Wikipedia talk:Search engine test

vandal
68.154.104.124 wrote: First of all I would like to know the purpose of this web page. It seems a little foolish that anyone can type anything that they want to see here! You know what that can lead to...don't you??? No? Ok so what are you waiting for? Go ahead.Try your hand...

Useful contributions OUTWEIGH vandalism by 50:1, so it's worth it. -- Karada 22:54, 1 Aug 2003 (UTC)


 * discussion of this, like any other page should be open. Waveguy


 * See replies to common objections

Alexa
I added Alexa - thinking in terms of a general "external utilities" page. Martin 11:33, 1 Sep 2003 (UTC)


 * After retrieving the results from Alexa, how does one analyse how that subject has performed on Alexa? --Daniel C. Boyer 14:42, 8 Sep 2003 (UTC)

I like the Alexa part, but not in the "fails the Alexa test" way. The Google test is enough imo. BL 18:00, 22 Sep 2003 (UTC)

Alexa is a useful utility for determining the popularity of a website, but I don't think that popularity is a criterion that we should use in deciding whether or not to have an article on something. Articles are built up from verifiable information. Therefore, the decision on whether or not to have an article on a subject should be based on the amount of verifiable information there is about that subject. If a website is popular but nobody writes about it, we probably shouldn't have an article on it, because there are no independent sources to draw on for information. Conversely, if a site is unpopular but a lot is written about it, of course we should have an article on it. I expect there must be a high correlation between the popularity of a website and the amount of verifiable information on it, but that's just a statistical generality. An Alexa ranking does not in itself tell us a thing about the amount of writing about a website in individual cases. So I don't think we should use it. The Google test will give a much better indication of the amount of verifiable information, because it is itself a search for information. So I don't think we should use the Alexa test as a substitute. (Or, in summary: I agree with BL.) -- Oliver P. 03:58, 23 Sep 2003 (UTC)

Alexa has been retired as of May 1, 2022. So? Greenlio (talk) 17:39, 6 May 2022 (UTC)

BuddhaInside-ism
BuddhaInside keeps adding "a quick check that an individual fact, within an accepted subject, passes the criteria of requiring 1000 google hits to avoid deletion."


 * This is not true and should not be added to the page or people will start deleting every fact that does get 1000 hits. Angela 05:44, Sep 23, 2003 (UTC)


 * Did you miss the discussion at Talk:List of nicknames for George W. Bush? That is exactly what Cyan is now doing. -BuddhaInside


 * Having a more stringent rule on a controversial page is different from adding the rule here and trying to apply it to everything. If no-one is objecting on that page, then Cyan can do that, but I am objecting to it being applied more widely than that. Angela 05:52, Sep 23, 2003 (UTC)


 * If should be enforced across all pages equally, or not at all. -BuddhaInside


 * Not if there is a consensus (my definition, not yours) on one particular page to adhere to the rule. There are lots of rules that do not apply to all articles. Angela

As one who edits here without looking behind the curtain too often, perhaps someone could elucidate the origin of these tests. Who first suggested the Google test? (I'm simply trying to better understand the nature of the wiki and how these practices evolve, not questioning the practice itself.) On a related topic, why is this article on the Wikipedia rather than the meta? (I still don't fully understand the relationship, sorry!) -- A Profoundly Perplexed Paige


 * I can't answer the first part, but it's on Wikipedia rather than Meta because it relates directly to editing as it's sort of a tool for checking whether a page should exist. My personal understanding of it is that pages that are for editing are here whereas pages that are about editing are on Meta. Someone else could probably explain that better though. Angela 22:12, 25 Sep 2003 (UTC)

belatedly moved from VfD (sorry, Fridge):


 * Wikipedia - relatively minor website, fails Alexa test. Fridge 20:24, 25 Sep 2003 (UTC)
 * NB. Meant to be a comment about Alexa Test not Wikipedia! Fridge 20:25, 25 Sep 2003 (UTC)
 * Firstly, Wikipedia passes the Alexa test easily. Secondly, the article is about the project Wikipedia, not the website Wikipedia. Wikipedia is not just a website. Angela 20:41, 25 Sep 2003 (UTC)
 * Alexa test says things about sites being in the top 100 and sites in the top 1000 (and also top 100,000). Wikipedia is not in top 1,000 (though might soon). OK Wikipedia is not just a site... but then most websites are not just sites: they are companies and people and projects :just like wikipedia is a project. Anyhow I wanted to say that some sites higher in Alexa wouldn't be in Wikipedia (like porn sites) but others lower than Wikipedia might be so Alexa is not a good test. Fridge 20:47, 25 Sep 2003 (UTC)
 * If you want to discuss the policy - go to the talk page of that policy. Trying to prove a point by listing things here really isn't a good idea. Angela 20:52, 25 Sep 2003 (UTC)

end of moved text

Please note, the policy says:


