Forum Moderators: open

Message Too Old, No Replies

SERIOUS Google update algo analysis thread (Dominic)

NO whining or cheering about how your site is doing in this one.

         

rfgdxm1

6:21 am on May 6, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



This is a continuation of an idea for a thread that I started a few updates back. The topicality is listed below, and the expectation is that this thread will be restricted to just that. In the case of this Dominic update, GG has stated that other aspects of the update will be rolled out as the update develops. Thus, for this update it is possible that the observations made early on will not hold true by the end of the update. This is OK, because if patterns like this hold true for later updates, members here can use the search feature to find this thread and see how past updates developed.

----

I'm starting this thread because another member suggested such would be a good idea because the main Google update thread is cluttered with posts like "OMG, I've been dropped in the new index!" and "Yippee, I'm now #1 on a key SERP". This thread is ONLY for serious, generic discussion of changes that you are observing with the new algo in this update. As in things like "Looks to me like PR is less important this month, and anchor text of inbound links counts more.", etc. How your site is doing has no relevance here unless you can explain why you think so in terms of a general algo update.

mfishy

1:47 pm on May 6, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



SEO Guy,

This may or may not be the case but I am missing links from Yahoo and DMOZ that have shown in the past and are PR 6-7.

pixel_juice

1:47 pm on May 6, 2003 (gmt 0)

10+ Year Member



Very strange results checking PR from sj and fi - seems to be a pr calculation from 2 updates ago on sj and fi, or maybe this is just a consequence of the old link data. What puzzles me is that the PR and links do not seem to be from last month, but definitely seem to be at least one update older. Don't have access to my PR records at the moment so I could be wrong.

pixel_juice

1:50 pm on May 6, 2003 (gmt 0)

10+ Year Member



Anyone know what's going on with sj/fi. Which is the newer/more likely to reflect final index?

UK_Web_Guy

1:50 pm on May 6, 2003 (gmt 0)

10+ Year Member



Sorry if this question has been asked before, but how do you tell the PR as based from a specific datacentre , e.g. sj & fi

albert

1:57 pm on May 6, 2003 (gmt 0)

10+ Year Member



@athloninside:

I found interesting what GoogleGuy said here:
[webmasterworld.com ]

GoogleGuy msg #298:

albert, what you said, except I wouldn't be surprised to see SJ show up at other data centers first, and then to start applying the newer data/filters after that.

My message was #292.

mfishy

1:59 pm on May 6, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



My best guess is that fi will more likely be the index even though it certainly hasn't factored in all links yet

trillianjedi

2:00 pm on May 6, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Sorry if this question has been asked before, but how do you tell the PR as based from a specific datacentre , e.g. sj & fi

Change the hosts file - a search on here will bring up the relevant thread of about 1 hour ago.

TJ

HenryUK

2:03 pm on May 6, 2003 (gmt 0)

10+ Year Member



I agree with the theory that backlinks are being taken from a couple of months ago.

For one month only (for reasons not important here) my site was dropped from the index: no backlinks showing.

Prior to that it had plenty; on www now it has plenty. On 2 and 3 it has none. This is not a question of other sites being downgraded - we have links in from more than one PR7 site - they can't all have been downgraded to <4!

However, I am getting some very bizarre entries on 2/3/sj for domains that were 301'd at about the same time.

No explanation I am afraid (and no panic either) - just extra facts for analysers to digest.

globay

2:05 pm on May 6, 2003 (gmt 0)

10+ Year Member



I found pages on #1 spot for fairly competitive 2-keyword-terms, that did not have the keywords in the title, nor in the meta keywords, nor in the site description nor in h1 and h2 neither a high keyword density, only once first in the paragraph, and with keywokeyword2 in the url, missing part of the first keyword.

HenryUK

2:10 pm on May 6, 2003 (gmt 0)

10+ Year Member



Further info. On -fi I am page 1 for my main keyphrase, about 5 million results, despite showing 0 backlinks.

On -sj also no backlinks and can't find the site at all.

Not whining or whinging at all (it's very early days) but I suspect that each of these is showing a slightly different "iteration" of the update, which is clearly a different kind of update to those we have become used to.

Seems to me that -fi is possibly a step further on from -sj.

Of course it could all change again next time you look at either!

icebane

2:16 pm on May 6, 2003 (gmt 0)

10+ Year Member



My site has the same number of backlinks as 2 months ago (I believe)

Some of the backlinks are coming from sites that are haven't been online for at least a month, and those sites offline were not included in last months update as backlinks.

So this index of sites is 2 months old.

HOWEVER...

