McDar Experiment

Thank you Shawn, your a good man...I don't care what your friends say .


The idea of this experiment serving no purpose is interesting! I think this thread stands as a historical cyber monument. I have read it a few times. Although it really tells us nothing of the future in terms of SEO, SEM, SES we can look and see what once was.

The flux we are seeing may be here to stay . I personally think much of this algo is based on quality and page positioning on ibls...link schemes have been targeted, uncovered and devalued if not penalized. To do this they must have developed a very complexed program that can spot the patterns! I say this because much of the programming, LV for example, is Server side. So they must be seeing an off page pattern?

I can also see a massive clean up on Google's perception of spammy sites/ pages.

Much of what I read in the article posted earlier in this thread seems accurate.

Less stability with Google leads to this conclusion! Most SEO's are scratching thier heads while this is going on. EXACTLY what Google wants. So if this was to be ongoing...well you draw your own conclusion .

 

as I consider this thread very helpful, even becoming a piece of history both of the entire forum and also serps via google, cheers to Shawn for bringing it back were it belongs

 

Let me rephrase my comment.

The experiment started with Compar's points about McDar's tool which compared serps with allinanchor rankings.

There were questions about the relationship of these two measures.

The experiment evolved with an experiment page being developed and members here running links to the site with specific anchor text *sleeping bags*.

As Caryl mentioned, this experiment showed that clearly this worked to boost rankings.
1. There was an apparent relationship between the two rankings.
2. The anchor text didn't have to come from *related sites*.
3. Significant serps rankings could be achieved with *google bombing* of a specific anchor text.
4. The *google bombing* of anchor text would also drive the allinanchor rankings up.

I'm generally impatient. I like quick answers.

It seems to me the dc's are coalescing (more so than at any time since 9/22).

If they do move to settle on a ranking around 400 (something) and about 100 in allinanchor rankings...then the apparent relationship we witnessed for most of 1 and 1/2 years has been significantly changed.

(now I do hope as Caryl suggested that the serps and allinanchor settle at better rankings.)

But if they don't move up...then what are we watching...and what are we doing?

Dave

 

Dave,

I can appreciate your impatients - But I tend to want to evaluate a situation thouroughly before jumping in a changing things.

Below are the changes I've recorded just since this morning!


The way the page keeps bouncing around is either due to Google still tweeking OR the page is being thrown in a more "random" placement then the top 100 or so results.

Either way - How the hell do you evaluate this mess?

What if I deleted a ton of links this morning? Would I look at this afternoons results as confirmation that deleting the links worked? Afterall, the page advanced 70 positions on some Datacenters.

We could do a lot of things right now but it would be like shooting in the dark at a moving target.

Caryl

 

Let me run a different question.

The experiment page started seeing Jagger changes on 9/22; earlier than many sites and continues to see fluctuation between the dcs.

In my industry everything has settled. All the dc's agree on serps for 9 phrases that I check. (I haven't checked allins)

During Jagger there was dramatically less fluctuation for the top 20 sites than what the experiment page has experienced.

The jagger impact started later and ended earlier than the experiment page.

Can we take a look at this.

Some of us have guessed that the nature of the links (non-related to sleeping bags) is the cause for the *fragility* of the experiment page. This might have caused the wide fluctuation for both allinanchor and serps.

We haven't confirmed this or not.

Does anyone have ideas on how to check this?

 

Dave,

Would you agree that links from the co-op ad network produce similar "unrelated" links?

If so, have you heard of sites depending on co-op links recovering from or not being effected by Jagger?

Caryl

 

I have.

I'd like to offer more examples but I'll just mention it for now as I don't want to skew things.

 

Caryl:

Good idea.

I'm not part of the co-op network or link vault. I can't comment on my experiences there. I'm sure 99% of the links would be considered "unrelated".

But certainly we can find out about the impact from people here.

I've been searching for forum (many forums) comments on Jagger. I'm a bit suspiceous of some of the articles...though I think I read the one you referenced.

