. Maybe I haven't read enough of the posts and comments, but this couldn't be a blanket change in behaviour applied to ANY anchor containing an onclick attribute?
Rgds
Richard
You need to use square brackets for the code here
The links are interesting reading if you haven't seen them before, and you should probably read the 2 threads on Sphinn as well, both are on the front page currently.
One of the sites linked though to was a useful tutorial on Urchin.
I am saying it should still pass pagerank, Rose is saying it isn't based purely on 2 threads.
Matt Cutts is also now saying the link: commend doesn't effectively mean anything at all as far as which links pass value.
Maybe every site should just use nofollow on every single link, start a national nofollow day
We had a guy over on Google Groups recently proposing something similar - nofollowing every link on your site in case Google penalises you... *sigh*
So just that I'm clear about what you're saying Andy - my reading of the above gives the impression that your testing of an anchor with an onclick attribute set passes little or no juice?
I see Half chiming in with some interesting thoughts on whether a page would be indexed without pagerank (and he's one guy who knows his PR).
Whatever test you set up for this, it is never going to be conclusive.
I have always maintained that Google could possibly be doing some kind of 2 pass system for link allocation
1. Decide which links they should use
2. Decide which weighting to give them
Sitewide links in the sidebar people suggest have less value than links in content, and the same is possibly true of comment links if detected.
At what stage in the calculations is that being discounted?
How does the juice deducted get allocated?
If all you have is sidebar links on your front page, then it doesn't matter what percentage they are allocated, it will all be in proportion 0.15:0.15 is the same as 1:1
Matt is effectively saying that just because links show up, that doesn't mean they have value, thus even the links I have found to sites from Blogcatalog could be totally meaningless.
That effectively makes the link: command about as useful as "Are you feeling lucky"
We had a guy over on Google Groups recently proposing something similar - nofollowing every link on your site in case Google penalises you... *sigh*
So just that I'm clear about what you're saying Andy - my reading of the above gives the impression that your testing of an anchor with an onclick attribute set passes little or no juice?
I see Half chiming in with some interesting thoughts on whether a page would be indexed without pagerank (and he's one guy who knows his PR).
Whatever test you set up for this, it is never going to be conclusive.
I have always maintained that Google could possibly be doing some kind of 2 pass system for link allocation
1. Decide which links they should use
2. Decide which weighting to give them
Sitewide links in the sidebar people suggest have less value than links in content, and the same is possibly true of comment links if detected.
At what stage in the calculations is that being discounted?
How does the juice deducted get allocated?
If all you have is sidebar links on your front page, then it doesn't matter what percentage they are allocated, it will all be in proportion 0.15:0.15 is the same as 1:1
Matt is effectively saying that just because links show up, that doesn't mean they have value, thus even the links I have found to sites from Blogcatalog could be totally meaningless.
That effectively makes the link: command about as useful as "Are you feeling lucky"
It is actually relevant to this conversation, because Rose seems to think the only directory which might have bad signals is Blogcatalog, whereas I would personally be more concerned with the links on MBL if I was worried about link credit.
It is actually relevant to this conversation, because Rose seems to think the only directory which might have bad signals is Blogcatalog, whereas I would personally be more concerned with the links on MBL if I was worried about link credit.
Got to love the SEO circle... I can't call Andy here a Guppy yet users can call me a dumb a**.
I can't mention BC but Andy can.
Andy if you think this all has to do with my being removed for pointing out their lack of terms of service and that they were a clone- you better think again. The funny thing is NOW Daniel is making changes.
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
You just don't get it and you never will Andy.
Here let me dumb it down.
It was Mr Doubts who first found this June 7, 2007.
He reported it here.
http://moredoubts.wordpress.com/2007/06/07/blog...
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments to wake them up.
Credit goes to Mr. Doubts though.
Got to love the SEO circle... I can't call Andy here a Guppy yet users can call me a dumb a**.
I can't mention BC but Andy can.
Andy if you think this all has to do with my being removed for pointing out their lack of terms of service and that they were a clone- you better think again. The funny thing is NOW Daniel is making changes.
Quoting Danny " Thank you for bringing it to our attention, don't be surprised if you see it implemented soon."
You just don't get it and you never will Andy.
Here let me dumb it down.
It was Mr Doubts who first found this June 7, 2007.
He reported it here.
http://moredoubts.wordpress.com/2007/06/07/blog...
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments to wake them up.
Credit goes to Mr. Doubts though.
Andy- why did you delete my comment?
Andy- why did you delete my comment?
He was replying to Sebastian.
It was Mr Doubts who first found this June 7, 2007.
http://moredoubts.wordpress.com/2007/06/07/blog...
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
Credit goes to Mr. Doubts for first finding this though.
Andy this was not personal. It had already been pointed out back in June and I was only stating that it should be fixed.
Halfdeck why it ok for Andy here to bring it up yet not me? I was only quoting what Andy said to Shawn. That certainly did not WIN brownie points in the debate.
He was replying to Sebastian.
It was Mr Doubts who first found this June 7, 2007.
http://moredoubts.wordpress.com/2007/06/07/blog...
BlogCatalog should have listened then, but no they were to stubborn to.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
Credit goes to Mr. Doubts for first finding this though.
Andy this was not personal. It had already been pointed out back in June and I was only stating that it should be fixed.
Halfdeck why it ok for Andy here to bring it up yet not me? I was only quoting what Andy said to Shawn. That certainly did not WIN brownie points in the debate.
He was replying to Sebastian.
He was replying to Sebastian.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
It took a post of 58 Comments on Sphinn to wake them up.
The good thing is I hope they now change things around to be fair.
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a dumb a**."
Look, Rose, I know its not fair. The only justification for Greg Boser's behavior is "she started it" which doesn't fly. Like I posted over on Sphinn, his idea is dumb. I don't think anyone else in that thread objected as strongly and bluntly as I did to Greg's post. But you did cast the first stone, and what happened next isn't all that unexpected.
It's Friday night - make yourself a drink and relax (I'm way ahead of you).
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a dumb a**."
Look, Rose, I know its not fair. The only justification for Greg Boser's behavior is "she started it" which doesn't fly. Like I posted over on Sphinn, his idea is dumb. I don't think anyone else in that thread objected as strongly and bluntly as I did to Greg's post. But you did cast the first stone, and what happened next isn't all that unexpected.
It's Friday night - make yourself a drink and relax (I'm way ahead of you).
"Well as Andy said it was him that brought it up"
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a ..."
Rose, I don't think anyone else objected as bluntly as I did to Greg's post. But you did cast the first stone, so what happened next shouldn't have surprised you, even if it was completely uncalled for and unjustifiable.
It's Friday - make yourself a drink and relax (I'm way ahead of you).
"Well as Andy said it was him that brought it up"
Yeah, I know.
"and now you say oh nevermind. Well why is it ok for Andy to bring up off topic issues?"
Andy explained why he didn't think the issue was off-topic, so I let it go.
"I can't call Andy here a Guppy yet users can call me a ..."
Rose, I don't think anyone else objected as bluntly as I did to Greg's post. But you did cast the first stone, so what happened next shouldn't have surprised you, even if it was completely uncalled for and unjustifiable.
It's Friday - make yourself a drink and relax (I'm way ahead of you).
A big thanks for sebastian on sphinn for digging up a solution.
A big thanks for sebastian on sphinn for digging up a solution.