3

Our help-center lists explicitly one symptom that makes a question "off-topic":

It is not about:

Questions answered by a simple Google search or to be found in a Wikipedia page

I see 'a problem' there.

First, the capitalised word "Google" is really an advertisement, unpaid, for one search engine. It is not the genericised 'to google something' (meaning searching the net via an/any engine).

It almost follows from concentrating on this 'evil' engine that we disregard two aspects:

  1. What in general is meant with "simple search"?
    (First "I'm feeling lucky"-hit after entering a single search string? First 3–5 hits? First 3 pages of entering up to 5 keywords or the entire actual question into the search fields?)

  2. Results from Google are not universal!
    The results depend on a long list, not entirely public, of parameters.
    It is known that results presented depend on region and/or IP, local censorship or other laws, cookies set in browser (also from other sites?), status of logged-in or not, personal search history and more (company doing experiments).
    Most bluntly: US users see different results compared to EU citizens or compared to Chinese users? We cannot blindly rely on two people from different parts of the world using the exact same search string to get identical results from Google.
    To be clear: other search engines might suffer from this phenomenon as well, although to a much lesser degree. But the point to observe is that two users using Google might see a different list, with different results or in a different order — much like two users using entirely different search engines: dynamically generated content is not a shared and stable frame of reference.

(Apart from making Google results generally less reliable than many might think this also paints the common phrasing in many comments less ideal than they are intended to be: ~"literally the first link on Google" — is really just the same as "what I found". A conclusion from that alone that 'OP didn't search/google' is not necessarily true (although I agree that it makes OP's prior research somewhat less likely to have been done. Again emphasising that this kind of prior research needs more explicit mentioning in any question?)

(Apart from suggesting that we re-word the free ad for Google in the help page; there are alternatives)

What research effort by a poster do we consider "answered by a simple search"?

9
  • 1
    Why would you assume that results from any other search engine are universal? To the best of my knowledge, none of the companies publish their algorithms for ordering search results. With good reason - that's the part that adds value. Jan 20, 2020 at 18:31
  • @sempaiscuba Thanks. A: I/we just can't assume that! It's just experience that let's me assume that the result skewing at G is much bigger than with startpage (which gives you one slightly more univeral access to Gpagerank), duckduckgo, metager, (for eg ecosia, bing, yandex etc I can't even comment on for this). Jan 20, 2020 at 19:10
  • 1
    Better, but why do you still assume it's to a much lesser degree? Any evidence to support that? A search using duckduckgo for duckduckgo censorship got this subreddit as the first hit. A search using Yahoo got this Wikipedia page, and using Bing I got this Wikipedia page. (That's just the most commonly used search engines). Jan 20, 2020 at 19:16
  • Why not simply state that "results from search engines will vary for different individuals, based on geography, search history, etc.". That is my starting point. But if people can't be bothered to tell me what research they've done, then I'm going to assume they get the same result I've found (and in general, the results on someone's 1st page from Google will appear on almost everyone's first few pages of results unless prohibited by local laws) Jan 20, 2020 at 19:21
  • @sempaiscuba I usually start with G, then check with an alternative. Using rotating IPs, noting which you use, you should notice that US/UK/FR/DE IPs gives you different results (especially if you can construct a search to 'locally forbidden content'. Example: German primary source material 33–45 with German IP on G: tough cookies. Switch to alternative engines or different IP: instant paydirt.) I only single out G in Q because we do so in help pages… Am unsure whether your last comment should go into my Q or your A? If you feel strongly about "my Q" then feel free to edit it directly. Jan 20, 2020 at 19:30
  • My observation above was just about removing your apparent bias against Google. People are more likely to engage with the question if the language is neutral. But it is your question. Jan 20, 2020 at 19:34
  • Oh. For that, I am biased against all search engines, I believe. We should not be biased so much for Google in our help and comments. But I might refactor this eventually nonetheless. (In a few hours. The low view count so far let's me infer that the body text was so far not much of a detriment, btw. I thought of not much evil in my words when writing, and I would be fine to re-use yours and find them in it anyway. Neither smearing nor hagiographing G, just reflect on how we use it.) Jan 20, 2020 at 19:41
  • From my observations, Duckduckgo will always provide a result that's not what you were looking for. But a good example of Google was here when I searched for something that I knew about and couldn't find the specific term, while Lars or someone else was able to give me that specific idea without any problems. I even searched both on Wikipedia and Google and had three out of the five terms in the title right, but still that page didn't come up...
    – gktscrk
    Jun 9, 2020 at 18:42
  • @gktscrk Yeh, depends of course. The assumptions that build ddg are sometims really boring… GG is usually the fastest and 'best', for what you/I want; market leader for quite the reason here. But: Try a search with an IP where you know that GG is instructed or wants to censor certain content, or abuses your history & cookies to skew results too much. Then having an alternative should prove its worth quickly. Recently, GG and bing were becoming too cocky for my tastes and a metacrawler made for an amazing change of perspectives in results. Jun 9, 2020 at 22:04

