Update Dominic - Part 2

[teeceo]

Continued from Part 1: [webmasterworld.com...]

---

Thanks to google for letting us peek into there database( they could very easy close that door to us)and a thanks to googleguy for "DONATING" his time to answer question(that he don't have to) and for keeping thing here (somewhat) calm. Also, thanks to all that work so hard to keep this forum going strong(I for one don't know how I would get along without it:). I could go on and on but, thats all I will say. Later.

teeceo.

[spacewar]

i mean i'm worried about the google update, scares me every month, google being such a powerful monster, so sj is secret sauce but is it an update as in 'the' update

[eraldemukian]

GG: Does that make sense?

yes

it makes allot of sense.

on the other side I can understand why the -sj behaviour creates so many postings. The results are quiet different from the other datacenters.
And it appears to be a new phenomen. So people can not help but to
speculate about whats going on.

Any info that is real helps. I rather see brief, yet acurate messages, than more info-noise based on speculation which is often fueld by pure SERP-drop-paranoia for indivual sites/keywords.

And that these brief messages have a natural limit in their scope is very understandable.

[edited by: eraldemukian at 2:37 am (utc) on May 5, 2003]

[Chris_D]

Thanks GG

I think that your last couple of posts have told us heaps.

Thank you again for taking the time out to help and clarify.

rfgdxm1 - all datacentre results are available for public view - as are www2 and www3 etc. And this isn't the first time we've seen 'things' slip in and out of the live index - remember when we first saw freshbot appear?.

I think its now that we pay attention to 'glitches' more than previously, we read more into them - and more people share their views here!

Best

Chris_D

[GoogleGuy]

Happy to try to help. Some people get into quick knee-jerk reactions, but even in the 500 post thread, some solid info came out in the first 16-17 posts or so. Good qualities for an SEO include curiosity and looking for new opportunities whenever something changes.

[eraldemukian]

Happy to try to help. Some people get into quick knee-jerk reactions, but even in the 500 post thread, some solid info came out in the first 16-17 posts or so.

my knee-jerk reaction to that was to read the first 16-17 posts again.

Which basically said that -sj is a datacenter, and the links are down on it. [doh].

tounge-in-cheek question:
how literal is '16-17'

[Stefan]

Well, I think I understand... hope so anyway. I can imagine everyone going through those first 17 posts examining every word for clues.

Thanks for the guidance GG. (Please note that I did refer to Google as "a very good SE".)

[GoogleGuy]

Stefan, no worries. Competition and diversity are good things--they make search better for users. :)

[Stefan]

Good thing. :-)

[steveb]

"... even in the 500 post thread, some solid info came out in the first 16-17 posts or so."

Message 11:
"This looks like an old database to me. Probably something they are toying with."

Message 12:
"I thought it was an old database too until I noticed some recently acquired incoming links to my site, and relatively new cache."

Message 13:
"The database is old but definitely a different algo."

[Stefan]

Yep.

[Powdork]

BZZZZZZZZZT. Wrong.

Hey, you got that from resource zone;)
and no I don't get much traffic from surf safely

or even(gulp)dmozguy?

Dude! Its MozillaMan:)

I've been wondering if -sj is not a bit screwed up on purpose, with a mixture of an old database and new to keep us from finding out too much about this special sauce

[deanril]

I enjoy seeing the data on -sj. Because I made some new statics(my site is dynamic), awhile ago and was getting frustrated that they werent being crawled.

Then a few weeks ago saw my googlebot 2.1 hits go up another 30, for a couple days (30 new statics) and when I look at -sj there they are. Not on www. but -sj has them, so it relaxes me a bit, to know they got indexed and its just a matter of time.

It would be nice if google could index my other 1000 pages, maybe I could rank a little better :)

[rfgdxm1]

>Hey, you got that from resource zone;)
and no I don't get much traffic from surf safely

Ahh...I see that since that thread was closed for arguing with the umpire, you had to respond here. ;)

>I've been wondering if -sj is not a bit screwed up on purpose, with a mixture of an old database and new to keep us from finding out too much about this special sauce

Very unlikely. Why not just take sj offline so it can't be publicly seen? Although, I wonder if there is perhaps even though this is madness, there is a method in it? Consider the possibility that some Google intelligence agents (we do have a Google programmer posting here) decided to "accidently" let this leak, and Google is using us as beta evaluators? If things are seriously hosed, someone here (or, another forum) might spot things that the Google crew would miss. Google may have decided that the risk of letting some low life SEO types get some clues about the algo was less of a risk than letting loose a buggy update on the whole Internet. Lots of dissatisfied surfers would be worse than a small handful of SEOs who could use this to advantage.

[GoogleGuy]

deanril, glad to hear that you have some more pages indexed in SJ. rfgdxm1, every index has to pass a really stringent battery of tests before it becomes visible. SEOs might notice a slightly different build of backlinks, but things like that could be balanced by factors that improve search quality more in other areas. The other thing to bear in mind is that it's easy to re-sync something like backlinks or spam snapshots once you're convinced that an algorithm or method is an improvement.

