10

Why on the first page Google says there are thousands of results but on the last page there are less than a hundred?

That's kind of a recursive Google Search that should be self-explanatory. It did take me some time to tweak the sentence and trying to find a number which would correspond precisely, the reason why I changed it to "hundreds" instead! :P

If you have used Google long enough you certainly have stumbled on situations similar to this, many times. Whenever it shows less than 10 pages it's quite easy to see it. The point is: the number of results on any page except the last one seems like a terrible estimate.

What's even weirder, some random times the estimate suddenly becomes much better and almost exactly the same on all pages than it is on the last page - the only page that's always correct.

So, anyone knows why?

  • 1
    Right now on page one for that search it says 3200 results, but on the last page it says 75 (for people interested in an update) – dkuntz2 May 1 '11 at 2:25
  • @DKuntz it's 2990 and 65 to me (while I'm logged in) while if I sign out (which is how I did it the first time) it turns into 3120 and 77. Maybe later I'll change the question slightly and build up a more consistent recursive query - but anyone is welcome to do it as well! :P – cregox May 1 '11 at 5:12
  • Google is weird. – dkuntz2 May 1 '11 at 13:45
6

I don't know the exact answer - perhaps no one outside of Google knows. But I have two data points which may be useful.

The first is XKCD's take on the issue:

So it really looks like there’s a certain threshold of result volume beyond which Google apparently says “screw it” and throws out a gigantic number. I imagine this is probably due to incompetence rather than intentional deception; I’m sure it’s hard to generate pages quickly from many sources, and maybe for searches with a lot of results they don’t have time to get it all synced up. So they fudge the numbers. The fact that this makes it look like they have way more results than they do is presumably just an unintended bonus.

The second data point is from a computer science professor I had who was quite curious about the issue and bothered some friends at Google about it. They told him that sometimes it becomes very expensive to figure out how many results were really returned from a search, so they just make a best guess and return the first page of results, which is all most people care about. Only when you actually start drilling down through the pages does Google bother to fully calculate the remaining pages.

So neither of these is a definitive answer, but hopefully they will be useful at explaining the kinds of issues that may be causing Google's inaccuracies.

EDIT:

Sathya's answer on this page has an answer from Google's Webmaster KB.

When you perform a search, the results are often displayed with the information: Results 1 - 10 of about XXXX.

Google's calculation of the total number of search results is an estimate. We understand that a ballpark figure is valuable, and by providing an estimate rather than an exact account, we can return quality search results faster.

In addition, when you click on the next page of search results, the total number of search results can change. In this case, we realize that some of the query results are duplicates, and collapse those duplicates so that you can find the specific result you're looking for more easily. Collapsing the duplicates decreases the estimated number of results, as well as the overall number of results pages.

  • While I love XKCD and deary former teachers, what you're telling me is that I should probably ask this on stackoverflow.com :P – cregox May 2 '11 at 4:15
  • I don't think anyone at stackoverflow will have a definitive answer - if you really need to know, you should ask Google. I can only relay the answer from someone who asked Google a few months ago. – dsolimano May 2 '11 at 12:46
  • not looking for a definitive answer, just one with a reasonable deeper explanation, preferably good sourced. Don't get me wrong yours is reasonable, but saying "because it's hard" even if it is authoritative and makes sense, is too little to me. Hmm... maybe skeptics.se :o – cregox May 2 '11 at 15:44
  • Nice edit, now that's good sourcing! Though it's still saying "because it's hard", which means I'm still hoping for more details. :P – cregox Jun 21 '11 at 14:03
7

No - the number is just an approximation.

When you perform a search, the results are often displayed with the information: Results 1 - 10 of about XXXX.

Google's calculation of the total number of search results is an estimate. We understand that a ballpark figure is valuable, and by providing an estimate rather than an exact account, we can return quality search results faster.

In addition, when you click on the next page of search results, the total number of search results can change. In this case, we realize that some of the query results are duplicates, and collapse those duplicates so that you can find the specific result you're looking for more easily. Collapsing the duplicates decreases the estimated number of results, as well as the overall number of results pages.

  • I think this is not a good behavior if they always limit the number of results to more or less 1000 only. In case user want more they should show more. In my point of view, some common keywords should return really large amount of search results (such as flower, book...) and I think tons of different pictures are existing out there on the Internet! – user11656 Jun 21 '11 at 9:58
  • @user11656 you're assuming that Google Image Search works the same way as a search in a standard app you might use. Things typically work very differently at that scale. There likely isn't a single answer for "all images that match flower" as a server in Australia might have a recently uploaded image that has not yet propagated to the server in the US. Then you have to deal with things like how to present duplicates, and how to do all of that across millions of images for millions of users instantaneously. Its hard stuff and you have to make architectural compromises. – George Mauer Apr 7 '17 at 17:32
3

Google result counts are a meaningless metric. Aside from the collapsing of duplicates, you also have stemming, the way that the counts are calculated in the first place by multiplying frequencies, and the fact that Google imposes a cap that sets the maximum number of results that will ever be returned.