3 Answers 3

3

Here are a few suggestions (the first of which is from a comment you posted here):

  • A single link answer found on the first page of a search should be strong candidate for 'too basic'.
  • Include the other main search engines rather than just 'Google', or don't mention a specific search engine (or use the verb google).
  • Let's not forget that not everyone is an expert at searching, so let's give a specific instruction. My suggestion would be: Please google your intended title question and check the results before asking here.
  • Put our criteria for legit questions in the yellow box on the Ask a question page. They are much more likely to be seen there as people asking questions have to visit this page.
  • Use clear, concise bullet points and keep the guidelines as brief as possible (the longer the guidelines are, the less likely people are to read them).

Concerning the problem of different users seeing different results based on location, cookies etc., I'm not really sure what the solution to that is other than to ask users to say what they found (as suggested by sempaiscuba in a comment here).

3

Caveat - as above, please ignore any diamonds you may see.

I acknowledge all of the problems you cite with googling. No contest.

If the OP states in the question that they attempted to search for an answer, I'll trust that assertion. But I believe the overwhelming majority of questions that are closed as too basic on the grounds that OP did not attempt a search of standard references are correctly closed on that basis.

1

This is a personal view. Please disregard any mod diamond that you might see.


A 'simple search' is just one which does not rely on the use of additional qualifiers to return the results you are looking for.

If I enter the title or main keywords from a question into a search engine, and it returns a link on the first page of results that appears to answer the question, then that question is 'too basic' according to the definition in our help centre.


From my perspective, your second question is linked to the issue of documenting prior research in questions.

Writing an Internet search engine is almost trivial. I've done it, and so have the (thousands) of other students that completed the same (free) course on Udacity.

The difficult bit is sorting the (potentially millions of) results that you obtain.

Google accounts for between 75% and 95% of searches because they do that part 'better' than other search engines (which is precisely why their algorithm isn't public!).


I suspect that whoever wrote the text for our Help Centre (it was before I joined SE) used the capitalised form of 'Google' simply because Google are the dominant search engine.

(I tend to use Google for history-related searches because I find that it returns the results I'm looking for more reliably than any other search engine that I have tried.)


The reality is that, if the OP hasn't documented their research, then the most likely scenarios are that either

  1. they haven't done any research, or that
  2. any Internet searches they did were carried out using Google.

But for our purposes here, that is still largely irrelevant.


For me, the 'acid-test' is this:

If I enter the keywords from the question into Google, or any other search engine, and find a result that appears to answer the question, I'm generally going to assume that would also be the case for the OP, unless they have documented their research to show that they did search and didn't get the same hits on whatever search engine they are using. Their question is thus off-topic, according to the definition in our Help Centre.

In cases like that, it should be up to the OP to show that the question is, in fact, on-topic.

5
  • Conceptually, I'd like to add that Google has a very good pagerank algo, but we can't use 'it', because of all the filters massaging over it. Remove forbidden, remove geoIP, push commercial (but punish some spam), weigh political bias, create a user-specific-bubble, etc. Then the main reason why you can't explore anything on Youtube autoplay: monetise no matter what. No doubt the Google algo is 'good', and no doubt that it is abused beyond our influence to our detriment. We can't dismiss it, but we shouldn't push it further, or uncritically. Jan 20, 2020 at 12:07
  • @LаngLаngС Well, that is an opinion (and not a particularly surprising one, given your characterisation of the Google search engine as "evil"). Those "filters" are part of the algorithm, and generally (with the exception of China) just reflect the user's online history. I have no strong opinion about whether we use 'Google, 'google' or 'search engine'. I just expect users to carry out and document some research before posting questions, because that is required by our site rules. Jan 20, 2020 at 12:51
  • The "evil" shorthand may look primitive, but 'China' is also just a small subsetted item (& dysfunctionally prejudiced?) on a long list even WP scratches. On SE we often see posts relying on Gmain or Gbooks, concluding "not there, ergo doesn't exist, for sure". My goal here would just be to praise google only where it's due and criticise it when that applies, making users aware of limitations & alternatives. (NB,IE: mainly agreeing on "document any research!") Jan 20, 2020 at 13:11
  • @LаngLаngС Silly me. I thought your goal here was to clarify when we should be closing questions because they can be "answered by a simple Google search". Jan 20, 2020 at 13:21
  • It still is, kind of. The title above the Q left out the word 'Google' because that causes imo additional problems. Only these additional problems are the goal here (in that comment). What I want to clarify overall is how we phrase and advise for "simple search". Jan 20, 2020 at 13:27

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .