ResaleBroker: Well Google doesn't show all the backlinks you have. You need to go to Yahoo and do a linkdomain:www*evliving*com -inurl:www*evliving*com -inurl:evliving*com which gives you all your inbound links minus the internal links from your other 2 sites www*evliving*com and evliving.com . Yahoo reports 327 links and they seem to all have high page rank. 5,5,3,5,3,5,4,2,4,4 just in the first 10. Replace the asterisks with periods.
Sorry Compar, but your article is an oversimplification of the Pagerank algorithm. I have studied your article in detail and it omits a huge component of calculating Pagerank and that is that the algorithm must be iterated over many times before you reach a stable state of Pagerank values. I'll post again the link that I believe most clearly summarizes how Pagerank is calculated and if you study the article you will realize that a site's internal link strucure has a significant impact on the Pagerank of any page within the site. http://www.iprcom.com/papers/pagerank/index.html By only looking at incoming links to a page you can only measure the amount of Pagerank that is being sent to the page; not the amount of Pagerank that will be applied to the page after iterating over its own link structure.
Where in my article did I say otherwise? PR is the sum of all the pages that link to your page. If in turn the page in question links to pages that link back -- internal links and link structure -- then this will effect the final PR of the page. This is entirely consistent with my article. I never said or suggest otherwise. So one more time you are wrong. You haven't absorbed the entire message of my article. It is simple. If a page has a large number of outbound links, the amount of PR it passes is reduced. Therefore the PR passed to internal pages is reduced and as a result the PR that is passed back is reduced. There is nothing in my article that is inconsistent with that. Before you criticize my article try and understand it.
Virginia Realtors: You address two issues. 1) That the algorithm must be iterated over many times before a stable state of PR is gotten, (This is a commonly known fact to any one that knows anything about the ragerank algorithm) and 2) Only looking at incoming links to the page you can only measure the amount of PR being send to the page, not the amount of PR that a page will have after being iterated over many time in its own link structure. Compar states above that his article is not inconsistent with your comments in point #2, and having read it over one more time I would agree. He didn't address those issues, but this article is not inconsistent with those issues. His chart shows the PR value of inbound links and didn't discuss the effects of outbound links or the structure of internal links. I would have to say that again like the required alogorithm iterateds to get a stable state, that any one that knows anything about ragerank is fully aware of the fact that outbound links and internal link structure effect a pages PR and not just the inbound links. So considering the depth of understanding the article displays the writer has, and how elementary that issue is (that outbound link and internet structure effects PR), to think that the writer was/is not aware that outbound links and the structure of the internet links effects the PR after writing an article like that, tells me that you many not be relizing how elementary those issues are. As for your your #1 issue I didn't see where Compar's article or calculation notes how many times the algorithm is being iterated (it is generally accepted that 20 to 40 tmes is enought for simple calculations before they stablizes). Also Compar didn't address that issue in his return post to you. Compar: Could you inform us the number of times the algorithm was iterated in your table calculations.
Nobody know how many time the calculation is done, or how many iterations Google makes before declaring a PR value for a page. And my chart doesn't try to determine that. What I present is the probably value, and range of values, for various PRs after all then iterations are done. Certainly the internal structure of a web site will impinge on the final PR of each page. However a detailed analysis of internal linking is another subject. And as you suggest I did not attempt to cover it. That said however, the values that I present in the chart are still valid. The value of a page may be higher or lower due to internal linking, but my ranges are still valid no matter how they are arrived at.
what are ? PRx? TOL? Nelson Homer www.guide-for-beginners.com www.ehanapbahay.com www.pinoyoutsource.com
Google Information for Webmasters Design and Content Guidelines "Offer a site map to your users with links that point to the important parts of your site. If the site map is larger than 100 or so links, you may want to break the site map into separate pages." "Keep the links on a given page to a reasonable number (fewer than 100)"
As I said in my post, you're article is misleading because it leaves out a crucial element of the Pagerank algo which is that it must be iterated over many times to reach a stable state. Because of this omission, in my opinion. your article and especially your Pagerank Calculation Chart is misleading. By its omission, it gives the impression that by simply looking at the static PR values of the pages with links to your page, you can determine the PR of your page. If you understand that the equation must be iterated over many times, then you understand this is not the case. From your article: You can argue that you never said NOT to iterate over the equation many times, but you do not even include a discussion of it. Your chart showing number of links from a page with a certain PR to achieve a PR value leads a novice to believe if they get these incoming links from another site, they should achieve the PR you have listed. The reality is, it will also depend on the link structure of their own site. I don't think I attacked compar's knowledge of the Pagerank algorithm. What I am saying is that his article has a serious omission and by this omission presents an overly simplified view of Pagerank that could mislead people not already intimately familiar with the algo. He has posted his article many times to help people understand Pagerank. I think the article I posted gives a much better explanation of Pagerank. While you seem to believe the iteration is "elementary" - I would argue that most people just starting to understand Pagerank do not understand it.