Comments on Jagger are drying up. But we ought to ask people here who are in the co-op network and see what the results were. (Frankly, since I'm not a member of either network I glossed over comments in earlier Jagger threads)

One interesting comment I remember though was asking about the relative level of co-op links versus "related" links as to the impact of Jagger.

We can leave this here for responses or start a thread.

Dave

 

Unfortunately some of the Trolls seem to have made it here. I have been ridiculed pretty hard by some people..Others have been very decent..I have noticed by jumping around this forum that those same trolls do it to just about everyone. It is very unfortunate when you have to deal with people like that however I have learned to ignore the trolls and the red they have given me and speak my mind when i feel I am making a good point... By the way awesome tool Caryl I use it almost every day.

 

From our perspective jagger doesn't appear to be a straight change in allinanchor, nor the co-op. Daily changes in serps from 800's to top 10 and back to the 800's continue for us.

 

NEW PAGE UPDATE:
*Google PR/Backlink update
**33 Active DCs(6 New DC Found)**


The "New Page" sleeping-bags.htm
04/07/2004 "New Page" went live

Search for "Sleeping Bags"
_____________
NOTE: Due to size constraints, the entire report can no longer appear in a post.
You can view the entire report here... Complete Report
________________
Last 10 days

Postion #17 allinanchor: #81 --11/18/2005 [16 DCs #17, 4 DCs #275, 3 DCs #303, 10 DCs #304]]**(Allinanchor #28 on 16 DCs, #81 on 17 DCs )
Postion #16 allinanchor: #83 --11/19/2005 [16 DCs #16, 4 DCs #269, 3 DCs #305, 10 DCs #306]]**(Allinanchor #34 on 16 DCs, #83 on 17 DCs )
Postion #16 allinanchor: #84 --11/20/2005 [16 DCs #16, 4 DCs #331, 10 DCs #332, 3 DCs #334]]**(Allinanchor #32 on 16 DCs, #84 on 17 DCs )
Postion #331 allinanchor: #84 --11/21/2005 [10 DCs #15, 2 DCs #63, 4 DCs #302, 4 DCs #309, 3 DCs #316, 10 DCs #331]**(Allinanchor #20 on 2 DCs, #33 on 10 DCs, #83 on 7 DCs, #84 on 14 DCs )
Postion #461 allinanchor: #102 --11/22/2005 [10 DCs #132, 4 DCs #339, 3 DCs #422, 10 DCs #461, 2 DCs #483, 4 DCs #484]**(Allinanchor #27 on 10 DCs, #85 on 4 DCs, #102 on 19 DCs )
Postion #430 allinanchor: #105 --11/23/2005 [12 DCs #430, 4 DCs #434, 9 DCs #446, 4 DCs #451, 4 DCs #487]**(Allinanchor #102 on 6 DCs, #105 on 27 DCs )
Postion #482 allinanchor: #105 --11/24/2005 [10 DCs #430, 2 DCs #462, 4 DCs #464, 7 DCs #480, 10 DCs #482]**(Allinanchor #104 on 6 DCs, #105 on 27 DCs )
Postion #485 allinanchor: #108 --11/25/2005 [10 DCs #425, 2 DCs #481, 10 DCs #485, 4 DCs #488, 7 DCs #492]**(Allinanchor #108 on All DCs )
Postion #425 allinanchor: #107 --11/26/2005 [2 DCs #392, 7 DCs #402, 10 DCs #425, 10 DCs #426, 4 DCs #472]**(Allinanchor #107 on 26 DCs, #108 on 6 DCs )
Postion #425 allinanchor: #100 --11/27/2005 [4 DCs #421, 10 DCs #425, 6 DCs #437, 10 DCs #463, 3 DCs #485]**(Allinanchor #100 on All DCs )
Postion #421 allinanchor: #102 --11/28/2005 [2 DCs #398, 10 DCs #421, 4 DCs #443, 6 DCs #444, 4 DCs #450, 4 DCs #463, 3 DCs #467]**(Allinanchor #100 on 10 DCs, #102 on 21 DCs, #116 on 2 DCs )

