 Anthonyparsons.com Test, Placement Of Code Test Results |
| Important Announcement: ***Need an Affordable SEO Website Review?*** |
High Rankings Search Engine Optimization Forum > Search Engine Optimization > Search Engine Friendly Design & Usability  |
 Mar 31 2004, 01:06 AM
Post
#1
|
|
 33 & Retired ... What to do, What to do?  Group: Active Members Posts: 653 Joined: 13-October 03 User's local time: Feb 10 2010, 04:59 AM From: Nth Qld - Australia Member No.: 1,052  |
Anthony Parsons vs. Jill Whalen Test In Progress
This test was conducted from another two posts here at High Rankings Forum. You can read each for all the introduction: Bullet Point vs. Informative Text & Keywords Text On Top Of Page Threads. Reading Background Information, Viewing Pages & Complete Code Is Essential Results test1.htm Code Layout (Partial Code) QUOTE <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <title>Anthony Parsons vs. Jill Whalen Test In Progress</title> <meta NAME="description" CONTENT="Anthony Parsons vs. Jill Whalen Test In Progress"> <meta NAME="keywords" CONTENT="Anthony Parsons vs. Jill Whalen Test In Progress"> <meta NAME="MSSmartTagsPreventParsing" CONTENT="TRUE"> <link REL="stylesheet" TYPE="text/css" HREF="http://www.anthonyparsons.com/style_sheets/seo.css"> </head> <body> <table CLASS="tab1" SUMMARY=""> <tr> <td> </td> <td CLASS="nav" WIDTH="760"> <img BORDER="0" SRC="../images/blank_01.jpg" ALT="search engine optimisation" width="760" height="29"><br> <img BORDER="0" SRC="../images/blank_02.jpg" ALT="search engine optimisation" width="760" height="43"><br> <img BORDER="0" SRC="../images/blank_03.jpg" ALT="search engine optimisation" width="760" height="29"></td> <td> </td></tr></table> <table CLASS="tab3" SUMMARY=""> <tr> <td ROWSPAN="3"> </td> <td WIDTH="760"> <h1>Anthony Parsons vs. Jill Whalen Test In Progress<br> <span class="text1">A Relative Test For Text Positioning</span></h1> test2.htm Code Layout (Partial Code) QUOTE <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <title>Anthony Parsons vs. Jill Whalen Test In Progress</title> <meta NAME="description" CONTENT="Anthony Parsons vs. Jill Whalen Test In Progress"> <meta NAME="keywords" CONTENT="Anthony Parsons vs. Jill Whalen Test In Progress"> <meta NAME="MSSmartTagsPreventParsing" CONTENT="TRUE"> <link REL="stylesheet" TYPE="text/css" HREF="http://www.anthonyparsons.com/style_sheets/seo.css"> </head> <body> <table CLASS="tab3" SUMMARY=""> <tr> <td ROWSPAN="3"> </td> <td WIDTH="760"> <h1>Anthony Parsons vs. Jill Whalen Test In Progress<br> <span class="text1">A Relative Test For Text Positioning</span></h1> When searched for "Anthony Parsons vs. Jill Whalen Test In Progress", being the page heading, test2.htm is ranked and not test1.htm? Yes, we all know that Google is not going to keep identical content. So why did it keep the page that contained the text closer to the top of the code than the other? All that know about this test have already assessed the test, so; does Google favour the code closer to the top than further down the code??? Is it a benefit to absolute position your tables through CSS to ensure the code is closer to the top but to the eye the page looks the same? |
 |
 
|
|
Mar 31 2004, 01:22 AM
Post
#2
|
|
 Peanut Butter Lover Group: Active Members Posts: 502 Joined: 6-August 03 User's local time: Feb 9 2010, 02:59 PM Member No.: 319 |
I would say that test is not exceptionally conclusive as far as positioning goes. I think the amount of content and coding is more important (especially in that small of an example). The second example had way less coding and a higher keyword density than the first
|
|
|
|
|
Mar 31 2004, 01:57 AM
Post
#3
|
|
|
33 & Retired ... What to do, What to do? Group: Active Members Posts: 653 Joined: 13-October 03 User's local time: Feb 10 2010, 04:59 AM From: Nth Qld - Australia Member No.: 1,052 |
Awall
QUOTE This test was conducted from another two posts here at High Rankings Forum. You can read each for all the introduction: Bullet Point vs. Informative Text & Keywords Text On Top Of Page Threads. You may want to read all the background information first and view the tests? View the actual full code possibly? I am sure that heaps of people will attempt to post at this without even reading the background information. The background information is essential. [Edited HUGE text. - Jill] This post has been edited by Jill: Mar 31 2004, 08:37 AM |
|
|
|
|
Mar 31 2004, 02:04 AM
Post
#4
|
|
 HR 5 Group: Active Members Posts: 330 Joined: 16-January 04 User's local time: Feb 9 2010, 02:59 PM From: kansas city, Kansas USA Member No.: 1,945 |
tables shouldn't be used for css.....divs only...tables are for dat sets and data sets only.
