Talk:Numbeo

Biased current article
I have asked ChatGPT to generate the article about Numbeo, this is one of the version with chapters:

Overview

Numbeo is a crowd-sourced database and online platform that provides information about the cost of living, quality of life, crime rates, health care, traffic, and other indicators for cities and countries around the world. The website was launched in 2009 by Mladen Adamovic, a Serbian-born computer engineer and economist, and is based in Belgrade, Serbia.

Features

Numbeo allows users to compare various aspects of life in different cities and countries by providing data on a wide range of topics. These include:

- Cost of housing, transportation, food, and entertainment - Pollution levels - Crime rates - Traffic congestion - Health care quality

Users can contribute to the website by submitting their own data on these topics, which is then used to update and improve the database. One of the unique features of Numbeo is its crowd-sourcing model, which allows for more detailed and up-to-date information than would be possible through traditional sources.

In addition to its data-gathering features, Numbeo also provides tools for users to analyze and visualize the data. Users can create custom charts and graphs to compare different cities or countries, and can also use the site's cost-of-living calculator to estimate the cost of living in different locations.

Methodology

Numbeo's data is based on user contributions and is therefore subject to some degree of variability and potential bias. However, the site's methodology and transparency have been praised by many. The site's founder has stated that the goal of Numbeo is to provide accurate and useful information to help people make informed decisions about where to live, work, and travel.

Numbeo uses statistical algorithms to filter out outliers and ensure the accuracy of the data. The site also provides a system for users to report errors or inaccuracies in the data, which is then reviewed and corrected by the Numbeo team.

Reception

Numbeo has become a popular resource for people who are considering relocating or traveling to different parts of the world. The site's data has been cited by news organizations, academic researchers, and government agencies, and it has been featured in publications such as The Wall Street Journal, The New York Times, and The Economist.

While Numbeo's data is not always consistent with official government statistics, the site is often able to provide more up-to-date and detailed information than would be available from those sources. However, users are encouraged to use Numbeo's data in conjunction with other sources and to exercise caution when making important decisions based on the site's information.

Conclusion

Numbeo is a valuable resource for people who are interested in comparing different aspects of life in cities and countries around the world. Its crowd-sourcing model and user-friendly tools make it easy to access and analyze a wide range of data. While users should be aware of the limitations of the data and exercise caution when making important decisions, Numbeo can provide a useful starting point for research and analysis.

______________________

Compare this with the current article on Wikipedia. I think it needs NPOV rewrite. Mladen.adamovic (talk) 08:24, 9 May 2023 (UTC)


 * ChatGPT output is in no shape or form remotely acceptable as a source for articles. AndyTheGrump (talk) 08:34, 9 May 2023 (UTC)
 * Aside from the fact we absolutely do not write articles with ChatGPT —  how is this actually more neutral than the current article? It appears to have removed the bulk of criticism and written it in more promotional manner. It's less neutral than the existing article. — Czello (music) 08:53, 9 May 2023 (UTC)
 * It also cites no sources whatsoever. Though I'm sure if you asked ChatGPT nicely it could invent some: it's quite good at this. AndyTheGrump (talk) 08:58, 9 May 2023 (UTC)
 * If you need sources for something, you can try to find it, right? Mladen.adamovic (talk) 09:39, 9 May 2023 (UTC)
 * That's working backwards. We don't write an article with a certain objective in mind (that is, removing criticism) and then try to search for sources to fit that afterwards. Instead we construct the article based on sourcing, which is how we have our current article. — Czello (music) 09:41, 9 May 2023 (UTC)
 * First, ChatGPT article has chapters Features that describes what features Numbeo has, it has chapter about Methodology, the chapter about Reception. The current article doesn't have none of those chapters. Mladen.adamovic (talk) 09:36, 9 May 2023 (UTC)
 * Wikikipedia does not base article content on BullshitBot-generated promotional pap submitted by the subject of the article. AndyTheGrump (talk) 09:39, 9 May 2023 (UTC)
 * So why the current Wikipedia article cites what blogger Ray Woodcock (who the hell is that guy?) states about Numbeo, but it doesn't mention what The Wall Street Journal, The New York Times, and The Economist stated about Numbeo. Does that represent neutral point of view? Mladen.adamovic (talk) 09:41, 9 May 2023 (UTC)
 * Provide citations. AndyTheGrump (talk) 09:43, 9 May 2023 (UTC)
 * There are multiple citations by The New York Times https://news.google.com/search?q=site%3Awww.nytimes.com%20numbeo&hl=en-US&gl=US&ceid=US%3Aen I cannot put all of them. Same applies to The Wall Street Journal, The Economists, etc. Mladen.adamovic (talk) 16:05, 9 May 2023 (UTC)
 * You know what, this is a valid point. That source wasn't reliable - I've removed it. — Czello (music) 09:43, 9 May 2023 (UTC)
 * there is more that needs to be done, the remaining sources needs to be checked out. See bottom: Clarification on Criticism of Accuracy - Sources Mladen.adamovic (talk) 15:58, 9 May 2023 (UTC)
 * "More" doesn't mean "better". — Czello (music) 09:42, 9 May 2023 (UTC)