ResaleBroker: Yes it works very well if Yahoo has your site indexed. nelsonhomer: TOL is is noted on the first page of this thread, "Total Outbound Links (TOL)." Total outbound links is the total number of links from a page that are outbound vss inbound (links that are made to a page.) In PRx the "PR" is a short form for Page Rank and the "x" is between 0-10. PR(x-1) just mean what ever value is asigned to X minus 1 etc. compar: Yes I kind of got to thinking and realized that later that you probley didn't have to iterate the algorithm as you were just providing a statement of relationships. I agree we don't know how many times the calculation is done by Google on the internet. We do know that for simple calculations the numbers settle down after about 20 to 40 interates. If we take 0.15 + 0.85(x) and put in any number from 0 to 20 it will move to ~1 within 40 interates. But yes the internet is a different story.
I agree most people do not know about the iterations required to determine PR. But it has nothing to do with my chart. PR is not declared until after all the iterations are run. There is no PR before that. My chart shows the relationship when PR is reported, and after all the iteration are run. It then represents the probable number and value of all the pages that have been included in the calculation, which has resulted from many iterations. You can't have PR without iterations. And it doesn't matter if anyone knows about iterations or not. When the PR is calculated my chart shows the probable relationships and values between all the pages. I have tested this against actual results and as I've said already if I change the average links per page assumption to 40 I get a very strong correlation with the real life pages that I have tested. If the chart correlates with real life pages I fail to see how it can be too simplistic.
We'll have to agree to disagree on that point. Many times people will ask questions like: "If I get a link to my homepage from a page that is PR7 and has 50 links on it ... what will the PR of my page be?" Using your data, the median PR passed would be 8,291. Your article leaves the reader with the impression that is all there is to it. Based on that link (let's assume it is the ONLY incoming link to keep things simple), the PR of the page being linked to is 8,291 --- a PR5 using your tables. BUT - depending on how the site is setup, the actual PR retained by that page could be as low as 8,291 * .15 = 1,243 --- a low PR4 using your tables. Unless you understand that the PR calculation is recursive, you will miss this. For very large sites where the site itself generates significant PR, this point becomes even more signifcant. Perhaps your explanation/article is the best way to introduce people to PR ... my $.02 ... it would be a more complete article if you make some mention of the fact the PR calculation is recursive and that internal site structure will have a large impact on the retained PR of the page being linked to. Again - I don't mean to attack you personally... just saying I think the article is misleading in this area.
Virginia Realtors: I didn't mean to infer you were attacking compar's knowledge, but it seems clear that you are infering that compar doesn't understand how internal structure effects page rank and the important of algorithm interates by your first 2 posts: "Take a thorough read of this article and study the example and you will see understand why the strucuture of your site has such a huge impact on the PR." "I'll post again the link that I believe most clearly summarizes how Pagerank is calculated and if you study the article you will realize that a site's internal link strucure has a significant impact on the Pagerank of any page within the site." My comments did point out that to someone that is in the know on pagerank these two issues are elementary, and to say to some one that comes up with a good article like compar does, that they need to read up on algorithm interates and that outbound links and internal linking structure effects PR, could be taken as an insult to their intelligence. I am sure you didn't mean it that. Further I noted that the fact you don't seem to relize how elementary these two issues are to those that are in the know, could be seen as implying that you many not be as far along as you think you are. I trust that is not the case, and I certianly don't mean that as an attack, just an observation. Again the two issues are elementary to those who know what they are talking about when it comes to pagerank, I was not refering to those issues being elementary to those just starting to understand pagerank. As compar has noted he doen't deal with those 2 factures in his article. He does touch on internal link structure in his following article PageRank & How to Get It but doesn't go into algorithm interates in either of them. Personally I am not sure you have to get the formula out and try to get people to undersand all that right off the bat. The formula can be overwelming for those that are just being introduced to pagerank. Yes there is a place to go over the formula, and yes understanding the formula is important as is understanding algorithm interates and how internal linking structure effects PR. I think what compar's article does is, without getting into forumals tells a person in general, hey, you need 18 PR5 coming into a page to get a PR5. I do agree that perhaps his article could be improved by noting that the chart doesn't take into consideration internal link structure or outbound links. You many want to suggest that to compar instead of making posts that seem to imply he is lacking elementary understanding on the subject of his article. I have already made changes in my Page Rank Calculation article in responce to your comments. I took your comments as helpful, made changes, and have overlooked the way you brought them out. I agree that going into the pagerank formula will help some one that can grasp it, and is ready for that, but for some one that don't know anything about pagerank and is just wanting to get a rough idea of what kind of links they need to get a certian PR, I think compar's article is helpful.