Postion n/f allinanchor: #119 -11/29/2005 [4 DCs #402, 3 DCs #409, 4 DCs #411, 10 DCs #436, 12 DCs n/f]** (Note: Allinanchor #118 on 12 DCs, #119 on 21 DCs )

Cache Date: Nov 09, 2005 08:11:11 GMT
_____________________________________________________
note: these pages are found using ( site:www.compar.com +sleeping Bags )
number of Bob's PR6 and PR5 links found:

142 - 11/16/2005
Unchanged from previous date
140 - 11/19/2005
Unchanged from previous date
149 - 11/21/2005
142 - 11/22/2005
Unchanged from previous date
105 - 11/24/2005
Unchanged from previous date
101 - 11/26/2005
100 - 11/27/2005
Unchanged from previous date
__________________

note: this page is found using ( site:www.ski-france-ok.com +sleeping Bags )
Foxy's PR5 and two PR4s
0 - 4/20/2004
3 - 5/04/2004
Unchanged from previous date
2 - 6/17/2004
Unchanged from previous date
__________________
Bob and Foxy added link Sleeping Bags to their sig files on this forum 5/06/2004
Results for - ( site:forums.digitalpoint.com +sleeping bags )
Note:Foxy & Bob removed links from Signatures - 6/11/2004
4060 - highest links reported 6/12/2004
Note:Foxy & Bob replaced links in Signatures - 7/21/2004
Note:Foxy & Bob removed links in Signatures - 11/16/2004

418 - 11/04/2005
409 - 11/05/2005
Unchanged from previous date
2300 - 11/09/2005
2290 - 11/10/2005
Unchanged from previous date
409 - 11/17/2005
2290 - 11/18/2005
Unchanged from previous date
2280 - 11/27/2005
Unchanged from previous date
__________________

Caryl and Foxy added link Sleeping Bags to their sig files on the SEO Chat forum 5/11/2004
Results for - ( site:forums.seochat.com +sleeping bags )
Note:Foxy removed links in Signatures - 11/16/2004

156 - 9/08/2005
Unchanged from previous date
155 - 10/06/2005
Unchanged from previous date
138 - 11/04/2005
Unchanged from previous date
147 - 11/19/2005
Unchanged from previous date
145 - 11/27/2005
Unchanged from previous date
__________________

*August 18, 2004 - (10) Project Supporters donated Links from 13 different IP addresses to page. Additional info will be added as it comes in.

Total links found by using site:www.domainname.tld Researched daily by Foxy

Daily Totals Report (click to see full report)

606 - 1/29/2005
457 - 1/29/20052/6/2005[/B](148 lost links from one IP) [/b]
251 - 2/12/2005
261 - 3/03/2005
273 - 3/18/2005
275 - 3/19/2005
508 - 4/09/2005
560 - 5/20/2005
627 - 6/14/2005
536 - 6/25/2005
584 - 7/11/2005
631 - 7/29/2005
15,785 - 8/19/2005
3891 - 10/20/2005
3904 - 10/26/2005
3699 - 11/08/2005
3461 - 11/28/2005
__________________

Dec 17, 2004 - digitalpoint's "Patient" Added link to 116 different urls
(Located on 3 different IPs within same class 3 IP address Address - 217.36.188.*)

Jan 20, 2005 - (moved all urls to dif server - now all 116 are on same IP + changed from Coldfusion to Java)

Number Indexed by Google

40 - 12/23/2004
48 - 12/24/2004
107 - 12/30/2004
115 - 1/2/2005
116 - 1/21/2005
108 - new urls & 70 old urls- 1/31/2005
__________________

PR/Backlink info

~ April 22 Update ~
sleeping-bags.htm has PR5
143 Backlinks reported
___________

~ March 3 Update ~
sleeping-bags.htm has PR5
113 Backlinks reported
___________

~ February 3 Update ~
sleeping-bags.htm has PR5
260 Backlinks reported
___________

~ January 1 Update ~
sleeping-bags.htm has PR5
268 Backlinks - see Table of Links reported
___________
~ December 16 Update ~ (only Backlinks updated in this update)
sleeping-bags.htm has PR6
155 Backlinks
(Details can be found for this and the following in Complete Report)
___________

"A nice tool that Shawn didn't write -" DISCUSSION
__________________
Note:Google API has page at #n/a

 

NOT FOUND on 12 Datacenters!