|
|
|
|
|
Mar 31 2004, 07:06 AM
Post
#5
|
|
|
33 & Retired ... What to do, What to do? Group: Active Members Posts: 653 Joined: 13-October 03 User's local time: Feb 10 2010, 04:59 AM From: Nth Qld - Australia Member No.: 1,052 |
ghergich,
Tables are not used for CSS, rather the tables are positioned through CSS only. This enables the code with the text to be placed as close as possible to the top of the <body> tag. You must read the background to have a complete understanding of what has been tested. |
|
|
|
|
Mar 31 2004, 10:46 AM
Post
#6
|
|
 HR 3 Group: Active Members Posts: 50 Joined: 26-March 04 User's local time: Feb 9 2010, 02:59 PM From: Northwest Indiana Member No.: 3,048 |
QUOTE All that know about this test have already assessed the test, so; does Google favour the code closer to the top than further down the code??? Is it a benefit to absolute position your tables through CSS to ensure the code is closer to the top but to the eye the page looks the same? I think it would be important to implement the test again on a different host to see if Google would again favor the code closer to the top. As I'm moving more and more to tableless design and absolute CSS position, it's definitely important to know if the content should be first and the positioned items later. Thanks for posting this! (IMG:http://www.highrankings.com/forum/style_emoticons/default/appl.gif) |
|
|
|
|
Mar 31 2004, 10:55 AM
Post
#7
|
|
 HR 6 Group: Active Members Posts: 786 Joined: 20-August 03 User's local time: Feb 9 2010, 02:59 PM Member No.: 583 |
One important thing this test doesn't consider is that Google prefers a page where the keywords are dispersed evenly throughout the whole page over a page where the keywords appear only near the top (or bottom, for that matter). Keyword spread is problably more important on a content rich page than either of the things being explored here.
G. |
|
|
|
|
Mar 31 2004, 04:33 PM
Post
#8
|
|
|
33 & Retired ... What to do, What to do? Group: Active Members Posts: 653 Joined: 13-October 03 User's local time: Feb 10 2010, 04:59 AM From: Nth Qld - Australia Member No.: 1,052 |
Absolutely Grumpus, but keyword positioning are not what are being tested here. Both pages are identical in text. So keywords are not the issue.
What I am going to do is remove the pages from the server, get rid of them from Google, change the file names over and see if Google takes up and holds the same file with the different file name. That should substantiate whether it likes the code closer to the top or not? The pages are identical in code, with the exception that one has the header table before the text where the other is after the text. Same code, just moved in position. That is where the absolute positioning takes affect through the CSS. If I change the file names now, that should pretty much well rid all doubts. |
|
|
|
|
Apr 1 2004, 09:32 PM
Post
#9
|
|
|
HR 3 Group: Active Members Posts: 50 Joined: 26-March 04 User's local time: Feb 9 2010, 02:59 PM From: Northwest Indiana Member No.: 3,048 |
QUOTE What I am going to do is remove the pages from the server, get rid of them from Google, change the file names over and see if Google takes up and holds the same file with the different file name. That should substantiate whether it likes the code closer to the top or not? I'll be anxiously awaiting the test results (IMG:http://www.highrankings.com/forum/style_emoticons/default/appl.gif) I took one of the pages in my website, and switched my code so that the menu is now below the content (tableless layout with absolute positioning). I'll be checking to see what happens with this one page versus the other pages. (IMG:http://www.highrankings.com/forum/style_emoticons/default/cheers.gif) |
|
|
|
|
May 2 2004, 12:36 AM
Post
#10
|
|
|
33 & Retired ... What to do, What to do? Group: Active Members Posts: 653 Joined: 13-October 03 User's local time: Feb 10 2010, 04:59 AM From: Nth Qld - Australia Member No.: 1,052 |
Ok then, test results finalized.