Why the current article doesn't state entities that are using Numbeo and for what purposes?
As per the title. Mladen.adamovic (talk) 09:44, 9 May 2023 (UTC)


 * Provide citations to reliable independent sources. AndyTheGrump (talk) 09:45, 9 May 2023 (UTC)
 * Sure. Mladen.adamovic (talk) 09:48, 9 May 2023 (UTC)
 * https://news.google.com/search?q=numbeo
 * This is for example Forbes articles that are using Numbeo as the source:
 * https://news.google.com/search?q=numbeo%20forbes
 * You can find other articles from any other newspaper of your choice. Mladen.adamovic (talk) 09:50, 9 May 2023 (UTC)
 * https://news.google.com/search?q=numbeo
 * This is for example Forbes articles that are using Numbeo as the source:
 * https://news.google.com/search?q=numbeo%20forbes
 * You can find other articles from any other newspaper of your choice. Mladen.adamovic (talk) 09:50, 9 May 2023 (UTC)


 * No. What we want is in-depth coverage actually discussing Numbeo in independent sources. AndyTheGrump (talk) 09:52, 9 May 2023 (UTC)
 * So I did look into the first article in Google News, https://africa.businessinsider.com/local/lifestyle/top-10-african-cities-with-the-highest-cost-of-living/d7lfn55
 * It states the following:
 * Numbeo uses user inputs and manually gathered data from reputable sources (such as grocery websites, taxi company websites, websites run by governments, newspapers, other polls, etc.) to gather data. Twice a year, manually gathered data from reputable sources are entered. The Numbeo website provides a more thorough explanation of the process.
 * Numbeo has created a methodology to determine the cities with the highest cost of living, and below are 10 African cities with the highest cost of living.
 * How you define discussion? Mladen.adamovic (talk) 09:58, 9 May 2023 (UTC)
 * It doesn't have to be the discussion although. According to Wikipedia: All encyclopedic content on Wikipedia must be written from a neutral point of view (NPOV), which means representing fairly, proportionately, and, as far as possible, without editorial bias, all the significant views that have been published by reliable sources on a topic.
 * All the significant Views that have been published by a reliable sources on the topic should be included. Mladen.adamovic (talk) 10:07, 9 May 2023 (UTC)
 * All the significant Views that have been published by a reliable sources on the topic should be included. Mladen.adamovic (talk) 10:07, 9 May 2023 (UTC)
 * All the significant Views that have been published by a reliable sources on the topic should be included. Mladen.adamovic (talk) 10:07, 9 May 2023 (UTC)


 * I asked for sources discussing Numbeo in depth. Not a source discussing the cost of living in African cities. Per Notability giodelines, for any article to exist at all, there must be evidence of significant coverage in reliable sources that are independent of the subject. Coverage of the subject itself. Not sources discussing something else. The subject itself. In this case, Numbeo. Provide articles on Numbeo. AndyTheGrump (talk) 10:10, 9 May 2023 (UTC)
 * So is some of these articles deep enough?
 * https://www.makeuseof.com/tools-estimate-cities-cost-of-living/
 * https://www.euronews.com/travel/2022/09/24/food-healthcare-and-restaurants-compare-cost-of-living-around-the-world-with-this-tool
 * Scientific articles indexed by Google Scholar:
 * https://www.sciencedirect.com/science/article/pii/S0264275117305905
 * https://www.inderscienceonline.com/doi/epdf/10.1504/IJMDA.2016.081078 this article mentions Numbeo 21 times, I guess it is relevant content.
 * https://www.mdpi.com/2227-7390/10/12/2091 mentions Numbeo 69 times Mladen.adamovic (talk) 11:00, 9 May 2023 (UTC)
 * Also, the current article cites source Walsh, Alistair (2012-08-29). "Harare has poor healthcare and political violence, but property is relatively inexpensive". This is about Harare not about Numbeo, should it be removed?
 * Also, these current citations are not about Numbeo but about Bradford or Yorkshire
 * Meek, Natasha (19 August 2022). "Bradford is Europe's 'most dangerous' city? Police force slams viral claim". Telegraph & Argus. Retrieved 23 September 2022. - not about Numbeo per se
 * Mwitumwa, Monde (20 August 2022). "Yorkshire city is 'most dangerous in Europe' and police boss is furious". YorkshireLive. Retrieved 23 September 2022. Mladen.adamovic (talk) 11:20, 9 May 2023 (UTC)
 * Also, the article is currently Disputed in this Talk page, while the article itself is not market as Disputed. Mladen.adamovic (talk) 16:00, 9 May 2023 (UTC)

Clarification on Criticism of Accurancy - Sources
The current article states: "In 2017, a Swedish man showed how easily Numbeo stats could be manipulated, by pushing the Swedish city of Lund to become the most dangerous city in the world on the website's "Crime Index Rate" in less than a day. Trulsson added that Numbeo should hardly be considered stats, as anyone can change the data, as many times as they want, in complete anonymity."

