Forum Moderators: open
I just checked all 9 datacenters. Only -sj and -fi have the new index.
I can't find anything where they are even close to the quality of the results of two months ago, the same time of the same backlinks. Searches on -sj for eveything I can think of serve up 50% more trash than before.
The real issue though is if the trash results are the result of Google changing its vision, away from pagerank to this almost bizarre fetish for "fresh". It's as if the engineers have locked the content folks in a closet or something. Hello googleplex, just because you *can* do something doesn't mean you *should.*
The deepcrawl is the foundation of Google's success. Find everything, rank everything, judge everything, make sites *earn* their search results. That philosophy seems to be abandoned in favor of insta-pages, "fresh" garbage that may appear today and disappear tomorrow that has not been "voted" on at all. Junk that not only hasn't stood the test of time, but hasn't stood the test of a month.
Did Google intentionally botch the deepcrawl two months ago, deliberately going with incomplete data? Has Google decided it can't accurately deepcrawl the Internet anymore? I don't know the explanation, but the fallout from Google not fundamentally relying on the deepcrawl would be a really huge thing. The other engines gotta be giggling themselves crazy at the chance of Google abandoning quality and relevancy in favor of "new".
I don't think anything can be analyzed until its finalized, although its fun to see what the other servers are doing, even if it drives a lot of people crazy.
regards
Mark
I don't think anything can be analyzed until its finalized, although its fun to see what the other servers are doing, even if it drives a lot of people crazy.
regards
Mark
That's an interesting take on things Steveb and you make some good points. Do Google Guy's comments on better "topicality" on SJ refer to the same thing you perceive as fresh fetish?
That botched deep crawl strikes a chord too. My observation is more of a change in the crawling pattern. It looks from here like all of the top level pages of a pyramid were crawled at the normal time and then deeper pages picked up one by one spread over the rest of the month. That would imply a different process of building, or rebuilding the index. It would also explain back links rebuilding at a different pace and in phases as well as a continuous process.
[edited by: Troppo at 10:40 am (utc) on May 15, 2003]
If so how long until the crying ends?
for my new site from March 20 freshbot yesterday got everything which was missing in -fi and -sj i.e. 50% of pages and I find it in -cw today. They are by no means new pages, freshbot seems to have taken up from where -fi and -sj left.
If this is the pattern then I do not need deepbot any more.
New is fine, as long as it doesn't impact quality. There's a balance there of course, and an important one. Whether they have got it right remains to be seen, as we are still pending the update. Things might look an aweful lot different when (if?) the April crawl results are integrated.
I too will be looking closely at the effect of the so called 'fresh' boost. I can see a 'page freshner' tool around the corner - certainly if 'newness' proves to be the potential back door to get ahead of quality sites which you suggest.
However, at the present time, most problems seem to pertain to missing 'newness' - in other words, recent links missing. I remain confident though that we are still in the middle of the road map described earlier (software/algo upgrade, then data/backlinks).
So, I see -sj and -fi with the same poor results, -cw showing better results (what I actually expected from the dance), and all other centers showing the pre -sj serps.
L
L
Endlessly is a bit optimistic but hopefully less stressful for all of us than those awful monthlies.
"If so how long until the crying ends?"
No idea but they'll probably try to reduce things back to subdued sobs and curses before really going public. Looks like part of Google Guy's mission actually.
If Google is looking for fresh results, it's dishing up some pretty stale pages for my keyword phrases.
All I can hope for is that they apply the spam filters soon... Many of the keywords I look at have horrible non relevant links coming up.
It's a shame if things stay this way. I don't mean just for any links I have indexed but because I use and have come to rely on Google myself for searching...
All we can do is what we have always done... Wait and see.
I'm curious to find out when your home page went MIA.
I have had the same problem in the last 24 hours. My home page went missing from the www server. It shows up in sj and fi database though.
Have stressed out all day trying to find out if I have been penalised for something.
Have never had this happen to me before.
How many others are in the same boat as us?
Earl: What's that Lem?
Lem: I'd like to see the end of the Dominic update
Earl: You would?
Lem: Yea-up ... not for me now mind you. I've seen an update end .... but for my grandkids.
(sorry, couldn't figure out how to type a decent primal scream)
www old rank: #1 *
www new rank: #3 **
www-sj: #4
www-fi: #6
* My rank on www for the last month.
** My rank on www for about one in five searches this morning.
Something appears to be happening, but whatever it is, it doesn't seem to be a pure, unadulterated import of -sj or -fi into www.
eh?
-for main single keyword dropped to #349 from (old www:#48, sj/fi:#42)
-this listing has a May 14 fresh date
-cw is showing 5 backlinks (same as sj/fi)
-as site is PR3 it has never shown backlinks before, so I have been speculating about a promotion to PR4 this update (but this #349 ranking has me scratching my head...)
-for 2 and 3 word queries, site ranks similar to last update
For another (new site):
-site is ~1 week old (all listing are "fresh")
-www has 28 pages, some with "May 13" date, some without
-cw has 37 pages, some with "May 14" date, some without
I don't know if anyone has stated this explicity yet, but cw is fresher than www.
[edited by: shaadi at 2:02 pm (utc) on May 15, 2003]
.co.uk uses Inktomi - cant access .com from the UK :(
<edit>Prev said co.uk use ATW when they use ink</edit>
[edited by: Dayo_UK at 2:01 pm (utc) on May 15, 2003]
Hotbot quary 4 different engines, one at a time. When you search the google one, you get new results.
all these people seeing all these changes, but we see the same things we've been seeing, excepting for the occasional momentary switcharoo Google pulls here and there.
we have have recently started to see geographic designations (i.e. (www.google.il) on our stat. manager.
is this something new, or have we just never received traffic from these indices in the past? haha. anybody help on this? thanks!
is it better to just leave my site alone until its finished?