Some of the new pages I've added to my site for the last crawl are showing up in the google update. So... I'm thinking Google spidered new content in the crawl, but they haven't implemented new PR/backward link calculations yet.

taskmaster

2:22 pm on May 6, 2003 (gmt 0)

10+ Year Member



It could be that they have a new system for the backlinks, because we all have droped about 50% in backlinks and I have noticed that my backlinks has been droping for the last 3-4 month, without much change on the site, so maybe they have just completed the new filter. My site has not droped anyware even if the back links has gone from 500 to 83 now, so I dont think we will see more backlinks in this update.

zeus

salmo

2:34 pm on May 6, 2003 (gmt 0)

10+ Year Member



It seems that most of the top sites have been affected to the almost the same extent. Our site has lost many back links (still trying to figure out which ones and why) but so have our competitors. The result is that our rankings have remained much the same, in fact we have gone up one position from last month despite losing over 300 back links.

It is satisfying to know though that our closest competitor lost over 1000 back links which put them below us.

Spica

2:34 pm on May 6, 2003 (gmt 0)

10+ Year Member



Regarding the significance of the drastic changes in backlinks noted by everyone in the strange snapshot of the upcoming update that GG is letting us see, it might be worth considering the following:
links to dmoz are 204,000 on www, and 338,000 on -sj

mipapage

2:52 pm on May 6, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Uk-Web-Guy
here. [webmasterworld.com]

Welcome to WW, not need to be sorry, but don't forget to try the search function :] (I think I've seen you elsewhere...)

taxpod

2:57 pm on May 6, 2003 (gmt 0)

10+ Year Member



I won't argue about whether we are dancing or not even though GG replied to at least one comment indicating that the band is actually playing. I am seeing nearly constant flux on WWW, 2 & 3.

Some of my observations regarding the SJ data are that Zeal links are showing up on the "new-old" database. Also my Yahoo Group is showing up in the serps whereas previously it wasn't there. There are other odd pages included also.

The number of allinurl pages of two of my sites has nearly doubled after many months of consistently going up by a 1,000-10,000 per dance. This time 10s of thousands of pages were added. I can't account for when they were crawled as there are more than double the number of allinurl's as number of pages crawled in the past 30 days.

In my space I'm not seeing a drastic change in the serps for important keywords although I don't doubt the spammy results people are reporting. I just haven't seen it for myself. It looks like a normal update from that point of view.

ncsuk

2:58 pm on May 6, 2003 (gmt 0)

10+ Year Member



Maybe Google is experimenting with a new real time update that will only update www once it has completed the rounds?

Allergic

3:07 pm on May 6, 2003 (gmt 0)

10+ Year Member



I agree Ncsuk
I also found this last month :
Adaptive Methods for the Computation of PageRank
[dbpubs.stanford.edu:8090...]

Dominic based in part with this new algo?

swerve

3:09 pm on May 6, 2003 (gmt 0)

10+ Year Member



Here are some observations from one of my sites:

The site has been a PR3 for some time, hence "link:" searches always returned nothing. On -sj, "link:" now shows 5 links, so perhaps the site will become a PR4 after the update is complete:). My toolbar is pointed at 216.239.33.100 and is still showing P3.

My -sj backlinks include my DMOZ listing (which is not new), though the DMOZ listing just shows the URL (no title, snippets, or cache). I know this happens when a page has not been fetched by Googlebot -- but if that is the case, how does Google know that it links to my site? This may provide some support for some of the theories regarding new/old index, new/old backlinks, etc.

As far as rankings (on -sj) for my main single word query, the site has fallen 3 places (from 38 to 41), so little change there (so far).

stevegpan2

3:17 pm on May 6, 2003 (gmt 0)

10+ Year Member



taxpod
I see one of my site indexed pages from 2100 to 3000 pages.

the site sure has many more pages, but pr4 does not ge them all indexed.

need pr up

french tourist

3:23 pm on May 6, 2003 (gmt 0)

10+ Year Member



I am keeping on trying to understand what changed in the link: command

Here is a funny test (which proves that link: command and/or Google's understanding of redirections was modified -and which proves also that I don't understand well how it worked before).

Among the three URLs :
[fr.yahoo.com...]
[yahoo.fr...]
[yahoo.fr...]

the first one returns a document (HTTP code 200) while the other two return a temporary direction towards the first (HTTP code 302)

(I suppose I am allowed to post such general URLs? I swear I am not Yahoo's webmaster trying to make some self-promotion -))

On old Google, link:www.yahoo.fr gives the same answer that link:fr.yahoo.com while link:yahoo.fr only returns the few pages who explicitly contain this URL.

On www-sj and www-fi, link: to the three URLs give three different results.

(Edit reason : one small typo)

[edited by: french_tourist at 3:31 pm (utc) on May 6, 2003]

mrguy

3:29 pm on May 6, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I went through all the links looking for some ryme or reason and on several sites I work on and found several links that I removed prior to the last update.

So, for the sites I looked at, it is showing an old link structure.

Now, why would some sites show an old link structure while others would not?

This gets stranger by the minute.

I don't think we can properly analyze this update because I don't think it's the real deal. Those crafty guys at Google are playing with what they want to use for the real update before they apply it to their freshest material.

I think this is why GG remarked that current crawl and data would be filtered in after the fact.

