Forum Moderators: open
Kackle - can you explain the "dictionary" for me? And how I might benefit from it - Im reading your posts hard but dont see where youre coming from.
Sure. But you have to act quickly. Google will fix this one just like they fixed the hyphen.
1. Google is depreciating pages/sites that are over-optimized for certain keywords or keyword combinations. It does this by looking up search terms in a dictionary of target keywords or keyword pairs that it has compiled. This dictionary is Top Secret, because if you knew what was in the dictionary, you could avoid these words in your optimization efforts.
2. If the search term or terms hit on a dictionary entry, the search results for that user's search are flagged. This means that before the results are delivered, the order of the links, or even the inclusion of links, are adjusted so as to penalize pages that have overoptimizated for those terms. Most likely the title, headlines, links and anchor text are examined. It's possible that external anchor text pointing to that page has also been pre-collected and is available for scanning, but this is much less likely. (Besides, external links are not something within your immediate control, so don't worry about it right now.)
3. You want to find out which keywords that are relevant to your site are in Google's dictionary. Compile as many relevant keywords you can think of that searchers might use to find your site. Now take these words singly and in pairs, according to how users might search. Run two searches for each combination and compare the results.
4. If the results are strikingly different for the pre-filter and the post-filter search on a particular term or combination of terms, it means that some variation of those terms has been flagged because something was found in Google's dictionary.
5. Do lots of searches and you can come up with a list of "sensitive" words that you'll want to avoid when you re-optimize your pages.
It's a nice weekend project.
Could you point us to some of this evidence please. Like everyone who is affected I want to know what to do next. To test the idea of a dictionary I wonder if others here could share, what evidence they have found that all/some of this recent Google effect can't be explained by on page text and link back keywords?
There has yet to be a report that noncommercial sites are affected by this filter. And there has yet to be a report that non-English language sites are affected by this filter.
Both of these suggest that there is some initial threshold. You don't have to call it a dictionary. In fact, I'm not at all clear as to whether it's a dictionary of single words, or a dictionary of word pairs, or some combination of these. I don't think it's a dictionary of three-word combinations -- this would get too complex because there are too many possible combinations. However, there may be an initial parsing algorithm so that longer search terms get broken up and two passes are made at the dictionary.
"Dictionary" is just a euphemism that describes the initial decision about whether to apply a filter to the searcher's results or not. Make up a new word if you don't like the word "dictionary." You'll still have to explain the evidence by coming up with some sort of concept to explain the initial decision.
But in fact, lookups in dictionaries are extremely fast for computers. With a dictionary of only 20,000 English keywords, I can do an amazing amount of screening at a speed that boggles the mind, using B-tree lookups. I can keep the entire dictionary in less than half a meg of memory, which is peanuts these days. I've programmed with custom dictionaries, and it's just amazing how few words are really needed to catch 99 percent of what you want to catch.
There are always linguistics experts and artificial intelligence people who insist that Google would never do anything as crass as a dictionary lookup. They're locked into looking at things from the perspective of a particular profession. No engineer has ever gotten a promotion for coming up with the idea of a dictionary. It's too stupidly mundane.
Yet you have to wonder whether some sort of dictionary wouldn't be adequate for a filtering task such as the one we're seeing. And certainly, since the algo is implemented on the fly, Google would be very interested in CPU overhead. If you read their early papers, Google engineers have always been very conscious of CPU overhead and storage space -- down to the point where you save a single CPU cycle or save a single bit in a bit-masked byte of storage.
You don't scale the way Google has managed to scale over the years by doing things the most complicated way possible. PageRank was an exception -- it required too much overhead. I think that's the main reason it is on its way out. Besides, the early engineers couldn't criticize PageRank as being unnecessarily complex. In the early days, I'm sure no one dared tell Sergey and Larry that there were easier ways to accomplish the same thing without spending several days a month computing Pagerank.
i cant even begin to imagine why either brett or gg would say that we would need a 301 redirect from www.somedomain.com to somedomain.com to avoid a duplicate content penalty. i could believe the sky is falling (literally) before i could believe that one. <shaking my head in utter disbelief>
All your conspiracies and concoctions are all null and void.
This is not the index, this is between beginning and end of this update.
Only thing I can see is it helps release some of people’s anger towards the current serps, other then that this is useless.
