Here's How to Build One-Way Links

[Tyler Banfield]

I've been meaning to make this post on my blog, but I thought I would just go ahead and put it on DP...

Many people want to know how many one-way links they can build between their sites. The answer is simple:

[(W x 0.5) - 0.5] x W whereas W equals the Number of Websites

For example, if you would like to build one-way links between ten websites, simply plug ten into the formula:

[(10 x 0.5) - 0.5] x 10
[5 - 0.5] x 10
4.5 x 10 = 45

Once you know how many one-way links you can create, you can use a simple grid to map out which websites to link together:

1 links to 2, 3, 4, 5, 6
2 links to 3, 4, 5, 6, 7
3 links to 4, 5, 6, 7, 8
4 links to 5, 6, 7, 8, 9
5 links to 6, 7, 8, 9, 10
6 links to 7, 8, 9, 10
7 links to 8, 9, 10, 1
8 links to 9, 10, 1, 2
9 links to 10, 1, 2, 3
10 links to 1, 2, 3, 4

As you can see, regardless of the amount of websites you are using, the first half of the sites link to exactly half the amount of sites, while the last half links to exactly half the amount of sites minus one. For example:

20 Websites
1 to 10: Each site links to ten other sites
11 to 20: Each site links to nine other sites

Remember, when it comes to one-way links, quality still matters. An unrelated one-way link is no better than a reciprocal link.

[career_info_fun]

awesome post, useful info thanks a lot:cool:

[xxfb123]

This is great info. Never thought about it in this direction. I have always trying to put all my websites links on other website. Making them all reciprocal links. I am going to try this. Good Job.

[jimkarter]

Good info. thanx.

[carmaniac]

Very useful information.. thanks!

[sebastya]

I'm confused :confused:

[Tyler Banfield]

I'm confused :confused:

What's confusing you? I can probably clarify...

[seoindiaweb]

This is simple... but the only problem is.. Google doesnt give much weight or rather no weight to links from same Class C IP's or i have 60 domains all with PR4+ ...

[jelqing]

I think the best way is to get other sites to link to yours - either paying them to do so, or offering such good content they are going to want to link - especially offering something interactive that is difficult to replicate for themselves.

[sebastya]

What's confusing you? I can probably clarify...

see you have 45 there...45 what? one-way links?

[Tyler Banfield]

This is simple... but the only problem is.. Google doesnt give much weight or rather no weight to links from same Class C IP's or i have 60 domains all with PR4+ ...

That's why you might need to use different IPs ;)

I think the best way is to get other sites to link to yours - either paying them to do so, or offering such good content they are going to want to link - especially offering something interactive that is difficult to replicate for themselves.

I have to disagree

see you have 45 there...45 what? one-way links?

Yes, 45 one way links total

[bookworm-seo]

Brilliant formula Tyler. It seems every time I find one of your posts, there's more quality content to read and learn about. What blog were you going to put this on? I'll be happy to give you a free one-way link to it when I post about this idea at my SEO blog (see sig).

About class C ... how do you tell? And how do you avoid having all the same class C IPs? Do you need to go with different hosts for all your sites??

[sebastya]

Yes, 45 one way links total

But how are you getting 45 links from 10 sites?

[Asen Kovachev]

How deep can you link before a link stops being a one way link?

I mean obviously if you do 1 -> 2, then 2 -> 3, then 3 -> 1, they are all one way but are easily detectable... (or 3way links as they call them)

so in your case you have 10 way link... I'd say that's a safe assumption...

but still how do you gather 0.5 to be the number?

[wasted soul]

nice calculation tyler
thank's for the info ;)

[MattUK]

I've never tried that before, but I'd guess that the bigger you went (more sites) then you'd stand a good chance of getting tagged as it being a link network.

Search engines can spot patterns in links. Though on a small scale it wouldn't be as noticeable.

[geegel]

I tend to doubt that this sort of scheme would remain undetected, on the long term at least. Perhaps a better way to avoid detection would be to use less than the total of 45 links, maybe 25 or 30. There is also another danger looming, especially with websites that allow feedback from users. I imagine the case of a forum: if a user posts a link that refers to a site in the network of websites, the whole scheme would pretty much fall.

Anyway brilliant piece of SEO math.

Best regards, George

[flagday]

Tyler,

I have found many of your posts in the past to be helpful and insightful...

But this!

It seems clear to me that PAGE strength (in terms of inbound links) is far more important than something as transient as SITE strength. This equation, while elegant, does not appear at all accurate.

[Tekime]

I've been meaning to make this post on my blog, but I thought I would just go ahead and put it on DP...

