homepage Welcome to WebmasterWorld Guest from 54.221.175.46
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Visit PubCon.com
Home / Forums Index / Google / Google News Archive
Forum Library, Charter, Moderator: open

Google News Archive Forum

This 41 message thread spans 2 pages: 41 ( [1] 2 > >     
Parked Domain Ranks #2 of 26,800,000 Results for Single Keyword
Google Algo and Expired Domains
nativenewyorker




msg:38886
 10:37 pm on May 9, 2003 (gmt 0)

Despite being an expired domain (former discount broker) showing 0 backlinks, the #2 result on main Google server for single keyword search "discount" is a generic Network Solutions page. Shouldn't the algo have spit this page out already? As a user, this is an awful search result with 0 relevance.

Google serp shows:
Web Page Under Construction
Network Solutions - Original domain name registration and reservation services with variety of internet-related business offerings. Quick, dependable and ...

And it still ranks #4 on -sj server.

Ted

 

ariff44




msg:38887
 10:41 pm on May 9, 2003 (gmt 0)

that definitely is poor but perhaps Google will hit it in next update...when did it expire?

nativenewyorker




msg:38888
 10:45 pm on May 9, 2003 (gmt 0)

I'm not sure when the site was taken down, but Ameritrade bought National Discount Brokers two years ago.

Ted

Chris_D




msg:38889
 12:17 am on May 10, 2003 (gmt 0)

Hi Ted,

That is weird - really weird - because the page you mentioned has PR0, Google says it isn't ranked and has no backlinks, and the Google cache says that the search term only apprears in links to the page?.

Someone else nailed it a few hours ago - there is no update - we've all been chasing shadows.

The only explanation for a result like that in www. is that something is broken, or half baked

Chris_D

rfgdxm1




msg:38890
 1:40 am on May 10, 2003 (gmt 0)

>The only explanation for a result like that in www. is that something is broken, or half baked

We can hope half-baked, but my guess is Google is just seriously broken. Time will tell. Check Alltheweb. ndb.com has gobs of backlinks. This is why they are ranking so high. Google is just doing such a rotten job they can't spot a standard holding page like this.

1milehgh80210




msg:38891
 1:50 am on May 10, 2003 (gmt 0)

I figured it out!
Results are now ranked ,least-to most-relevant!
:)

BigDave




msg:38892
 1:57 am on May 10, 2003 (gmt 0)

Good find, you should report it to google.

But I have to take issue with "As a user, this is an awful search result with 0 relevance." as that is a single word search that will not meet the needs of just about any searcher on that word.

Let's do a search on some other verbs next. "run" does not return useful information at all when I want to run my model trains. That find does not damage the SERPs.

grifter




msg:38893
 2:05 am on May 10, 2003 (gmt 0)

Google is just doing such a rotten job they can't spot a standard holding page like this.

Frankly, I think the query smells a little worse than the result. Who searches for just "discount?"

1milehgh80210




msg:38894
 2:14 am on May 10, 2003 (gmt 0)

If this site ranks #2 for "discount", would'nt that have a bearing on other searchs containing that word.I seem to have noticed that effect anyway...0h well

BigDave




msg:38895
 2:22 am on May 10, 2003 (gmt 0)

That is hypothetical. Find some.

1milehgh80210




msg:38896
 2:30 am on May 10, 2003 (gmt 0)

discount broker #3 and #4 a little shaky?
I do think G is a great SE overall!

rfgdxm1




msg:38897
 2:35 am on May 10, 2003 (gmt 0)

>That is hypothetical. Find some.

[www-sj.google.com...]

Take a look at #2 also:

Index of /donaldj
Index of /donaldj. Name Last modified Size Description
Parent Directory 02-May-2003 16:21 -

Google sure seems to be sucking really bad to me. More evidence of how widespread lousy Google SERPs are.

1milehgh80210




msg:38898
 2:39 am on May 10, 2003 (gmt 0)

I was using (www.)rfgdxm.
But I would have to disagree with you----the right side of the page has very relevant listings!

rfgdxm1




msg:38899
 2:47 am on May 10, 2003 (gmt 0)

>But I would have to disagree with you----the right side of the page has very relevant listings!

"Though this be madness, yet there is method in't.". - (Hamlet, Act II, Scene II).

One might wonder if Google broke the algo on purpose to increase profits?

"Never attribute to malice that which can be adequately explained by stupidity."
-Hanlon's razor

Then again, it could be just that Google doesn't know what they are doing. ;)

CCowboy




msg:38900
 3:15 am on May 10, 2003 (gmt 0)

Off topic, but I wish Verisign would play by the Stickin Rules!