The comment: This exploit was closed 2 days within the report. If somebody vandalizes some page it's not necessary mean that the project is innacurate. Where is comprehensive study of accurancy comparing to other sources? If they are no study you cannot conclude that something is innacurate just because someone at specific point of time vandalized the statistics by using multiple IP addresses, Tor networks or whatever. Mladen.adamovic (talk) 15:56, 9 May 2023 (UTC)


 * If there is no study, one certainly cannot conclude that a website based on user-contributed content is accurate. AndyTheGrump (talk) 16:54, 9 May 2023 (UTC)
 * but it cannot be concluded that is inaccurate as well, without the study? Mladen.adamovic (talk) 17:55, 9 May 2023 (UTC)
 * As far as Wikipedia is concerned, the only conclusion that is relevant is that claims about accuracy need to be cited to independent reliable sources, per policy. Along with everything else in the article. AndyTheGrump (talk) 18:03, 9 May 2023 (UTC)
 * There is no proof that The Local is reliable source, and still that article is linked here at this page for several years. Neither, no proof that those claims in the article are correct. I'll now propose the article for deletion, as it seems there is no interest in maintaining this article. Mladen.adamovic (talk) 18:27, 9 May 2023 (UTC)
 * There may possibly be legitimate grounds for deleting this article, but the owner of the website it describes not being able to use it for promotional purposes isn't one of them. AndyTheGrump (talk) 18:43, 9 May 2023 (UTC)
 * You are free to discuss if there are legitimate grounds or not. Mladen.adamovic (talk) 18:49, 9 May 2023 (UTC)
 * Any such discussion would have to start with the question we normally address when discussing article deletion: are there sufficient independent sources available to meet Wikipedia notability requirements? Given that when I asked for such sources earlier, you provided a list of sources you seemed to think met the requirement, we should probably at least take them into consideration. Personally, I'd still say that notability was marginal, but I'd rather see what other people think. Possibly the fact that Numbeo data has been used in scientific studies might tip it over the edge? Like I say, I'm undecided... AndyTheGrump (talk) 18:58, 9 May 2023 (UTC)
 * While the community (if any, since in 12 hours there are only a few people here) is deciding, please go and review that current source Thelocal.se, since it is not in the list of reliable sources, and it states "The southern Swedish city, known for its cathedral, university and quiet cobbled streets, is currently ranked top of website Numbeo’s “Crime Index Rate”, ahead of the likes of Caracas in Venezuela and San Pedro Sula, Honduras." the article is obsolete as by looking into Numbeo website one could conclude that it is not correct. So TheLocal.se article is using incorrect/obsolete information. Mladen.adamovic (talk) 19:04, 9 May 2023 (UTC)
 * The validity of one specific source isn't relevant to notability discussions. Not while other sources offering more in-depth coverage appear to exist. AndyTheGrump (talk) 19:08, 9 May 2023 (UTC)
 * So where is the notability discussion? Mladen.adamovic (talk) 03:25, 10 May 2023 (UTC)

Open Statement (1) by Numbeo to Wikipedia to address Wikipedia's article bias - TheLocal article
The current article states that: "In 2017, a Swedish man showed how easily Numbeo stats could be manipulated, by pushing the Swedish city of Lund to become the most dangerous city in the world on the website's "Crime Index Rate" in less than a day. Trulsson added that Numbeo should hardly be considered stats, as anyone can change the data, as many times as they want, in complete anonymity." It uses the following reference: https://www.thelocal.se/20170117/how-one-swede-made-a-city-the-worlds-most-dangerous-to-expose-fake-stats/

The website "The Local" is not on the list of Wikipedia realiable sources: https://en.wikipedia.org/wiki/Wikipedia:Reliable_sources/Perennial_sources and the article itself:

1) incorrectly states that "The southern Swedish city, known for its cathedral, university and quiet cobbled streets, is currently ranked top of website Numbeo’s “Crime Index Rate”", that could be checked at the Numbeo page.