I too watch what people with over 1000 posts say, because the rest of this is pure Looney tunes.
Im certain you will see them very soon, if not from one of our properties them something similar from elsewhere!
I too watch what people with over 1000 posts say, because the rest of this is pure Looney tunes.
Too easy for google "moles" to jump in and say how much they love the current results. IMHO # of post alone doesnt mean increased credibility.
As for being over - its 11.23.03 - Its OVER two day!
There has yet to be a report that noncommercial sites are affected by this filter.
My editorial content site has been affected. More specifically, several index.htm pages for my "sites within a site" have disappeared from their usual positions (#1, in one case) in the last 36 hours or so.
I posted a hypothesis on why this happened in the "How many front pages did you lose?" thread:
[webmasterworld.com...]
I seem to recall that this may have happened before, quite a while ago, and it got corrected in fairly short order. (And it isn't a disaster, since most of my Google referrals arrive on inside pages anyway).
There hasn't been any recent communication from GG about the current situation, unlike the last big update, when he said more data will be folded in over time - "less than months, more than weeks". Now that is worrying, because we all twiddled our thumbs and waited till they popped back last time. Now? Who knows. According to GG, the data is in there and as every update goes - some sites drop, some gain.
Why would sites MIA suddenly appear just by adding -blahblah after the keywords? Obviously some sort of Scroogle filter is in place.
Has anyone tried stripping out those "poison" keywords from the Title and Body of the index page? - Did it make any difference?
Well at least I can tell my clients that their competitors are gone too LOL.
I think Kackle's point more precisely is non-commercial search phrases as opposed to non-commercial sites. So the question is - are your index pages missing for search phrases that have commercial connotations?
Too easy for google "moles" to jump in and say how much they love the current results. IMHO # of post alone doesnt mean increased credibility.As for being over - its 11.23.03 - Its OVER two day!
What does 11-23-03 have to do with anything?
Yeah # of post do mean a lot of things, experienced members, you'll notice this board has a lot of experienced members but how come they are not posting in here?
Because they know....
That's not exactly true. Searches on non-commercial, non-competitive sites HAVE been affected; the result has just been less dramatic and less negative. The Florida update has caused a shift in rankings and a vanishing of irrelevant porn spam from educational searches I've looked at. I still say the effects are just less disastrous in less competitive searches because there aren't thousands of sites jockeying for position, so a change in the algorithm (or a bug for that matter) won't vault 250 sites over a previously high-ranking one, just two or three. Which, obviously, is an acceptable vagary of fate.
I wonder if this is just a simple miscalculation on Google's part--that is, if they rolled this update out with the successfully improved non-commercial results in mind, not quite realizing the chaos it would inflict on the 5% of most competitive searchterms.
My editorial content site has been affected. More specifically, several index.htm pages for my "sites within a site" have disappeared from their usual positions (#1, in one case) in the last 36 hours or so.
I did see one site that was a *.co.uk that got zapped for optimizing on the term "law essays." But both words are, perhaps independently, somewhat commercial terms. One because of lawyers who advertise, and the other because of college students looking to buy their homework. There's going to be a gray area between commercial and noncommercial. I still think it's generally true that this filter is not aimed at noncommercial sites.
There is a clear intent here to FILTER out pages based on the search term used. The same page could appear at the top of the SERPs under a different search term.
The filter definitely has something to do with the page being over-optimised. I thank people for the -fufuf trick, it helps to explain a lot.
It doesn't matter if it's a one word search term, or a four word phrase, the filter still works. I know this because I have a page being filtered out for a four word phrase, and at least two pages that I know of being filtered out for just a single word.
I have no idea if it's only "commercial" words being filtered, but if Google wanted a list of commercial words and phrases, the obvious source for that list is the Adwords database. Just add up all the money being bid for a particular word or phrase, and then you know how commercial it is. I don't see why Google would program some complicated algorithm that might now work in all circumstances when a very simple query on the Adwords database could figure it out with perfect accuracy.
To the extent that commercial sites are filtered out, to be replaced by sites that have nothing to do with the keywords in question, Google will undoubtedly increase its Adwords revenue.
The danger of "unoptimizing" the site to get back into the free Google results is that you could lose your ranking in Yahoo and MSN and still not get back into Google. You'll be throwing out the baby with the bathwater.