2

None of the answers above are correct.

Google's estimate must be closer to the real number of results than what is given.

This can be shown by a simple example. Pick a somewhat frequent word like "Russia" or "michael". In truth there must be thousands, if not hundreds of thousands, of websites on the internet that contain those words someplace on their site. But Google results might only give you, say, 700.

The truth is not that Google gives vast overestimates, but that now Google vastly limits the number of results that it will give you. It abridges the results severely, to our major detriment as individuals. It is unfortunate because I do want to read the thousands of sites with my terms in some cases.

I know this because I have used Google for a long time, perhaps 10-15 years and have noticed that as a general rule, the results for terms have gotten shorter, not longer, even though the number of sites with the same terms generally must have increased in the time that I have been using Google.

Whereas in the past, years ago I could get perhaps 1000 results for some terms, now I am getting 500 results for the same terms that I would have gotten 1000 results for.

  • 2
    Please note that "above" has no context in answers, since they can be sorted in several different ways. Even if you assume the default sort by "votes", their positions will change as up- and down-votes are added. – ale Sep 4 '15 at 11:10
  • Very interesting! Russia gives me right now only 322 results and it only shows so on the very last 34th page. Up to then it would say 799 000 000 results! While this is a nice insight it still doesn't seem right. I don't think the first estimates are correct either and the intention is bringing only results that are shown, not "how many it thinks exists on the web", hence, last page fixing it. – cregox Sep 9 '15 at 11:08
1

The issues with Google (Image) Search engine is that it aims for relevance by dropping variance. Firstly, it's important to understand how it works. An image uploaded to the internet needs to be indexed in two ways:

  • based on originality (image gets to be analyzed by a "smart bot" based on color, size, patterns, shape recognition, type, etc.)
  • based on similarity (image gets to be analyzed by a "smart bot2" and cross-referenced with already existing images in a database and further tagged with one or two category labels: "image is match" and "image is similar to...")

After indexing is done, the image will inherit "keywords" as a result of recognition to avoid showing off-topic results to enduser. The thing is that each keyword is a standalone phrase with assigned value in % based on the strength of relevance (that's why grouping more words in search box results in fewer results and also when you search for an image by uploading it or entering URL, Google will assign a "guess" with only top keywords, which means that the search will never show you all pictures simply because the search is not designed to do so)

So at the end of a day when you perform an image search and right under the search box, you will see a tremendous number saying "25,270,000,000 results (0.55 seconds)" its mostly always kind of a fake number, because you will get no more than 200 (max 500 but only after user requests it) results which are also filtered to exclude:

  • external duplicates (if the same image is 2+ times on the same site page)
  • relevance duplicates (shows only "best results" - see image below)
  • images that have issues with the law (see image below)
  • images that violate DMCA (see image below)
  • images that originate from SPAM, MALWARE, PHISHING sources
  • images that are hidden (if a user didn't disable Safe Search)
  • images that are blacklisted by Google (article)
  • images with different AR (basically all images that have a different aspect ratio than the image search reference. eg. you can miss all desired results if you search by your chosen image - let's say with weird ratio 9:7 - while all images indexed in Google are with ratio 3:4 - which is maybe the biggest downside of this search engine because it always obeys Aspect Ratio as the first criterion)
  • and a bunch of more

To summarize it: search results of an image are never full-relevant. Sometimes you will get your final desired result with a search phrase "gate poster" otherwise it could be "gate cover", "gate bluray", "gate dvd", "gate 2015" or even "自衛隊 彼の地にて 斯く戦えり" - you just can't have "ALL" image results, because there isn't anything that is "ALL", that's why there is a need to play with it to get there. And also is relevant to mention, that there are other image search engines that can do the job a way more differently, because they work on different standards and criterions. It isn't and never been "Google"-only... >> https://www.yandex.com/images/

notable extensions:

  • Again, marking these as duplicates would be much more helpful than posting the answer on each question. – jonsca Jan 12 at 1:09
0

At the top it says page 70 out of 7000 results (example). It will give the page number and how many results. There aren't 70 results, but 70 pages. Hope this helps.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.