2) it contains references to broken or withdrawn Twitter posts

3) if crowdsourced page was vandalised for a certain period of time, that happens with Wikipedia all the time, is it the Criticism of Accurancy listed in Wikipedia

4) The article looks like a post of successful vandalism act rather than accuracy study

5) Numbeo perform the action of blocking multiple entries by the same IP address unless very long timespan was used in between, that could be checked at the Numbeo website. Mladen.adamovic (talk) 03:46, 10 May 2023 (UTC)


 * Just a question. What is to stop someone from using a VPN/Proxy to manipulate data? YorkshireiteAcademic (talk) 06:04, 7 September 2023 (UTC)

Open Statement (2) by Numbeo to Wikipedia to address Wikipedia's article bias - LeFigaro Used Source
The current version of the Numbeo page states that: "In 2022, French newspaper Le Figaro used Numbeo's ranking to compare violence and insecurity and titled "Insecurity: Nantes, Paris, Marseille, French cities unscrew in the world ranking of the safest cities" without criticism on the method of calculation, which is based on ‘reviews’ left by anonymous users, even if numbers are not realistic, like saying France is more dangerous than Mexico. While Mexico have seen at least 13 journalists killed in 2022, no one have been killed in France."

1) First sentence seems to be more criticism of Le Figaro, not the Numbeo itself

2) There is no reference to show that there is strong correlation with the number of murder of journalist by the overall crime rate, so by using number of murder of journalists, it is scientifically unknown if someone can conclude that one country has higher or lower crime rates

3) The statement says "saying France is more dangerous than Mexico" without providing any reference to whatsoever (who, what, when, how). Mladen.adamovic (talk) 03:54, 10 May 2023 (UTC)


 * Any statistics are open to bias depending on the source of the statistics. There's nothing we're going to do to get around that.  We also aren't going to get around the fact that crowdsourced material is always going to be open to a level of manipulation, which is illustrated by all the countless times that Wikipedia has been manipulated.  Does that discount the value of Wikipedia?  No, it simply gives a caveat that needs to be clear when using it as a source.  So any expectation that Numbeo should have absolute accuracy or reliability is just silly here and inconsistent to the standards that Wikipedia holds itself to.
 * Is Numbeo open to manipulation? Yes, but as someone that has actually used Numbeo for more than a decade as I've lived in various places around the world, I have found the data to be useful and quite representative of my actual costs in various countries. JohnGB (talk) 12:56, 10 May 2023 (UTC)
 * A 15-year old account with just 15 minor edits, has no user page nor talk page, has somehow found their way into this talkpage to support the creator of website. Seems legit., I hope you know that engaging in off-wiki canvassing (or even meatpuppetry) is considered highly disruptive and inappropriate. Pauline Muley (talk) 13:34, 10 May 2023 (UTC)
 * I really don't know how long I've had a Wikipedia account, but I discovered today that I "need" to have a User page (Why?). What does that have to do with whether a person is allowed to have an opinion?
 * I, too, like Numbeo, though I agree the article isn't NPOV. Of course one should use the "grain of salt" with anything crowd-sourced, but there isn't much out there comparable to Numbeo.  The only one I'm aware of is Expatistan.  Numbeo makes it quite plain that it is crowd-sourced. 伟思礼 (talk) 15:53, 7 June 2023 (UTC)

Proposal to delete See Also content of vague relevance
The article contains See also: Boosterism, Deception, Propaganda, as the relevance of those topic is not detaly explained, I propose deletion and cleaning up of the page Categories. Mladen.adamovic (talk) 05:23, 10 May 2023 (UTC)


 * Agreed. I've removed them. AndyTheGrump (talk) 13:51, 10 May 2023 (UTC)
 * I agree that Numbeo’s data shouldn’t be taken as gospel. Neither can Wikipedia articles, or any other crowd-sourced items.  But an article that is 100% condemnatory, almost hostile, is hardly a “neutral point of view.”  I know of no other source as easy as Numbeo to get such an approximation.  I compared the customized view of my own city to my own expenses to get an adjustment factor I can apply to any other place.  I think the tone of the article needs to be softened.  Would be nice if any detractors can offer another source even half as easy to use as Numbeo.  伟思礼 (talk) 17:28, 11 May 2023 (UTC)

Not purely crowd-sourced
removed information about how the data is not purely crowd-sourced because it is also manually gathered in this revert. As the article now only describes Numbeo's data as being crowd-sourced when it is not exclusively crowdsourced, the article gives a partial explanation of where the data comes from. Sourcing is adequate per WP:SELFSOURCE, as I have explained on my talk page. I contest this removal. The content should be restored.—Alalch E. 10:34, 3 June 2023 (UTC)
 * I've restored it with more supporting secondary sources for every claim except for the claim about how the manually inserted data is weighed differently, which is sourced directly to Numbeo, but it is not a self-serving claim, and provides a useful explanation for how the data is combined (the reader certainly wants to understand what would the point be in manually inputting data into a crowdsourced database only for it to be dilluted and made irrelevant due to the predominance of user-generated data).—Alalch E. 13:36, 3 June 2023 (UTC)