Google Guy... can we all feel out fake or maybe not so fake spamm reports on them?

BikeMan




msg:38901
 3:57 am on May 10, 2003 (gmt 0)

Well there is a site returning between 4 and 6 for a two word combination for the last few updates inspite of not being in operation for the last two years.

Regarding the discount site this is only due to anchor text in the backlinks which have not been updated. If the backlinks are not showing they must be below PR4. Check alltheweb and you will get all the backlinks for that site

1milehgh80210




msg:38902
 4:17 am on May 10, 2003 (gmt 0)

The (questionable?) site is showing @ #4 shows 455 backlinks on allthe web.
Site #18 shows 15,001 links.
Site #11 shows over 2,000 links
etc. etc.

Are #4s links that good? For a site that is'nt working?

nativenewyorker




msg:38903
 6:04 am on May 10, 2003 (gmt 0)

rfgdxm1 may be on the right trail with reducing relevance of free search results in favor of increasing relevance (and returns) on Adwords. Who knows what kind of demands their investment bankers are making before Google IPOs?

Lately, Google results appear as if Google is wearing beer goggles. OO

Ted

rfgdxm1




msg:38904
 6:28 am on May 10, 2003 (gmt 0)

>Are #4s links that good? For a site that is'nt working?

And didn't Google implement in the algo that unless the inbound anchor text matched what was on the page, it would be ignored to prevent Googlebombings? Looks like Google hosed that part of the algo also. :(

digitalghost




msg:38905
 6:32 am on May 10, 2003 (gmt 0)

>>And didn't Google implement in the algo that unless the inbound anchor text matched what was on the page

No, never happened, and I haven't seen anything from Google to indicate that was in the works. Theming just isn't happening.

Google's algo is the property of Google. All we can do is make an educated guess about particulars.

why2kit




msg:38906
 6:39 am on May 10, 2003 (gmt 0)

and I'll throw this one in out of left field. while Verisign must have a large number of inbound links (due to the tags some folks put on their sites to show that they are kosher) check www,www2 and www3 for commission junction. Intesting results

nativenewyorker




msg:38907
 9:50 am on May 10, 2003 (gmt 0)

Actually, I should have titled this thread NetSol and not Verisign. It was a slip-up on my part as Verisign owns Network Solutions. If a moderator would like to correct the title, please do.

In any case, the search result is pointing at a parked domain and not Network Solutions itself. Network Solutions should not be benefiting from inbound links to the parked domain as there are no links on these parked pages.

Let's try GoogleBombing Google to see if the GoogleBomb still works. Anybody game? The term is SERP [google.com].

Ted

iThink




msg:38908
 1:30 pm on May 11, 2003 (gmt 0)

This is not the isolated example. Search for mp3:

[google.com...]

#9 is napster.com. The site that is closed since last 2-3 years. The only page hosted on napster.com is a holding page with 2 links going to roxio.com (its new owner?).

Most will agree that MP3 is one of the most competitive 50 single keywords out there and yet this holding page with "great content" is out there masquerading on page-1 most likely due to large number of backlinks because the site has been in news due to all the wrong reasons.

Kackle




msg:38909
 1:59 pm on May 11, 2003 (gmt 0)

Methinks Larry and Sergey have been too busy scooting around on their Segways.

One benefit we would see from a post-IPO Google is that their public relations people would be more responsive. For example, if a major journalist were to pick up on the "Google is broken" story, a public-corporation version of Google might actually investigate the problem. Why? Because they'd be nervous about the reaction of stockholders if they lose credibility.

But a pre-IPO, private Google just shuts out the critics as if nothing at all has happened. They have more to lose by responding than by keeping their mouths shut.

Meanwhile, those of us who have a lot to lose with a broken Google, depend on forums such as WebmasterWorld to glean clues about the true state of affairs at the Googleplex. While GoogleGuy can be helpful when it comes to the "small picture" of what's happening, he's not helpful with the big picture of how it's all fitting (or not fitting) together.

Maybe GoogleGuy doesn't know enough. The really scary thought is that maybe no one at the Googleplex can see all the ramifications of insanely complex algorithms. It's almost to the point where a new search engine, starting out with near-random ranking and a little bit of filtering, would be as interesting as Google's SERPs.

Keep those "I'm feeling broken" search combos coming. I know of a site that would like to do some screen shots for a new feature.

So far we have "discount brokers," with or without quotes, and www or www2. You have to be number one to make the "I'm feeling lucky" button work.