 * we should have articles on important websites, important being defined as Alexa-ranked higher than some threshold (100 or 1,000)
 * we should not have articles on unimportant websites, unimportant being defined as Alexa-ranked lower than 100,000
 * nothing at all about articles on websites in the fuzzy zone between important and unimportant

-- Cyan 00:56, 26 Sep 2003 (UTC)

Google Scholar
This was copied a couple of times; it was originally on the Village Pump. Maurreen 08:04, 20 Nov 2004 (UTC)


 * Google Scholar at http://scholar.google.com/ looks like an interesting new tool for research into scientific publications on the Internet. It is described as " Google Scholar enables you to search specifically for scholarly literature, including peer-reviewed papers, theses, books, preprints, abstracts and technical reports from all broad areas of research." At first glance, this seems to be far more helpful for finding relevant refences than other search engines. Please try out, and please give feed back. Should Scholar be recommended in our How-to pages? Kosebamse 15:21, 19 Nov 2004 (UTC)


 * Yes, I think it should be recommended. :-) &#8211;&#8211; Constafrequent (talk page) 06:27, 14 Jan 2005 (UTC)

Google test Template?
Does this exist?


 * If it did, what would it do? --Smack (talk) 05:11, 17 Jan 2005 (UTC)

Another kind of Google test
I often use Google to determine the most common form of a word, e.g. 'canoe touring' vs. 'canoe camping'. Shouldn't this be listed? --Smack (talk) 15:58, 6 Jan 2005 (UTC)
 * Whether this is a good idea is somewhat arguable, but no more arguable than the Google Test as stated now. Please be bold. Deco 07:15, 29 Jan 2005 (UTC)
 * I dunno if this matters, but this concept is commonly known as a "Googlefight", popularized by the website googlefight.com. --DropDeadGorgias (talk) 02:50, Apr 12, 2005 (UTC)

POV tool
The "Google test" is not being used as any sort of test at all. I would strongly suggest reducing this to something along the lines of a "Google researching" how-to article for newcomers to Wikipedia who want to learn how to detect copyright violations and then merging it to the appropriate location.

In no way whatsoever has this this "test" ever been evenly applied, even when comparing two articles within the exact same genre or category (or sub-sub-category). To describe it as such is false and misleading. --GRider\talk 21:53, 11 Mar 2005 (UTC)


 * The Google test is a useful tool. It is not perfect, nor is it the only measure of notability. But as long as we recognize its inherrant biases, I see no reason why it can not be used effectively. Your recent edits to this article have tipped it strongly into POV territory. For instance, it is certainly not necessary to list every single instance of where you think the test has been used unfairly. DaveTheRed 22:03, 11 Mar 2005 (UTC)


 * Censorship does not qualify as making something form a neutral point of view. Bastardization of the English language also does not equate with NPOV'ing.  Be advised.  --GRider\talk 22:20, 11 Mar 2005 (UTC)


 * Censorship? How is removing a POV section of an article censorship??? Censorship implies that I am trying to keep the public from knowing that there is a bias to the Google test. This is emphatically not true. I do maintain that your list is POV. It is a list of articles that you happen to think shows the google test to be invalid. The list is inherrantly POV because its inclusion criteria is subjective. I will not revert just now, because I have no desire to enter a revert war. And for the record, your comment about bastardizing the English language is way out from left field. DaveTheRed 22:48, 11 Mar 2005 (UTC)


 * No, calling this a "test" is left field. Raising a red flag is not.  --GRider\talk 23:14, 11 Mar 2005 (UTC)

How can you argue that it's not a test? You can argue that its a biased test, or a frequently misapplied test, but that doesn't make it any less of a test.