[d_fused]

What I have read recently is that Google will be stressing on the spam factor in the future...

<!-- it's easy to re-sync something like backlinks or spam snapshots once you're convinced that an algorithm or method is an improvement. ->

Sorry, I am not very good at reading comprehension, but... what if we just ignored the majority of spam results in SJ...

just a thought at 6:50 a.m.

Regards.

[deanril]

Yeah thanks Googleguy, im a newby to all this, and its quite fun, I must tell you.

The waiting bit is new to me, Im inpatient, but you know what, Im becoming more patient : ) Im learning other things here that reflect in my daily life, heheh

The fun part is watching my un-attentative competitors get bounced out of their cushy #1 slots.

I tell you what I really get a kick outa that! Most have enjoyed the good life for too long, time to get back into their sites and do some optimizing.

Meanwhile its fun to come out of no where and snag their position. I just get a kick because I can picture these people faces " Who's this So and So?" And by the time they try an take what was theirs back, I will all ready have optimized even more. This to me is a fun game!

[rfgdxm1]

>rfgdxm1, every index has to pass a really stringent battery of tests before it becomes visible.

Based on what I see, and many others have reported, what is on sj is failing the test.

[Krapulator]

"SEOs might notice a slightly different build of backlinks"

I'll say. I have gone from 642 backlinks to Zero on www-sj

Im in big doo doo if this becomes the main algo.

[polarmate]

Same here. Zero backlinks. Add to it the affiliate URL that Google awarded us in Jan - yep, that's the one that shows in sj. Luckily such an affiliate does not exist - the affiliate whose site Google picked thsi URL from had it badly formulated. And that badly formulated URL has since been corrected - way back in Jan itself. Would be sad (and a lot of other things!) to see Google give us an old and invalid URL.

"Secret sauce" needs to be cooked some more before it can be served.

[poet22]

Seems to me GG just said once they check the algorythm they can easily re sync or put back in all the backlinks, no?

[Krapulator]

***nods convincingly and pretends to understand***

[markusf]

Google guy said..
Competition and diversity are good things--they make search better for users.

Google guys comment here got me to thinking.

Nothing ever goes into a production environment unless it is tested and the SJ datacenter is no fluke. Here is what i think is going on.

1. A few data centers are going to have near real time calculation of on the fly PR, and pages that are fresh botted will go into the index right away and stay there. Right now this would only be the SJ datacenter.

2. The other 7? data centers are going to continue doing the once a month updates coming from the deepcrawl.

3. These results are going to appear on www.google.com 1/8th of the time so that google can moniter the performance of this datacenter and see how the new algo is performing.

I think the move to on the fly PR, and constant freshness, means that google can not test this index/algo in a testing environment as the resources needed are just to great to mirror a production environment. Don't be surprised if in the months to come other datacenters start acting like SJ...

[mil2k]

Yes poet22 i agree with you. I think he meant to say that they are testing "A Part" of an algorithm and testing the results. Hence this algo has tweaked the importance of one of their parameters and they are observing those results to gain some insights into the quality of those results.When they complete their experiments they would then decide whether or not to include the algo tweaks in the next update. Hence they are using 2 months old index with some fresh results. The results what we are seeing on -sj would not necessarily be seen on the main index after this month's update. So this is more of their R&D exercise (alongwith QC i suppose).

[madmatt69]

Hi guys,

www-sj is strange.

It shows a tonne more inbound links to my site - Definetely nice. BUT. My site itself isn't in the index! Not a single page. It's in all the other www's, and has been for the last two months (when it first went online).

Strange stuff, but like others, I won't get too worried until there's a real update. My site doesn't use any spammy tactics, so it should only be improving in the serps.

[CCowboy]

God some of you guys are so negative!

Google Guy said concerning www-sj: "bear in mind is that it's easy to re-sync something like backlinks or spam snapshots once you're convinced that an algorithm or method is an improvement."

I think we are going to see an update very close to what we all expect... If you are playing by the rules and have good back links.

[rfgdxm1]

>God some of you guys are so negative!

Yeah, slashing your wrists over your SERPs on a Google testbed datacenter does seem a little drastic. ;)

[nutsandbolts]

I have a good example to share once this baby is live on the main WWW as to why SJ isn't quite working correctly.

>Googleguy: it's easy to re-sync something like backlinks

That sounds like SJ has the new secret sauce with many backlinks ignored for now.

[daamsie]

So far to me it seems that they have purposely included old spam results in the sj results as a way of testing some new filters.. (something we have been told about)

It only makes sense that they put in old spammy (hand-edited) results for testing purposes to see if their new algo will manage to filter them out. If it works on the old ones, then the new results should be cleaner too and catch a lot of similar spam before it is on www3.. and save them a lot of time hand-editing the results.

Of course, just a wild guess ;-)

could I be close, GG? :-)

[mil2k]

Hey i am seeing different results on 5 datacentres! And some of those are being reflected on the main index.

[coconutz]

link:www.yahoo.com [google.com] = 384,000 on www.google.com

First time I've actually seen this on www. Anyone else?