SteveB - >Of course random is possible with "software"
"Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin." - John von Neumann
Yes google makes mistakes, that is not in doubt. However your suggestion that because some sites get hit with data refreshA and then not in data refreshB is evidence that errors are taking place is not a correct assumption. You are missing many other possibilities including deliberate confusion for webmasters by constantly changing the algo. Only this would be random.... not the software driving it. The concept that errors create random effects is nonsense. There will be a pattern, even if you are the victim of collateral damage and some bizarre site becomes number 1.
The important point is that it is extremely unlikely that a human is introducing a bizarre random element. There is possibly a random tweek done by engineers within established rules but I don't think this issue is anything to do with 'random'. There will be a pattern even if it is deliberate or a mistake.
shogun, my site is out of supplemental on those dcs, but currently there is no return of former rankings on those dcs.
We just noticed that google is very slow in indexing new pages this past week. Typically, we put up a page and it shows in the index after 3-4 days.
We added about 15 pages between Christmas and New Years.
Yahoo and MSN had them indexed in 3 days
Google still does not have them indexed.
Very strange google crawling behavior this week as well.
I too put up 50 new pages on Dec 26th. My server logs showed that Googlebot visited most of, if not ALL of, the new pages on Dec 27th.
As of today, NONE have shown up in the index.
I have also made numerous changes that have effected a total of 55 additional pages. In spite of healthy spidering - none of these changes are reflected in the currrent index either.
This is why I refer to the current googlebot as 'LazyBot'!
|my site is out of supplemental on those dcs, but currently there is no return of former rankings on those dcs. |
Same, here. The site commands come up nice and clean, but no difference with the rankings.
Looks like some kind of refresh (dare I say anything more...) despite constantly adding new content we've been stuck at 1,180 pages for about two months. Yesterday I noticed 1,190 page - that's good right?
Well about two hours ago I saw 1,170 and just now it was 1,150 - numbers I've NEVER seen before. So it SEEMS there is some kind of shake-up underway.
Its funny, we run adsense as well and usually the adsense bot makes it to the page as soon as we put the pages up. So its not an issue of getting the bots to the pages or the bots to understand the pages.
Its an issue of google tossing in a cached page in the index.
Folks - the hacked websites spam is back in full force, after Google seem to have it cleared up the past couple of weeks.
Which makes me think that the current results are quite temporary - that's what happened last time - those websites hit the SERPs and shorty after - a bigger than usual refresh, and the hacked spam was pushed back.
Also, our main page shows a Jan 4 cache on some keywords and an old one on others.
Something is cookin'...
Another thing - on our sitemaps, the keyword queries have not been updated for probably a month - anyone else seeing this?
|Another thing - on our sitemaps, the keyword queries have not been updated for probably a month - anyone else seeing this? |
Yes, the query stats under statistics on GWT show old results as before my site was hit on December 2006. I wish to have that rank again.
>> now the data refreshes seem to be every two or so days <<
I thought I saw Matt Cutts or Adam Lasnik confirm that is how things now work as of a few weeks ago?
"However your suggestion that because some sites get hit with data refreshA and then not in data refreshB is evidence that errors are taking place is not a correct assumption."
You miss the point. It is not "some sites" hit in data refreshA and but not in data refeshB. It is how pages on the same site are effected by both in different ways. Again, don't focus on your site. This is a page phenomenon. Over the past year and a half, pages on the same domain get penalized and unpenalized, now on a daily basis. Pages on the same domain with the exact same algo value get penalized and unpenalized, as explained above.
If you don't understand the random nature of the application of the penalty to different pages within the same domain, and also the penalization and unpenalization of the exact same page that has not changed at all, and also the arbitrary penalization of anything below a penalized page in a directory structure, you are simply ignoring the phenomenon.
What isn't random is that Google is attempting to put a certain type of page at #950 in the results. And what is true is they also put a collection of other pages down there that they are not trying to, and the status of those mistaken pages randomly changes with each data refresh. Not only is this demonstrable, there are a year and half of threads talking about it.
"the hacked websites spam is back in full force, after Google seem to have it cleared up the past couple of weeks."
Very true. The big daddy lazybot loves this stuff.
|Google is attempting to put a certain type of page at #950 in the results |
What TYPES of page do you think Google is actually trying to send to position #950 that results in these false positives?
My site fits the pattern for having this phenomenon. I have a page that is number 1 for a very competitive and commercial phrase and also ranks 950+ for an equally competitive phrase. It used to rank for both and the keywords are very different. In fact, many pages show this. You seem to think that it is a page penalty which it isn't.... it is related to what keywords are searched for and depending on those keywords, the page will be treated accordingly. In my case, a big factor is the anchor text I use to link to another site... here's how it works:
Search term "widgets in Australia"
My site's page links to site B with 'widgets in Australia" as anchor text. Site B appears 15th while I appear err... used to be number 1, but I am now 950+. I change the search to "widgeting in Australia" and this phrase is not on site B's page that I link to ( a different site B page ranks in it's place) so I am number 1 again. I change the anchor text to "widgeting in Australia" on my page. Using the search term "widgets in Australia" my page now ranks at number 1 and the same page I link to on site B is 20th.
I then test all the links to site B and the pattern is consistent. I have now completely removed some links and the pages rank with site B nearby in the rankings.
This problem only arises when the site has its supplementals above the normal pages. I think in most cases it will be more than just an anchor link problem which cures it. If the sensitivity of the algo is turned up for pages which are supplemental heavy, then any number of reasons could cause a page to trip up and crash to the bottom of the serps. It starts to get very complicated when you need two pages from a site to help each other rank well. Both need to be squeeky clean and tick all the boxes. At the moment I have a string of pages covering a theme and the top level one in that folder is still awaiting a new cache. The others are now cached and ready, but not ranking becuse the top page is still tripping the algo. Word density and other factors have also been changed, just to add to the mix.
>...also the penalization and unpenalization of the exact same page that has not changed at all
This is where you may have to look at other pages which help that page to rank. "No page is an island", so not only do you need to look at your site as a whole, but also other sites that rank within the search term and their relationship to yours.
>What isn't random is that Google is attempting to put a certain type of page at #950 in the results.
No, that page is failing for certain keyword combinations. The 'type' of page is not causing the problem, it is the fact that your pages are being subject to more intense examination by the existing algo. Other sites pages may have a very similar profile but not supplemental heavy, so they rank as before. The big question is, what triggered the site to go supplemental heavy in the first place? Although I have managed to regain rankings for my main keywords, my site is still suplemental heavy and tomorrow, if the algo changes, I may drop because of the critical optimisation that is required.
>And what is true is they also put a collection of other pages down there that they are not trying to...
Of course that happens. You can write a perfectly honest and non seo inspired page about a topic and still trip algos. That's just life.
>and the status of those mistaken pages randomly changes with each data refresh.
Probably because they are marginal in their problems that trip the phenomenom. Nothing random, they are close to the edge of innocent or suspect.
Just noticed my site is no longer supplemental heavy.... anybody else seeing this with theirs?
Added.... Other sites I follow have come out, perhaps something is now moving? Good luck.
[edited by: MHes at 11:50 pm (utc) on Jan. 6, 2007]
My site is back!
Checking on various datacenters they're showing me in the top ten again for my main term. I cant see it myself yet in Google but I'm getting hits on the term so it's showing in some places.
My supplemental pages are gone too, a site:mydomain.com shows my index as the first result and my other pages roughly in order of importance.
Good luck to everyone, I feel like I just won the lottery. Let's hope it stays that way.
MHes and steve8383
I guess its on these DCs you are seeing what you wrote about in your posts, right?
Has anyone seriously considered and detected that the differences between "tweaks" on different pages (and different indices) are phrase-specific?
All datacentres on mc.... data watch.
Ok. Maybe I need an early morning hours Cappuccino, because I don't see that happening on all the DCs for the site(s) I watch :-)
However, Good luck!
Reseller, I tend to focus on [220.127.116.11...]
In the last few days I was fine on a 'UK only' site:search but toast on a worldwide site:search.
In my excitement I thought some other sites were also out but I was doing a 'UK only' search and in the worldwide they are still supplemental heavy.
I'm not counting chickens... I have been here before! However, it has coincided with some critical pages getting cached.... early days.
[edited by: MHes at 12:28 am (utc) on Jan. 7, 2007]
"You seem to think that it is a page penalty which it isn't.... it is related to what keywords are searched for and depending on those keywords, the page will be treated accordingly."
No that is ONE phenomenon.
"This problem only arises when the site has its supplementals above the normal pages."
No that is ONE phenomenon.
""No page is an island", so not only do you need to look at your site as a whole, but also other sites that rank within the search term and their relationship to yours."
Which as I said makes the random factor obvious. This is very easy to see.
"No, that page is failing for certain keyword combinations."
No that is ONE (very rare) phenomenon. Far more often penalized pages rank terrible for anything.
What you see could be signs of some posative "international" changes. Who knows ;-)
Definitely see a new arrangement on 18.104.22.168
Not "fixed" by any means, but different.
Steveb - Are you a 'phenomenon' as well ;)
I'm seeing it on all datacenters, I'm using one of those google dance tool thingies. My stats are showing incoming hits from my keyword with the top ten Google page being the referrer.
|The big question is, what triggered the site to go supplemental heavy in the first place? |
Good question. And the next question is why should this increase in supplemental pages cause a non supplemental page to be placed in position # 950?
Many pundits say each page is separate unto itself, indexed and ranked on its own merit, its not about sites, its about pages. But it would certainly appear this new wrinkle indicates otherwise. This appearance of supplemental pages, and in particular ones that show at the top of the results when running a site command, are either causal, or associated to sites getting sent to position # 950.
|Steveb - Are you a 'phenomenon' as well ;) |
In fact steveb has a PhD in "Understanding Google Auxiliary Index" :)
Do you have a .uk or a .com site?
| This 205 message thread spans 7 pages: < < 205 ( 1 2  4 5 6 7 ) > > |