I'm going to go about working on my site and wait and see what happens. Life is just to short to get all upset over a Google dance! (Besides, there is nothing I can do about it anyway until the dust settles. Then, I'll adjust and move forward just like everybody else here:)

pixel_juice

3:31 pm on May 6, 2003 (gmt 0)

10+ Year Member



I have one site that has been particularly problematic on Google. On sj the site shows up in the top 10 for a 2 keyword phrase. On fi it is way down in the results, but only for that particular phrase. Any other phrase the site results are almost identical on sj/fi.

To me this suggests fi is running some kind of filter/penalty that's missing from sj - but perhaps a keyword-specific penalty? On a second look it almost seems like it's a home page-only penalty.

carlr

3:34 pm on May 6, 2003 (gmt 0)

10+ Year Member



Just an idea here.

Since they rolled out the expired domains filter, maybe they are re in the process of recalculating the backlinks for all domains... This would explain the loss in backlinks and why they say it will be back...

stevegpan2

3:42 pm on May 6, 2003 (gmt 0)

10+ Year Member



compare www to www-sj
I see 50% more pages indexed on www-sj
but 50% fewer backlinks on www-sj

carlr

3:45 pm on May 6, 2003 (gmt 0)

10+ Year Member



Sorry for repeating myself - Browser problems

swerve

3:52 pm on May 6, 2003 (gmt 0)

10+ Year Member



Some of my observations regarding the SJ data are that Zeal links are showing up on the "new-old" database.

taxpod, can you clarify this? Are you seeing Zeal categories showing up as backlinks, or just showing up in the SERPS? (The former would be definitely new, while the latter would be normal)

meinereiner

4:05 pm on May 6, 2003 (gmt 0)

10+ Year Member



I can say, that the pagerank is definitely the one from 2 months ago (the one just before last googledance) - I checked this on 10+ domains and I have a 100% match with the PR data from 2 months ago.

For some sites this would be good, 'cause they dropped a bit in the last update *g* but for most not, they are down now on the "updated server".

Calculating this I agree: THIS is not the googledance, it's just funny google-techs playing with our nerves *lol*

abcdef

4:05 pm on May 6, 2003 (gmt 0)

10+ Year Member



Just a big thank you to the moderators, GG, and not leastly the posters to this forum. It is of great help to small companies like ours that cannot yet afford expensive advertising campaigns for premium advertising space in effective venues. Without Google listing results, not quite sure where we would be. And, this forum provides valuable information on what to do, what not to do, and insight on certain changes that are going to take place from time to time at Google.

Google results are akin to a Tax Audit. If they gave away the secret sauce, audits wouldn't be much good to the government? Google results would not be relevant to surfers, if Google gave away the sauce. But having Google representatives in certain places available to help out at times, is a big help. Google compared to the IRS,,, sorry....

Google is really good. We believe there is a conscience operating there that factors in that there are many sites listed in Google results that without Google results, they simply would not exist to the surfing world. That would spell doom for many, even newer, small business' like ours that have a large online base to advertise to, and are on very limited budgets.

Truth is, that Google listing results are still by far the most economical way to bring the online audience to your website. They also created a niche for those so inclined (crazy enough. haha) to earn a living helping other people enhance listing results, which is also a good thing.

We would just end off by saying that for those of you, or clients of those that use you for optimization pros-- save the profits the economic benefit Google results can bring.... Man does not live and die by Google results alone, or man will live and die as the Google index will see fit. Having a permanent presense on the net is very important, but traditional paid online and offline advertising has it's place as soon as you an afford to invest in it, and is a necessity for any business or idea serious about getting exposure.... as we are saving our dollars to reinvest the benefit Google results bring, into quality paid advertising campaigns to help us grow....

No, we are not an advertising company.... haha.

Thanks again to all!

NewcastleB

4:25 pm on May 6, 2003 (gmt 0)

10+ Year Member



Some of my observations (for what they are worth):

I have one site that gets hit by freshbot constantly. It hits hard and goes deep. SJ shows all the previously indexed pages as well as new pages.

Another site hardily ever gets hit by freshbot. When it does, it only hits the main page (with one exception I’ll get to in a second). This site only shows results from before I did an overhaul in December. None of these pages even exist anymore. Now to the freshbot exception. Starting in February, the freshbot started trying to access these phantom pages. This continued until late March. It didn’t hit new pages, just these old pages that I had removed from my server after the January update. They had not been linked to since the overhaul. Then it started up again just last week. The results are showing up without a cache, so they are probably from freshbots crawling 404’s.

One more comment. I recently redesigned a site for someone else. It went live just before the April crawl. They didn’t get any traffic so I went ahead and deleted the old pages. Now, in SJ, I see both the new pages and the deleted pages. But this time the deleted pages have a cache. They still lead to a “Page Cannot be Found”.

Make of this what you will. But it kind of looks to me like Google is searching for a way to incorporate fresh pages into the index better than they have before.

This 263 message thread spans 9 pages: 263