For many years Brain Surgeons would not make changes in surgery techniques thinking that entry from the top of the skull was too much of a risk for the patient. That was until one surgeon proved that it could produce good results without danger to the patient. When he proved this, all of the surgeons accepted the practice themselves. If you have a valid point Bob is wise enough to accept it and improve his study also. I am sure he welcomes the new information if it is valid.
I do not see either of you standing still, if you can improve yourself or your content, you or Compar will both do that Bill Gates was wrong about the internet, but he got on board when he found out that he was missing the boat. So, Bob and Bob are smart enough that when they find a flaw in the theory published, it will be corrected or adjusted also. Hey Bob might even give credit to the guy that helped him find the flaw, Compar is a good guy after it is all said and done. Even if some of us try to make him look like a bad guy in jest
Can't we all just be friends? I certainly didn't mean to offend anyone by any of my comments and I apologize if my approach/langauage isn't as sensitive as it should be. I will say what I believe to be obvious ... anyone that takes the time to put together tools/articles or simply to respond to posts here to help others is obviously contributing positively to the community and should be applauded... so cheers to bobmultch and to compar for taking the time. My feedback for you both is to consider that a wide variety of people with very varied degrees of insight into Pagerank will be accessing your resources. I think if you are going to present a simplified view of Pagerank it should be with at least a mention or reference to the facets that have been omitted.
compar: Lets say I got 18 inbound PR5's links of medium value to the home page on a 10 page site that has a fully meshed internal link structure. Your chart shows that 18 PR5 Links to a "page" will give it PR5. Is your chart take into consideration the PR that home page will pass on to the other 9 pages and how that will effect the end PR of the home page. Or is your chart just take into consideration 18 Link of PR to one page and the effect it will have on that one page with out considering whether that page is linked to other pages or not. If your chart is only taking into consideration one page by its self with 18 inbound links then the person trying to apply this to there site must take into consideration how much of the PR will be transfered to the other pages that are in the site for the chart values to be meaningful. If this is the case I think the chart would be more meanful if it would note the value of the receiving page in for a 10 page site with a fully meshed internal linking structure and no outbound links. While you can't have a chart that would be adaptable to every sitituation it would be best to pick a site structure that is pretty common like a 10 page fully meshed site. I am not saying your calculations are off. I am just questioning to what can they be applied to.
Virginia Resultors: Well you didn't offend me all it did was make me think. Take a peek at the 4th paragraph I added to my article and let me know if that made things a bit clearer. If you have further suggests to make it clearly I am very open to them. You may even want to suggest how compar's article could be make clearer. I do now think I am seeing that the numbers in the chart probably just apply to a simple page with 18 links. If this is the case I am thinking that the readers of the article would be better serviced if the number applied to say a 6 page site that has a fully meshed internal link structure (no outbound links), or even better the ability to be able to put in the number of fully meshed pages they have. If you have 18 PR5 links coming into your home page, there is a difference in the PR retained on the home page and the PR passed to the connected pages depending on the number of fully meshed pages in the site with no outbound links. The smaller the site the more PR the home page retains. Most smaller sites link structures are fully meshed via the menu. It would be hard to make a model that would work for every site but one that would allow you to choose the size of your fully meshed site would be a starting place.