I have created a simple PageRank calculator. Just put in the number and type of incoming links for the page in question and it will estimate pretty accurately (but not exactly) the expected Google PageRank.
Yeah but what if that page is from DMOZ with two hundred unique URL's and yours in one of them. Then ( .85 * PR ) / 200 would be passed to you wouldn't it (forgetting about possible 50 link max, order etc. for now)? That can only be taken into account in this tool if there is a second column where you can enter how many (unique) links there are on that page that links to you. That assumption should be mentioned at least on that tool's page IMO.
Nice tool. I do agree with Tops30 that some disclaimer (or other set of fields for link counts, etc.) should be referenced on the page. This is especially true for those less knowledgeable about the formula. IMHO.
Agreed. I am currently working on an advanced portion however. Hopefully a lot more variables will be present. Any additions that you would like to see?
You could make it so that you can store the data and just fill in the url of the page that links to you. Then your script can count the links automatically, check the page's PR and calculate everything for the user. Just store the url's and your url in a cookie or a quick log in. Then you really have a kick arse tool for those who are interested in PR and want to predict the impact of those links they are planning on buying...
Doesn't a Google search for sites that link to you provide only a subset of all possible sites? I would also have to do something where you would enter in your Google API, just like what DigitalPoint does. These are all good changes. I'll probably phase them in slowly.
No I meant without the API. I once enter my page x.com/mypage.php Then list all my rerrering pages y.com/page1.htm & z.com/page3.html You script reads in the pages, counts the links and the number linking to my site x.com. Then it uses the PR checksum to determine each linking page's PR and there you have it... A fairly accurate estimate of my page's PR!
There's definitely a lot that can be done. An "Advanced PageRank Calculator" page has been added. Currently the only additional variable is number of links per page. More to come...
I only wish that my pages had a PR7 like your tool estimates, instead of the PR5 they currently have.
Are these backlink pages filled with more than 50 links? Having more links on pages should decrease the value of each backlink coming from that page.
I tried it using the PR6 field alone. When I typed "10", calculated result is PR7. But when I typed "9", result is PR5. Only one BL difference will cause a 2 increment drop in PR? Thanks.
But if you only wait a couple of weeks Google gives you this information without any typing or effort on your part. I think what ThinkBlink is trying to do is provide a supplement to the chart I published so that you can do "what ifs" with the assumption of links from various PR bearing pages. Since the possible PR passed by a page has such a large range of potential values, depending both on the number of links from the page, and where the page sits in the total range of values for each of the PR increments -- see my chart -- exact predictions are never going to be possible. All that can be done, at best, is to offer a type of order of magnitude prediction based on some reasonable set of assumptions. Now the assumption that the average value of all pages is the median value of their range is totally valid. But if you are trying to predict the PR contribution from a discrete set, or subset, of pages then you have no idea where they may sit in the range. They may be normally distributed across the entire range but they may also be at either extreme. So once again absolute accuracy is impossible for the individual case.
I looked back at my equations and it was not calculating the PR 7s correctly. You should need about 18 or 19 PageRank 6 backlinks to get a PR 6. Thanks for discovering the bug!
I have updated the Page Rank Calculator. There is only one view now (Not an advanced and a basic). All the functionality is in this one view now and everybody should agree that it looks much, much better. Negative comments are encouraged because it sculps a better app.