Postion n/f allinanchor: #119 -11/29/2005
4 DCs #402,
3 DCs #409,
4 DCs #411,
10 DCs #436,
12 DCs n/f

Allinanchor
#118 on 12 DCs,
#119 on 21 DCs

Curiously, the page has been NOT FOUND in allintext: search for the past three days.

HOWEVER - it is interresting to note that page 246 of THIS THREAD is in position #320 for that search.

Caryl

UPDATE:within the last 40 minutes 7 additional Datacenters have dropped this page completely from the results!

 

Now I would like to see all the people who claim that Google has a master plan, and all we have to is understand it, explain this one. What type of master plan simply loses a web page?

When will people finally acknowlege that Google is beyond any rational behavior. I don't know what they might be trying to do, but I know this it isn't working. There is no plan or objective that can be fulfilled by losing a web site from a large number of your datacenters.

The reason I haven't participated in much of the speculation and debate about what is going on with Google is because it is a fools game. Nobody can divine what is going on, because the major thing that is going on is that the system is broken.

In my opinion the only rational thing to do is sit back and wait until they at least learn to count and not lose web pages.

 

I saw this with one of my main page/keywords last week-- Google completely lost my cached copy, too-- it's been like that for about a week or so. Maybe the same is happening with the experiment page-- though you say the cache still seems to be there(though phantomly dated)...


--
Derek

 

I just checked to see if the cached page was being dropped on the Datacenters that were reporting the page NOT FOUND and saw that the datacenters are now showing the other "Phantom" cache date of Nov 19, 2005 23:13:23 GMT

Earlier this morining they were returning the "Phantom" cache date of Nov 09, 2005 08:11:11 GMT

So changes are a foot in Google Land - at least in this neck of the woods!

I agree, Bob - 100%!!!

Caryl

 

For all to recall; From 10/27/04-12/02/04 the experiment page had its all time highest rankings; obtaining as high as a #2 serps for sleeping bags and retaining a#2 ranking for allinanchor.

From 12/2/04-2/2/05 the site had similarly strong rankings.

then all hell broke out till 3/22/05.

Again from 3/22-8/13/05 the site had high rankings and high allinanchor.

Then sh*t starting happening with some higher serps leading to the jagger start at 9/22. Even during that period there were some strong serps days.

Now the page is ranked in the 400's with some flux and not even found on some dcs.

That is wierd.

This is a great experiment. How can Google explain this incredible fluctuation in serps? It points to tremendous changes in google.

It probably also reflects changes in google that reflect on the optimization strategy and methodology for this site.

Whether broke, perfucked, or going through filter changes it reflects the amazing variations in existance with google and its .

Lets keep going and see what transpires.

Dave

 

Determining that this is simply a database corruption issue is an achievement unto itself.

Ah... but it is more interesting than that. I have about twenty web sites. Only one was "lost" during Bourbon and the same one was lost during Jagger. What makes this one web site special among all of my web sites?

Figuring that out would bring major benefits to the bottom line.

 

It sure would-- sadly, all my websites were hit, so I have nothing to even try to analyze...


--
Derek

 

That is what I would like to check.

Dave

 

I suspect that this issue is NOT on-site/on-page specific but rather something as wierd as the age of the page and the particular grouping it is stored in on Google.

Will, have you checked cache dates and other factors that this experiment records and found any similarities with your "troubled page/site"?

Caryl