To get back to my previous point, I have made the assertion that your list of articles where the google test fails is inherrently POV. If you wish to keep it, please list any reasons you think it is not, and should stay. DaveTheRed 01:03, 12 Mar 2005 (UTC)
 * I agree with Dave on that point. It would be more useful to make some statistics, i.e. what is a 'substantial amount' of googles for a band? And for a company? And for a website? Radiant! 18:42, Mar 12, 2005 (UTC)
 * Actually, I'm kinda against giving statistics, because the google test is so imprecise. How many hits it takes to be notable is in many ways a subjective measure, and one that should be made on a case by case basis. Who's to say that a band that gets 1000 hits is notable, but one who gets 999 isn't? I think we need to stress that failing the google test is not proof that something is not notable, rather it is only circumstantial evidence against it. DaveTheRed 00:07, 13 Mar 2005 (UTC)
 * The fact that you are so vehemently against providing statstics on how many Google hits are needed in order illustrate notability of a subject only further proves the inherent subjective flaws of the so-called "Google test". The Google test is not a test, there is no way to reproduce the same results time after time on VfD because it is not evenly or fairly used based upon the subject.  In fact, if you applied this so-called test on the same exact subject one week and then tried it again 6 weeks later you would very likely have a different outcome.  This is not a test if the answers are different every time.   --GRider\talk 18:14, 15 Mar 2005 (UTC)

Since you have not defended the inclusion of your list, I have removed it again. Before you revert, please give a good reason why the list is not inherrently POV and should stay. DaveTheRed 02:57, 15 Mar 2005 (UTC)
 * They've been reinserted. Citing direct examples of how the Google test fails is what helps makes this article closer to a neutral point of view.  By not allowing in such evidence you are the one guilty of imposing a pro-Google test POV.  --GRider\talk 18:10, 15 Mar 2005 (UTC)

"This is not a test if the answers are different every time."

I must respectfully disagree with your assesment there. That just means its a test that's used in a subjective manner. Perhaps it would help to quote webster:

TEST 1. the procedure of submitting a statement to such conditions or operations as will lead to its proof or disproof or to its acceptance or rejection

Notice that this defintion states nothing about reproducible results. The google test submits a statement (I think X is not notable) to an operation (googling) that will lead to its acceptance or rejection. How is this not a test?

Now, as for your list, It is is a list of instances that Grider thinks invalidate the google test. Tell me how that is NPOV. I don't think any of them invalidate the google test. Some of them demonstrate its biases, but none of them invalidate it completely. The list as is comes off as a petulant attempt to get in the last word on your nominations that failed. Please give me a good reason why they should be kept. DaveTheRed 19:33, 15 Mar 2005 (UTC)

Foreign languages and non-Latin scripts
I think the following needs to be pointed out on this page: Claims for the non-notability of a topic is occasionally made based on few Google hits, where the claim is actually false because a considerably larger number of hits may be found if searching in the correct script or for various transcriptions. An Arabic name, for instance, needs to be searched for in the original script, which is easily done with Google, provided one knows what to search for, but one also has to take into account that e.g. English, French and German webpages will likely transcribe the name using different conventions.

In addition, different forms of a name used in the original language has to be searched for. A Russian personal name has to be searched for both including and excluding the patronymic, and any search for names and other words in strongly inflected languages should take into account that arriving at the total number of hits may require searching for forms with varying case-endings or other grammatical variations not obvious for someone who does not know the language.

Doing a search like this requires a certain linguistic competence which not every individual wikipedian possesses, but the Wikipedia community as a whole include many bilingual and multilingual people and it is important for nominators and voters on VfD at least to be aware of one's own limitations and not state conclusively a small number of Google hits for, say, a Serbian poet without pointing out the limited validity of a preliminary search using only one particular transcribed form of the name. / up◦land 08:11, 15 Mar 2005 (UTC)


 * I agree whole-heartedly. A section like that would be very useful. Feel free to be bold and implement that change. DaveTheRed 08:25, 15 Mar 2005 (UTC)


 * I am waiting to see if there are any objections or suggestions as to the wording of the section, better examples etc. / up◦land 15:32, 18 Mar 2005 (UTC)


 * no objection, I support the idea Tobias Conradi 05:37, 24 Mar 2005 (UTC)

Vfd link
This article was proposed for deletion in March 2005. See Votes for deletion/Google test   Joyous 19:27, Mar 19, 2005 (UTC)

A humorous aside
As of today: cesarb 00:26, 20 Mar 2005 (UTC)
 * "Google test": 9.810 hits, fourth hit is Google test
 * "Alexa test": 309 hits

