Pagerank Patent End of Life

[JS_Harris]

I thought this was interesting reading that will become relevant again now that the lifespan of some important patents reach the end of their lifespans. [news.ycombinator.com...]

Pagerank embodied a method of determining and assigning a final value to any given page and, while patented, it is rather hard to compete against. I am not referring to the actual values but rather to the methods used to obtain that final value in general.

Do you think the idea of displaying rank values within a browser window might resurge from another source when patents expire? If they do, do you think that webmasters may begin paying attention to it once again and possibly help make that new ranking system relevant?

note: Google's exclusive license expired in 2011 but Stanford retained rights, until now.

[martinibuster]

Seems like you're talking about two different things. PageRank is a ranking calculation. A display of a simulation of that ranking formula is something else.

There is nothing stopping anyone from displaying their own metrics. MozBar shows their proprietary ranking metrics. For a short time even Yahoo offered a Yahoo Rank meter in their toolbar.

Are you suggesting that a PageRank branded toolbar might be offered?

[keyplyr]

MozBar shows their proprietary ranking metrics.

Yes, and there's also the Alexa toolbar that continues to show rank, albeit a different type metric.

Google's PR was by far the standard for many years, so much so that it helped spawn the SEO industry. In all indication, that's why Google stopped using it, letting it eventually die, as noted above.

[JS_Harris]

I was talking about one thing, getting a pagerank metric back in action that displays the value of pages in Google's eyes. Obviously when a patent exists you cannot duplicate the same methods but now that it has expired you certainly can.

If someone was inclined to do the heavy lifting we could very well see something like an addon that reports a "pagerank" equivalent of Google's rankings. Many sites aggregate google rankings data and sell it for a monthly fee already so it would not be a huge leap to actually use formerly patented metrics to essentially turn pagerank back on, albeit not officially by Google.

A lot of people also have addons that add a lot of information directly onto the google results pages they see which isn't actually on those pages so I can totally see someone figuring out how to revive pagerank, now that they legally can. Link and website sellers would be delighted I'm sure, but there was some SEO benefit to seeing those figures too.

edit: I'm well aware that other companies have their own "value" assigned to pages. With the original patents expired someone could actually duplicate the old pagerank and provide a non-official value of GOOGLE's rankings once again, using much the same method they did. I'm not talking about creating a new value system, those exist already but are not as widely adopted because they were not beign applied against Google's rankings.

summary: Google created a blessing in pagerank early on which turned into a monster when the SEO community blossomed and ran away with it. I'm just curious if, now that the patents have expired, we'll see pagerank revived and applied against google rankings once again as an addon or service. And if so... will it once again become an important number to use in sales of websites, by link sellers etc.

[edited by: JS_Harris at 7:22 am (utc) on Feb 19, 2017]

[keyplyr]

I was talking about one thing, getting a pagerank metric back in action that displays the value of pages in Google's eyes

Well that's just it, Google doesn't use PR in the same way anymore so the endeavor would be futile without the other approx 7 hundred factors.

[JS_Harris]

Correct me if I'm wrong but pagerank was not a function of any equation, it was merely a reporting of where a page is ranked based on how many sites link to it and the weight those pages had. Sure, google has a bazillion claimed other factors but where the page is currently ranked and how many sites link to it(thus make it important) is data that is widely available.

I'm not talking about creating a method of ranking pages here, I'm talking about reporting where a page is ranked across a wide range of keywords as well as how linked to it is and how highly ranked those pages are etc to come up with a final value number.

NO calculations needed to determine the value of a page, this would be purely a reporting tool that the patent had protection against.

note: Pages could receive pagerank 5 by merely being linked to by a highly ranked page EVEN IF THEY WERE BLANK or a hello world post, and this was true for a logn time. It's a reporting tool, not a ranking tool.

If you're not seeing what I'm getting at, here's a visual.

- Brand new site, no links coming in, rank = 0
- a month later 3 new incoming links are found with ranks of 2,3 and 4. NEW rank for new site is 2(deja vu?)
- etc

This is a reporting tool, it does not do calculations besides figure out the value of pages based on values of other pages and reports it. Obviously if a page is #1 in serps the tool will find lots of incoming links and such and likely come up with a higher ranking... but it's not trying to do the 700+ or however many calculations Google did per page, pagerank never did that either. When a site is penalized it typically falls 50 or 500 spots so while we can't know why a page is penalized we can know it was and adjust rank accordingly.

None of those other systems by alexa and whatnot apply a pagerank-like value to Google's results, in part because of the legal issues surrounding their patents. I'm not saying revive pagerank, I'm just aware it could be recreated at this point, perhaps as an addon.

[keyplyr]

it was merely a reporting of where a page is ranked based on how many sites link to it and the weight those pages had

Part of what I'm saying is a PR report wouldn't apply the same way because all the ranking factors have changed, especially how linking (in all it's forms) is weighted.

If you remember, right after the PR green bar was frozen, then later removed from the Google tool bar and rendering all remote versions useless, Google enacted negative ranking factors to sites that had link exchanges, so most sites removed them... I did.

PR will never again reflect how Google ranks sites due to all the other factors that were added since then. Just my opinion :)

[Dimitri]

Didn't Google abandon the PageRank thing because it was patented at the time Google was still being developed at the Stanford Universtity, and so the University retained rights over it and Google didn't want to pay them anything ?

[martinibuster]

Didn't Google abandon the PageRank thing because...

No. PageRank is still in use.

Obviously when a patent exists you cannot duplicate the same methods but now that it has expired you certainly can.

I answered that in my first response. I will restate it in different words for your benefit:
What the toolbar showed was not PageRank. It was a simulation that had nothing to do with the calculations indicated in the patent. What was in the Toolbar was not a distillation of actual PageRank. Everything that Google ever handed out, from backlink data to allintitle searches were all crippled so as to be useless for backward engineering their algorithms. The Toolbar was similarly crippled.

it was merely a reporting of where a page is ranked based on how many sites link to it and the weight those pages had.

It was more than that. It was a calculation performed across the entire link graph, with some links counting less and some more, with extra doses of PageRank getting injected here and there to keep the flow moving. These details were outlined in general terms in the backrub paper (see section 2.1.1 of that paper) but it was also hedged with a statement that more work is needed and that is exactly what happened as successive versions improved on the original.

Good luck, I hope that makes things clearer for you.

;)
Roger

[tangor]

Even if the patent expires, and the underlying methods are now open to all for use in developing new products, the original "page rank" as described in the patent had/has serious flaws.

What would be essential for it to even work is a central data set (location and hardware to run it) of the entire web IN REAL TIME to have any value. It is possible to do the entire web on lesser systems over longer periods of time, but the data results would be that much less useful.

Lastly, getting the web to adopt a new "page rank" would be difficult. No existing search engine will integrate it as that would be in conflict of their own purposes. Yet, as an historical took of the web the way it was, it might be useful to play with to see how it stacks up against today's methods .... and might provide some insight in what the actual differences between then and now might be.