After swapping deleting the files until Google dropped the file, then renaming the folder from "test" to "test1" just so Google had somewhere new to go, then swapping the links around so "test1" then went to test2.htm and "test2" then went to test1.htm, Google still indexed test2.htm containing the text closer to the top of the code. I think it can well and truly be stated that Google favours text closer to the top of the page within the code, not how some think about it being closer to the top in the page as we view it (which we know is rubbish). So there you have it, Google prefers the text closer to the top in the code. CSS positioning to get your text up top is viable. Even though this is only one of the over hundred or so assessed features, is it another important factor to be considered IMO. Good news for those who already write using complete CSS design and layout. |
|
|
|
|
May 26 2004, 01:43 PM
Post
#11
|
|
 HR 6  Group: Moderator Posts: 918 Joined: 24-July 03 User's local time: Feb 9 2010, 01:59 PM From: Michigan USA Member No.: 17 |
Anthony, the only thing your tests indicate is that you don't know a whole lot about controlled testing. But even ignoring your methodology, your conclusions can only be appropriately applied to duplicate content. Were you trying to determine the criteria Google uses to select which pages with identical content to index? If so, how you linked to the test pages is probably more important to document and understand than is the amount of code in each page. Does Google choose a page, as many have contended, based on PR? Does it choose the first link it finds? Or does the last link it finds always overwrite previous links? There are probably a score of important factors you have neither documented nor controlled. I rather strongly suspect you're not even testing what you intended to test.
Conclusions backed only by faulty tests are worse than no conclusions because they too easily and quickly become accepted as Truths. There are enough [url=http://www.highrankings.com/forum/index.php?showtopic=15499]Seo Myths[/url] out there already, without any need to build and perpetuate more. It has become increasingly difficult over the years to devise meaningful ranking tests, and it's likely there's no longer a perfect methodology possible. That doesn't mean we shouldn't try to control as much as possible when testing, and it certainly doesn't mean we should accept dubious results without question. |
|
|
|
|
May 26 2004, 02:35 PM
Post
#12
|
|
|
HR 3 Group: Active Members Posts: 102 Joined: 20-September 03 User's local time: Feb 9 2010, 02:59 PM From: Ogden, Utah, USA Member No.: 856 |
Ron, ouch.
Granted, the poster probably should have refrained from using words like 'well and truly be stated', but that doesn't mean his results aren't interesting. You're right, there are unknowns and variables that have not been accounted for. Your mention of link order or link preference is definetly one of those unknowns. However, I think he's done a nice job of controlling most of the experiment. And, I think it's good enough in that it justifies further research. See, this is how science is. You do a quick experiment to get you half the way there, then you tighten the screws a bit and continue the experimental process. Eventually you account for all the unknowns and you've run the test enough that a certain hypothesis can be considered proven. Of course, it's up to each individual to decide how much evidence it takes (IMG:http://www.highrankings.com/forum/style_emoticons/default/wink.gif) . Look the point is, don't blast the guy for trying something out. It should be encouraged, in my mind, and I hope he continues his study trying to account for the other unknowns. Hopefully he'll learn to use more appropriate phrases as he describes his work, but you can't blame the guy for being over zealous. |
|
|
|
|
May 26 2004, 05:39 PM
Post
#13
|
|
 High Rankings Advisor  Group: Admin Posts: 29,201 Joined: 21-July 03 User's local time: Feb 9 2010, 01:59 PM From: Ashland, MA Member No.: 2 |
All that was shown that when there is duplicate content, Google tends to choose the page that has the content close to the top. How does that show that Google ranks said page higher than another with text lower?