Disputed?
For what it's worth, after looking over this article carefully I don't believe it exhibits any particular POV problem, unless I overlooked something. Everything it says seems sensible and neutral. If this has been resolved would someone remove the tag? Deco 00:30, 20 Mar 2005 (UTC)
 * Bkonrad has removed the disputed list, so I am currently happy with the article in its current form. I would accept the removal of the neutrality tag. DaveTheRed 05:23, 20 Mar 2005 (UTC)
 * No other opinions; tag removed. Complain in my talk page if required. -- Kizor 21:43, 26 Mar 2005 (UTC)

Quote from Jimbo
I think the Google test is often helpful, but should be used with judgment. I see no way to ban people from doing it, do you? It's useful information which ought to inform our judgment, although of course in many cases there can be reasons to override or ignore it. I trust people to make the right decisions thoughtfully, using the Google test as one element among many.--Jimbo Wales 01:01, 12 Mar 2005 (UTC)
 * Despite the signature, the note above was actually left by User:GRider. I'm not sure what his point is since it is opposed to his remarks above; it would be nice to see what context it was quoted from. Radiant_* 20:25, Apr 7, 2005 (UTC)
 * The context is and . --cesarb 01:27, 8 Apr 2005 (UTC)

Infotiger test
Why do we use the google test instead of the Infotiger test? Isn't it ultimately all just arbitrary? I have created several websites that did not show up on Google. Just because they crawl Wikipedia doesn't mean they crawl everything. The vast majority of the Internet's content will not show up in Google results, in fact. And not everything notable is on the Internet. The phrase "Failed the google test" is a pathetic reason to vote Delete, because, for instance, one school could hire a search engine optimization consultant, spam message boards, etc. and beat out another, more low-tech school (e.g. in Amish country) on the Google test. Are we going to say Delete on the hypothetical Amish Preparatory School for Learners just because Google can't find it? And in any case, is it a good idea to base deletion of someone's hard work on the output of one corporation's software? 205.217.105.2 17:33, 12 Apr 2005 (UTC)
 * We use google because it is a popular search engine. If you decide to vote at the vfd board, you are more than welcome to use yahoo or infotiger or any search engine you please. It is best to keep in mind that the Google test is not the only test of notability. If something fails the google test, that is only circumstantial evidence that it is not-notable, not an iron reason to delete. Everytime something passes the google test, it is necessary to check to see if it looks like the subject in question has been spamming google. It is true that the entire internet is not represented on google. But most of the notable parts are. The google test works on the assumtion that if a subject is notable, then someone somewhere will have mentioned it on a website that google links to. Dave the Red (talk) 18:17, Apr 12, 2005 (UTC)

Articles labelled "PR" or "blog" in Google search results
Hi, I'm seeing a lot of articles, particularly newspaper articles published in media located in developing countries, labelled "PR" or "blog" in Google search results, when in fact that there is nothing to indicate that the newspaper article was a blog/PR article and it is not clear how the Google algorithm has determined this. For example, this article in the Ugandan Daily Monitor is listed as "(press release) (blog)" in the Google search results. At AFD this can lead to these references being dismissed as not reliable or not independent based simply on what Google have published, which appears not to be reliable (the Daily Monitor is a well-established Ugandan newspaper and the article is clearly not a blog and very unlikely to be PR). Can we put something in here to say that simply because Google (or any other search engine) lists something as a PR/Blog article, does not necessarily mean that it is? FOARP (talk) 10:13, 11 October 2019 (UTC)

JSTOR
Does a JSTOR search test count as WP:HITS? Firestar464 (talk) 04:32, 25 December 2020 (UTC)

Assessing Common Name
A "Common Name" section should be added as assessing what the common name of something is is a very common use of search engines here. FOARP (talk) 15:16, 20 April 2022 (UTC)

Retirement of Alexa
Alexa has been retired by Amazon as of May 1, 2022. Greenlio (talk) 17:42, 6 May 2022 (UTC)

Add a template
the page currently does not have the template although it's semi-protected, when will it be added? thanks 103.47.210.34 (talk) 02:37, 3 May 2023 (UTC)
 * ✅ -- DB 1729 talk 04:30, 3 May 2023 (UTC)

"Wikipedia:ALEXA" listed at Redirects for discussion
The redirect [//en.wikipedia.org/w/index.php?title=Wikipedia:ALEXA&redirect=no Wikipedia:ALEXA] has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at  until a consensus is reached. Steel1943 (talk) 20:24, 1 June 2024 (UTC)