Interesting post. Tell me, is this a general theory then or do you have some data that brought you to this formula? Obviously there are no guarantees with search engines, just curious. :)

I'm interested, too, because I'm developing a webmaster toolkit which includes a linking engine. My goal is to make a very flexible system that can handle custom formulas such as this and monitor progress over time. Progress in this case is hard to nail down to just a few factors, though, considering the number of variables at work here in addition to just where and how many links there are. (IP, relevancy, placement, strength, etc.)

[Tyler Banfield]

But how are you getting 45 links from 10 sites?

Site 1 links to Sites 2, 3, 4, 5, 6
Site 2 links to Sites 3, 4, 5, 6, 7
Site 3 links to Sites 4, 5, 6, 7, 8
Site 4 links to Sites 5, 6, 7, 8, 9
Site 5 links to Sites 6, 7, 8, 9, 10
Site 6 links to Sites 7, 8, 9, 10
Site 7 links to Sites 8, 9, 10, 1
Site 8 links to Sites 9, 10, 1, 2
Site 9 links to Sites 10, 1, 2, 3
Site 10 links to Sites 1, 2, 3, 4

Total: 45 Links

How deep can you link before a link stops being a one way link?

I mean obviously if you do 1 -> 2, then 2 -> 3, then 3 -> 1, they are all one way but are easily detectable... (or 3way links as they call them)

so in your case you have 10 way link... I'd say that's a safe assumption...

but still how do you gather 0.5 to be the number?

Not exactly true. In this scenario, 3 does not link back to 1.

As far as the 0.5, that's just the number it worked out to be

I've never tried that before, but I'd guess that the bigger you went (more sites) then you'd stand a good chance of getting tagged as it being a link network.

Search engines can spot patterns in links. Though on a small scale it wouldn't be as noticeable.

I'm actually going to test this out (on a large scale), just to see how it works. I'll let you guys know ;)

Tyler,

I have found many of your posts in the past to be helpful and insightful...

But this!

It seems clear to me that PAGE strength (in terms of inbound links) is far more important than something as transient as SITE strength. This equation, while elegant, does not appear at all accurate.

This is a valid point, but not at all set in stone. I have many sites that rank well with the majority of links pointing to the site, not individual pages.

Interesting post. Tell me, is this a general theory then or do you have some data that brought you to this formula? Obviously there are no guarantees with search engines, just curious. :)

I'm interested, too, because I'm developing a webmaster toolkit which includes a linking engine. My goal is to make a very flexible system that can handle custom formulas such as this and monitor progress over time. Progress in this case is hard to nail down to just a few factors, though, considering the number of variables at work here in addition to just where and how many links there are. (IP, relevancy, placement, strength, etc.)

Over the weekend I actually sat down and drew up an experiment, so over the coming months I will put this equation to the test (on a large scale) and see how effective it really is.

[Rasputin]

Hey great thinking Tyler,

Right or wrong (and it looks good to me, in so far as I understand things) it has focussed my mind on getting rid of all reciprocal links among my own sites.

One small comment, when you say:

As you can see, regardless of the amount of websites you are using, the first half of the sites link to exactly half the amount of sites, while the last half links to exactly half the amount of sites minus one.

that applies only for an even number of sites, not an odd number, (unless i got my sums wrong)

Cheers

[Tyler Banfield]

As you can see, regardless of the amount of websites you are using, the first half of the sites link to exactly half the amount of sites, while the last half links to exactly half the amount of sites minus one.

that applies only for an even number of sites, not an odd number, (unless i got my sums wrong)


Very good point! Thanks for catching me on this, I had completely overlooked it. When it comes to an odd number, each site will have the same number of links. Let's use five as an example:

[(5 x 0.5) - 0.5] x 5 = 10

1 links to 2, 3
2 links to 3, 4
3 links to 4, 5
4 links to 5, 1
5 links to 1, 2

or seven:

[(7 x 0.5) - 0.5] x 7 = 21

1 links to 2, 3, 4
2 links to 3, 4, 5
3 links to 4, 5, 6
4 links to 5, 6, 7
5 links to 6, 7, 1
6 links to 7, 1, 2
7 links to 1, 2, 3

[Tyler Banfield]

that applies only for an even number of sites, not an odd number, (unless i got my sums wrong)

I also updated my blog post ;)

[homework62]

sorry for being a dumb noob but what are 1way links for?

[bookworm-seo]

Right or wrong (and it looks good to me, in so far as I understand things) it has focussed my mind on getting rid of all reciprocal links among my own sites.


SEOChat ranks for SEO... and it uses recip links with the other sites in the network, which links are coded in html. so they aren't necessarily bad. This is just another (more advanced) strategy for people looking for a different way of linking.

[Tyler Banfield]

sorry for being a dumb noob but what are 1way links for?

They are many SEOers who consider them superior to reciprocal links.

[geegel]

I observed that you theory avoids both reciprocal and 3 way links. If you would include 3 way links how many total links would there be in total?

A quick adaptation of the formula would be:

[(n x 0.5) + 0.5] x n

I might be wrong though.

Any thoughts?

Regards, George

[Sem-Advance]

Well this does help.

Though it should be pointed out there are 100 pieces to the algorithim puzzle so some relevant one way links to your sites on the same c class can help a bit....

You are not going to land 1 1 for viagra using the method described...

Spread it out to several different c class servers the effect intensifys....

but

You are not going to land 1 1 for viagra using the method described...

Good job link baiting ;)

Peace

[onedollar]

interesting formula, Tyler Banfield:)

[nezzz]

hmm I like this idea

[Herendethelesson]

But isnt there a risk of getting penalties from SEs for interlinking our own sites?

I read somewhere that the different class Cs may not benefit for long as G is getting smarter and looks at whois info?

Or am I being extremely paranoid?

[kkibak]

Cool idea, I'm glad to see someone else thinking the way I do. I think there's a potential problem, though:

1 links to 2, 3, 4, 5, 6
2 links to 3, 4, 5, 6, 7
3 links to 4, 5, 6, 7, 8
4 links to 5, 6, 7, 8, 9
5 links to 6, 7, 8, 9, 10
6 links to 7, 8, 9, 10
7 links to 8, 9, 10, 1


In this scenario, 1 links to 6 which links 7 which links to 1.

This is a straightforward "three way linking procedure" that I think search engines like Google would be able to detect.
---
My solution?

What I'm writing right now is a PHP script that would do the following:

1. Make it so that you start out with the program by entering a site on the network, it’s specifics (PR, etc.), and assign it an ID. Then have the user enter all the sites this site links to that are a part of the network.
2. Use the add tool to add a second site.
3. Enter a proposed link for the new site
4. Make sure this target site is not a site that already links to the proposed linking site. If it does, let the submitter know and make them say they are sure. Do this by both domain AND IP.
5. Make sure this proposed site does not link to a site that links to the proposed linker. If it does, let the submitter know and make them say they are sure. Do this by both domain AND IP.
6. Make sure this proposed site does not link to a site that links to a site that links to the proposed linker. If it does, let the submitter know and make them say they are sure. Do this by both domain AND IP.
7. Make it so that if a submitter ignores a warning message, this link is identified with a flag as a potentially detectable network link and make sure the admin is aware.
8. Make it so the program looks at sites and sees which other network sites are eligible link-mates. That is, have it check which pages the site could link to without throwing up a flag, and list those off so that the submitter knows what he can do to make the network more productive.

-----

Unless I'm making a mistake, the program above would prevent up to four-way linking... I doubt the search engines check further than that, but who knows.

I'm sure mine has some problems too, and it's working programmatically through trial / error just because I'm not good at coming up w/formulas..

Input?

[MrMarco]

Sounds good... Just to clarify something.
The idea for this works for 10 of your own sites only?
AND using different IP's for each?
Sub-Domains in that case won't work well, am I right?
I'm asking this because I'm considering the idea of adding sub-domains to my site made up of the different categories of articles.
I have over 20 categories of articles. Making sub-domains may make it easier for SEO. I'm not sure that's a good move or not... just thinking out loud.

again... I like the line of thinking - Thanks for the info
:)

[Sem-Advance]

But isnt there a risk of getting penalties from SEs for interlinking our own sites?

I read somewhere that the different class Cs may not benefit for long as G is getting smarter and looks at whois info?

Or am I being extremely paranoid?

I would not say penalties till you go to far....but yes there would be a filter placed if its found you are interlinking all of your sites and if they are not relevant a further dampening filter applied..

penalty would probably be placed if you linked every page of your site A to every page of your site B and vice versa, and both to site C and so on.

One link from each site should suffice.....

Peace

[Tyler Banfield]

I'm glad to see this thread is generating so many quality posts. Although my equation is obviously not perfect, my main objective was to generate good discussion, so I'm glad I posted the equation ;)

[Tyler Banfield]

I tried adding a link to my blog post to an article on Wikipedia, but big surprise, it got deleted in about a day :)

[MattUK]

They're really hot on any SEO postings ;)

[Tyler Banfield]

They're really hot on any SEO postings ;)

Yeah, I can't say that I blame them though :D