|
|
|
|
|
May 26 2004, 07:31 PM
Post
#14
|
|
|
33 & Retired ... What to do, What to do? Group: Active Members Posts: 653 Joined: 13-October 03 User's local time: Feb 10 2010, 04:59 AM From: Nth Qld - Australia Member No.: 1,052 |
Ron,
Thanks for the criticism and taken onboard, however; the way in which the pages where linked if you read the entire process correctly; the links where swapped and exactly as I stated, once Google dropped the pages, and the link order was changed, the same page was once again kept in a difference of order being fed to the engine. What it demonstrates is fairly much well as Jill has said. I never said it ranks better, but that Google Likes Text Up Top In Code. If you didn't know it before, then you know something now about Google that you didn't previously. What impact does this have? Well for starters, people who use complete CSS positioning to design! Are they ahead of us just by the method of design? If Google favours a page that has the text up the top in the code, then it is very possible that Google may use that as another assessed technique or comparison in their algorithm, which means, if one page vs. the other are locked and you simply cannot get over the top of them, that could be a deciding factor. Their is nothing definate about any technique we use because unless Google actually tell us, we are only guessing. Even testing something does not make it factual, because other factors unbeknown to us may be playing some role in it. Google have said openly that its algorithms use billions of factors within its assessments! Billions. That's a lot of factors that none of us are ever going to guess? I don't personally think that something you didn't know before is going to hurt anyone. Is anyone going to get carried away with it? Probably. Do people get carried away with stacking keywords in pages? Most definately. Why? Because one of us professionals or experts such as Jill made a comment that the search engines require or prefer keywords in the page or particular area of the page. What people do with that info and how they use it is up to them. But at the end of the day, if you're trying to tell me that the test conclusions are not complete, then why don't you test it for yourself and come back and tell us all. Simple solution to the problem Ron. QUOTE Conclusions backed only by faulty tests are worse than no conclusions because they too easily and quickly become accepted as Truths. There are enough [url=http://www.highrankings.com/forum/index.php?showtopic=15499]Seo Myths[/url] out there already, without any need to build and perpetuate more. I think you're all confused mate. If you read everything, and not just pieces, examined the pages as did most previously, then you may have a different opinion. The other thing is, is that the entire process was developed through this forum. All aspects of the test where discussed before implementing. I suggest another look possibly for things you may off missed. QUOTE anthonyparsons.com
then swapping the links around so "test1" then went to test2.htm and "test2" then went to test1.htm This post has been edited by anthonyparsons.com: May 26 2004, 08:09 PM |
|
|
|
|
May 26 2004, 10:17 PM
Post
#15
|
|
|
HR 6 Group: Moderator Posts: 918 Joined: 24-July 03 User's local time: Feb 9 2010, 01:59 PM From: Michigan USA Member No.: 17 |
Adam, my intent certainly wasn't to "blast" Anthony (and I think/hope he realizes that). Rather, I wanted to forestall yet another potential misconception by visitors perhaps too new to the game to realize that saying something forcefully enough doesn't necessarily make it so. I agree Anthony has a decent start, but his post made it sound very much like the end of research instead of the beginnings.
Anthony, in the writing field, your statement that, "Google likes text up top in code," is called personification. It can be useful for communication, but it can also be dangerous when taken too far. A program doesn't have taste buds or personal likes. It just responds to stimuli and, then, only in the circumstances for which it's been programmed. The ONLY circumstances you've explored at all are for duplicate content, and I'm honestly not convinced even there that your conclusions are merited by your results. What other factors might influence Google to index one dup page over another? Until you eliminate those factors as being potentially determinant, any conclusions are premature. |
|
|
|
|
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:
This forum is sponsored by High Rankings, a Boston SEO Agency
|
Lo-Fi Version | Time is now: 9th February 2010 - 01:59 PM |