Wikipedia talk:Google searches and numbers

Comment

Excellent proposal, well written. It all seems fairly obvious to me. Jezhotwells (talk) 17:54, 3 November 2009 (UTC)

Google results are estimates

As is stated at the link "Google result counts are a meaningless metric" it should probably be emphasized more here and at WP:GOOGLEHITS that Google hit results are just meaningless estimates. -- œ^™ 15:39, 3 December 2009 (UTC)

Also this and this (and older posts linked from there) show that Google hit counts are sometimes are logically absurd. A. di M. (formerly Army1987) (talk) 23:34, 28 July 2010 (UTC)

Any objection?

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.

This has been on proposal to become a guideline for a long time. If there are any objections, speak now. Stifle (talk) 09:46, 11 August 2010 (UTC)

I think it's RfC time... œ^™ 06:57, 17 August 2010 (UTC)

• Support Long overdue, IMO. --Trevj (talk) 07:37, 25 May 2011 (UTC)

• Support Drsmoo (talk) 00:26, 13 December 2011 (UTC)

• Support pbp 15:13, 19 July 2012 (UTC)

• Support useful to cite in deletion discussions, since many do argue that something is notable (or not) on the basis of GHits alone. This essay explains why something may not be. Sebwite (talk) 21:13, 19 July 2012 (UTC)
• Strongly Support See below Zaldax (talk) 19:28, 27 July 2012 (UTC)

RfC

I'm wondering if there is enough of a consensus to bump this up from an "AfD argument to avoid" to a guideline. Above, there are five supports, no opposes. But, for it to become a guideline, we'd need a few more eyeballs on it. Hence this RfC pbp 15:29, 20 July 2012 (UTC)

• Support. — Dmitrij D. Czarkoff (talk) 01:20, 21 July 2012 (UTC)
• Oppose I think the essay goes too far. While a google search alone is unlikely to ever be enough to establish notability, searches can often add value to a discussion. For example, a google search that turns up nothing can be very useful for analyzing a potential hoax. Someone with a very unique name that generates alot of google hits suggests (but does not establish) notability. The essay also doesn't seem consistent with practice at AfD, implying that finding sources that establish notability is not enough to save the article if its not improved to include them, but that isn't the way its analyzed. Either the subject is notable or they are not, having the sources in the article just allows you to analyze notability without research. While the essay points out how useless a bad search term(s), it doesn't mention the otherside of the coin, that well crafted search terms can return very useful results. Specific problems aside, while the essay provides alot of good advice, I think most of it is not really of the type that really fits as a guideline, and that that does is already covered by other guidelines. Particularly, the most actual actionable part of the guideline is already covered as an AfD argument to avoid. Monty₈₄₅ 20:16, 25 July 2012 (UTC)

  But AfD arguments to avoid aren't guidelines... pbp 23:31, 25 July 2012 (UTC)

• Strongly Support I for one am tired of seeing Google search numbers given as a rational for pretty much everything. We already know that google doesn't give every user the same results, and that the number of hits is an approximation. Making this a formal guideline should cut down on the number of, apologies, arguably meaningless statistics cluttering up numerous AFD discussions. Zaldax (talk) 19:27, 27 July 2012 (UTC)
• Oppose There are certainly times when a Google search is useful in such arguments. One case is when there are almost zero hits...no matter what ambiguity, a low hit count tells you something about the lack of notability. Another case is when the words don't have other meanings and it's really obvious that one or the other is better. That said, I completely agree that you need to be super-careful about how you interpret this kind of evidence. We need a more nuanced view of how to use raw counts of Google hits...but I don't think this proposal is it. SteveBaker (talk) 19:53, 14 August 2012 (UTC)
• Oppose unless modified Ghits are usually not an argument for keeping or deleting an article, but sometimes they are, because sometimes popularity is a factor, even though it is not the basis for a decisions. If we use this at all as a guideline, we need to say also that it applies both ways: the absence of ghits is not a factor disproving notability, -- except that sometimes it is, when something if notable would be the sort of thing that certainly would have many ghits if it existed and is important. And Google Scholar hits are not the least meaningless--they are by far the most accessible measure of number of publications and citations to them for recent scientists. Various studies including some by my former student Nisa Bakkalbasi et al [1] [2] have shown them generally consistent with Web of Science and Scopus, though the numbers are usually higher because GS covers a wider range of document types. DGG ( talk ) 21:06, 14 August 2012 (UTC)

  Could you please link to any policy or guideline that permits judgment on for unverified factors? — Dmitrij D. Czarkoff (talk) 21:12, 14 August 2012 (UTC)

  WP:NOTBUREAUCRACY, WP:COMMONSENSE...basically WP:IAR (which is a pillar)...all intend that good Wikipedians exercise judgement. That is why we don't need a rule stating that ghits are always a bad criteria - but (if anything) a rough guideline that says that we should exercise careful judgement over when to use or ignore ghit data. Yes, we need to use judgement - that is why Wikipedia has to be written by humans and not by some automatic WP:RS-gathering bot. SteveBaker (talk) 15:06, 15 August 2012 (UTC)

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

opinion

for me, ebe123 I think that the google search is a good idea but it would be nice if you could say if or not that you want it. Maybe that there could be another button for searching that would be for google. I'm in support of the idea. ~~Awsome EBE123~~(talk | Contribs) 22:09, 24 February 2011 (UTC)

Not constructive enough to be useful

This essay does not seem to be constructive enough to be useful. The heading "Why are Google results not valid" suggests that any and all Google results are totally worthless. If you don't know how to use Google, or any tool, then the results are unlikely to be useful. However, if you read "What a search test can do—and what it can't" you will learn that you need to page to the last page of search results to get an accurate number, and "Notability" explains how to search for "Madonna of the Rocks" and exclude the singer Madonna. "Not all websites are reliable sources" is true, but "Search engine expressions—examples and tutorial" explains how you can search for results just in the reliable sources that you specify. People who hate typing long search strings can use templates, as illustrated in the examples. If you take the time to learn how to use Google, it can be very useful and very accurate. LittleBen (talk) 16:16, 24 November 2012 (UTC)