Dynamoo




msg:38910
 3:49 pm on May 11, 2003 (gmt 0)

I can confirm that the PR0 expired domain penalty appears to have vanished for a couple of my sites on the www-sj, www2 and www3 indices.

However, the indexed pages on those servers are ancient.. it really looks like an index that's a couple of months old. The snippets are about 3 months old I'd think and the SE results look just like the old ones.

I did have a problem with Google PR0ing some sites when it shouldn't have, so I would hope for better results. However, I'm getting good search results on domains which haven't had new backlinks, so it looks like the expired domains algo hasn't been applied yet.

BigDave




msg:38911
 9:34 pm on May 13, 2003 (gmt 0)

I don't exactly see how one bad result on the front page of a few SERPs is going to destroy google.

The only people they matter to are those of you looking to find little faults, and the google hackers that are trying to make the best search engine they can.

With these isolated cases, would it be better to report it to google than run a bitchfest on WW?

It's very rare for me to get a page where I am willing to click on all 10 results. And of those 10, at least one will usually lead to a 404 or a parked domain.

Dynamoo




msg:38912
 10:00 pm on May 13, 2003 (gmt 0)

It's not just one bad result. I've got about 25 sites and it's like a blast from the past.. the results are about 3 months old.

Kackle




msg:38913
 11:53 pm on May 13, 2003 (gmt 0)

I've been studying the "discount brokers" situation, and have arrived at some conclusions, based on that one top link that shows only an empty directory (No.1 on www, No.2 on www-sj). Without quotes, on www, the two words show 494,000 hits.

I understand better now what is happening. It's not a bug, but rather a dramatic illustration of the role that link rot is playing in the decline of any search engine that uses link popularity.

First of all, by using the Wayback Machine, you can see that the empty directory was empty last November. The previous Wayback entry is April, 2001, a year and a half earlier. You can see what the page actually looked like in April, 2001. Google has had several crawls to detect this empty directory, and it still ranks at No.1 or No.2. It's probably been longer than several crawls, but I don't know how to figure out more accurately how long that directory has been empty.

Secondly, several major engines are ranking this empty directory at or near the top the same way that Google does. It's not a Google glitch, it's a generic search engine problem.

Third, the empty directory still has many backlinks that are valid, with the words "discount" and "brokers" in them. The problem is basically one of undetected link rot on the part of these sites that link to this empty directory page. Those that use link checkers to catch 404 pages cannot catch this empty directory, because Apache puts out a little HTML to display the directory listing automatically, and it looks like a 200 instead of a 404 to the link checkers (and to the crawlers). This is one of those Internet service providers that gives you a /username/ or /~username/ web directory with a basic service account. The ISP won't set Apache to produce a 403 Forbidden for directory listings, probably because many of their clients don't know how to make a real index.html page.

Fourth, I think we have to admit that since this page was about reviewing online stock and commodity brokers, and since the market has by now wiped out the day traders, this is a somewhat unique situation. Link rot is much less likely to be noticed when it's found in the middle of a huge graveyard. To put it another way, the half-life of the average URL, which is distressingly low to begin with, has been extremely low recently in the field of online brokers.

Finally -- and this is worrisome -- if it's not a bug, but rather business as usual at Google and other engines, it means that anchor text plus link popularity is just about the only game in town. On-page characteristics cannot hold a candle to anchor text plus link popularity. I guess this is what the Googlebombing phenomenon should have taught us a year ago.

PatrickDeese




msg:38914
 1:00 am on May 14, 2003 (gmt 0)

I have noticed that a site that I was once involved with that hasn't served content since 1999 still has pages in Google index. I first noticed this about a year ago, so it is nothing new to this new algorithm.

In fact, about a year ago, Google actually said it had the cached version of a page from this site (it couldn't produce it when I called its bluff).

However, I am sure, at least in the case of this stock trading site, and in the case of the site I ran with my former partners, the sheer mass of links to it dictates that google keeps "seeing" the site. It simply is giving the site the benefit of the doubt, and assuming that the site is having temporary difficulties.

Truthfully this should be great news to the folks who are always worried that their server was offline for 6 minutes during a crawl. A site with incoming links can live for years without a server. :)

1milehgh80210




msg:38915
 1:10 am on May 14, 2003 (gmt 0)

I don't see anything unique about 'brokers'.
Check #7 for discount autos. ( Internic )
G just does'nt like the word "discount" :O LOL

This 41 message thread spans 2 pages: 41 ( [1] 2 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google News Archive
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved