The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's
talk page or in a
deletion review). No further edits should be made to this page.
There used to be these two now deleted articles
Wikipedia:Articles for deletion/List of localities in England by population and
Wikipedia:Articles for deletion/List of largest United Kingdom settlements by population. There are also articles for each ceremonial county in England on the same subject based on the same data set and they have the same problems. They have no sources outside the primary ones. The source of which doesnt even split them out by county (many of the areas cross county boundaries) so any county splitting is Original Research. They also use the word settlements which the source doesnt mention at all referring to them as built-up subdivisons. These aricles are misleading.
Eopsid (
talk) 21:22, 21 January 2022 (UTC)reply
Delete all. Beyond the sourcing question, these articles have generally not been substantially updated for several years, and the information in them therefore can not possibly be accurate, since population fluctuates year after year. There is no current prospect that these can be kept up to date for the indefinite future.
BD2412T 03:45, 22 January 2022 (UTC)reply
Fix and keep. I created a number of these originally. Those sources at the time did forward to a county specific list of settlements. What's not being realised is that there have been changes on the source website since and so now the data is not on the landing page anymore. But if you drill down into the website, the data can still be found. The banner above states we consider alternatives to deletion, simply locating and updating the references will correct the problem.
On this page it does refer to the table as a list of Settlements, and there is a status column explaining whether they are
BUAs or sub divisions.
As to
BD2412's concern about outdated stats, the census is done every ten years which are official counts, and anything outside of those is an estimate. 2021 was the latest census and so up-to-date figures will be imminent.
The fact that interim data would be "an estimate" doesn't fix the speed with which periodic census data becomes outdated. If the updated title reflected "as of 2021" this would be more accurate, but I would still question the encyclopedic value of such a snapshot in time.
BD2412T 05:49, 22 January 2022 (UTC)reply
Eopsid, CityPopulation.de was used at the time because Nomis didn't have direct links to census area data in those days and the Neighbourhood Statistics census site which did was decommissioned. That London link seems to have been superseded by the agglomeration list
https://www.citypopulation.de/en/uk/agglo/E34004707A__london so again I think site pages have been 'improved'. The data is now better accessible through a primary source, and
Template:NOMIS2001 /
Template:NOMIS2011 have been developed to take advantage of that, so it would be a matter of updating the refs.
BD2412, usually the article refers to the shortcoming of the census data in the prose of the articles, and estimates are really only given for main settlements (because the data is only provided to district level, but not towns/parishes/settlements as it's never been that granular. There is talk of using other gov data sources to create future censuses because of the expense
https://www.bbc.co.uk/news/uk-51468919, that methodology is more in line with how estimates are created, but until that is accepted practice I personally think that decade long difference gives the best insight into the growth or not of a settlement and smoothes out any irregularities. But the articles are definitely notable as there will always be interest in what are the largest settlements.
The NOMIS data doesnt split it out by county. So splitting it by county would be Original Research. The NOMIS data also doesnt refer to them as settlements but as built-up area subdivisons.
Eopsid (
talk) 14:27, 22 January 2022 (UTC)reply
That link doesnt mention Built-up area subdivisions at all. I will have a deeper look into the NOMIS data to see if it does show counties.
Eopsid (
talk) 19:35, 24 January 2022 (UTC)reply
The maps are all based on Open Street Maps which isnt a reliable source, its basically another Wiki.
Eopsid (
talk) 19:16, 25 January 2022 (UTC)reply
Yes Wiki too then - nominally unreliable as it too can be changed easily. Yet we are here fighting to make sure it's relevant and the best it can be. So too at OSM I bet. You haven't kept apace with developments, as well as the Nomis academics and ONS who I would expect to be very fussy how the census data is shown there, Wikipedia uses it as a base for their
mapping service, and many well-known online aggregators and services use its geolocation or mapping service -
OpenStreetMap#Popular_services. The bit that actually matters, the county boundary datasets is from the Ordnance Survey and regularly updated -
https://wiki.openstreetmap.org/wiki/Ordnance_Survey_OpenData. Simply put those businesses wouldn't be risking their reputation on a rubbish mapping product.
Keep - Here are the
reasons for deletion. Neither 'outdated' nor 'needing improvement' are on the list. The sources used are not primary, they're secondary and tertiary, and routine calculations (ie adding output areas) are perfectly acceptable under
WP:Synth. --
Ykraps (
talk) 09:45, 22 January 2022 (UTC)reply
The secondary source some of them use is CityPopulation.de which I dont think is a reliable source. I dont think there are any reliable sources for settlements split out by county.
Eopsid (
talk) 14:30, 22 January 2022 (UTC)reply
Eopsid, at risk of repetiton Nomis does show what counties the built up areas and subdivisions are in.
The Equalizer (
talk) 10:20, 23 January 2022 (UTC)reply
Eopsid"Some of them use..." You cannot mass delete articles because some of them use! Request closure of this and reopen deletion requests for the some of them. --
Ykraps (
talk) 09:20, 24 January 2022 (UTC)reply
Sorry let met clarify. There are two sources these articles use. One is the census data on built-up area subdivsions itself (or urban area subdivisions which was they were called prior in the 2001 census data) and the other is CityPopulation.de which is based off the same data.
Eopsid (
talk) 19:30, 24 January 2022 (UTC)reply
You seem to think that the ONS stats are a primary source. They are not. The census forms are the primary source, which are collated, analysed and commented on by the ONS. This makes the ONS a secondary source. There also appears to be some confusion over original research. WP:OR is the synthesis of multiple sources, by Wikipedians. If a further source, not a Wikipedian, does this, it is not OR. If you think the source is unreliable, tag it and state your reasons (ie self-published). There are other sources available so there is no reason to delete all these articles.--
Ykraps (
talk) 07:57, 26 January 2022 (UTC)reply
Ok my mention of primary sources was irrelevant. The articles sources are using something called Built-up area subdivsions (which they've misleadingly referred to as settlements). When there are anomalies in the subdivisions Original Research has been used to mask it such as in
List of settlements in Gloucestershire by population where they are using the population of a parliamentary constituency when the built-up area subdivision fails to provide population for a town. The UK wide version of these articles (
Wikipedia:Articles for deletion/List of largest United Kingdom settlements by population) was deleted for similar reasons 9 years ago. The main original research masking back then was that there was no subdivision for London, so a number of them were combined together. I will open a discussion on the Reliable Sources noticeboard about the potentially unreliable source CityPopulation.de.
Eopsid (
talk) 21:00, 26 January 2022 (UTC)reply
Eopsid, utterly ignoring that ONS do link BUASDs to settlements, for which I provided the ONS methodology notes elsewhere in the discussion. Urban = built up = settled by residents. My response at the base of the page explains the logics of the mix of geographies - so find the relevant data and fix it.
The Equalizer (
talk) 00:52, 27 January 2022 (UTC)reply
Urban = built-up= settlements doesnt imply that subdivisions of said built up areas are settlements and anywhere that isnt a subdivision isnt a settlement.
Eopsid (
talk) 09:01, 27 January 2022 (UTC)reply
I'm not convinced that having (
Wikipedia:Articles for deletion/List of largest United Kingdom settlements by population) deleted was wise, and I certainly don't consider it a valid argument for starting a mass deletion campaign against any article with a similar title. Your concerns regarding built up areas, which I don't endorse, are not relevant to every article and, as I said earlier, there are sources other than www.citypopulation.de available. These are articles that need improving, not deleting.--
Ykraps (
talk) 06:58, 27 January 2022 (UTC)reply
There are no other sources which aren't flawed thats why the UK and England wide ones were deleted
Eopsid (
talk) 11:10, 27 January 2022 (UTC)reply
Yes there are! As has already been explained, the ONS figures are a reliable secondary source and routine calculations are permissible. In addition, local governments tend to produce their own figures. Here are the ones for Dorset, for example [
[1]]. Regarding your concerns about BUAs/SDs, you can request moving the articles to Built up ares in..., or similar, but this really is semantics because a built up area is a settlement. Also, as I keep mentioning, your arguments are not relevant to every single article. If you wish to delete them en masse, you ought to have a reason that is common to all. I am not going to be drawn into an argument about the rights and wrongs of a previous deletion, and I don't think this discussion is going anywhere productive. We all seem to be repeating ourselves now so why don't we let this request run its course? --
Ykraps (
talk) 06:30, 28 January 2022 (UTC)reply
But that Dorset page doesnt use the word settlements at all. Only parishes, towns, wards and unitary authorities, it would be useful for a List of Civil Parishes in Dorset by population or List of Towns in Dorset by population (but it doesnt include Bournemouth so big flaw there) but not List of Settlements in Dorset by population
Eopsid (
talk) 09:00, 28 January 2022 (UTC)reply
Are you proposing we delete these articles because you don't know what a settlement is?! A settlement is a built up area with a population. It saves having to put, list of hamlets, villages, town and cities in.... The term chosen by the ONS is built up areas. As I said above, if you want to propose a move to something else, then that is your perogative but you are just arguing semantics! The figures for Bournemouth are kept by the local government responsible (obviously, I would have thought). [
[2]] I appreciate that you are now heavily invested in this discussion and perhaps don't know how to bow out gracefully but, as I said earlier, this isn't going anywhere productive. Nothing you have said, or are likely to say, is going to make me change my stance!--
Ykraps (
talk) 06:41, 29 January 2022 (UTC)reply
But the link you gave me earlier wasnt about built up areas but civil parishes which can have multiple built up areas or only parts of one. These articles arent using built up areas (except for Cornwall and Bedfordshire which are using them and civil parishes and should probably be split into two articles) but for the most part are using built-up area subdivisions and the consensus from the UK and England wide discussions was that labelling these subdivisions of built-up areas as settlements was misleading.
Eopsid (
talk) 09:13, 29 January 2022 (UTC)reply
Oh dear. No! The link I provided earlier [
[3]] gives population figures for towns, wards, and parishes! More information than is needed! It seems to me, the problem is that you simply don't understand the terminology being used. Is that the case?--
Ykraps (
talk) 16:30, 29 January 2022 (UTC)reply
Keep The reasons given for deletion are absurd, and defy basic
WP:Common Sense. Some of them need a bit of improvement undoubtedly, such as consistent sourcing, and a consistent definition being used. But deleting them would not improve the encyclopedia.
G-13114 (
talk) 10:19, 22 January 2022 (UTC)reply
I don't understand how the reasons are absurd. The England and UK wide versions were deleted for the same reasons.
Eopsid (
talk) 14:27, 22 January 2022 (UTC)reply
The problem with the sourcing is that its a misleading use of it. The source doesnt refer to the areas as settlements but as built-up area subdivisions. I'd support moving all the articles to List of Built-up area subdivisons of X county.
Eopsid (
talk) 14:27, 22 January 2022 (UTC)reply
The source also doesnt split by county so they'd need merging by region to avoid Original Research.
Eopsid (
talk) 14:47, 22 January 2022 (UTC)reply
Eopsid I think there are bigger battles out there to fight. It's known a given built up area consists of one or more settlements, the vast majority of BUAs are within a given county, only a handful straddle or extend across county boundaries. And the counties of where the BUAs are within is visible in Nomis. It's not a problem to explain in the table that this is the case.
The Equalizer (
talk) 15:04, 22 January 2022 (UTC)reply
These aren't BUAs but BUASDs i.e. Built-up area subdivisions. Please dont confuse the two.
Eopsid (
talk) 15:23, 22 January 2022 (UTC)reply
There are quite a few that cross county boundaries, looking at
List of settlements in Gloucestershire by population which contains the largest cross boundary BUASD which is Bristol's. The anomaly has just been covered up with original research. The articles are full of original research where people can't see their town because its not in the source and then add it using a completely different definition to the main one.
Eopsid (
talk) 15:30, 22 January 2022 (UTC)reply
Eopsid, as mentioned by an earlier post extrapolating stats from the OAs to cover that overlapping area is not OR, so fix the problem as much of the list is fine. The articles should really be semi-protected and only changed on consensus as yes I have seen the same issue with editors not realising a certain neighbourhood is already within a BUASD.
The Equalizer (
talk) 10:20, 23 January 2022 (UTC)reply
Delete all due to the problems identified by other people with the sourcing and the fact that the only way to resolve said issues is by relying on OR. Which is clearly a bad way to maintain articles. There's also already consensus that this isn't the way to do things because the England and UK wide versions of similar articles were deleted for the same reasons. It's absurd to ignore prior consensus and rely on OR to justify keeping these instead of just deleting them. --
Adamant1 (
talk) 18:30, 22 January 2022 (UTC)reply
*Delete all Lacks sources establishing that
WP:NLIST is met. "List of Towns in XX" would be better achieved via categories.
MrsSnoozyTurtle 02:04, 23 January 2022 (UTC)reply
The source that was used CityPopulation.de did divide them by county so the original use of that as a reference was sound. The suggested category method does not gather population stats in one place so that they can be sorted by size, which the tables do.
The Equalizer (
talk) 10:20, 23 January 2022 (UTC)reply
Weak keep Thank you,
The Equalizer. I have changed my view of the topic, thanks to your reply and the discussions below.
MrsSnoozyTurtle 21:33, 25 January 2022 (UTC)reply
Weak keep seems like a valid list with a reasonably clear criteria which is that used by the ONS which is essentially independent of the settlements. I have some lists like
User:Crouch, Swale/Suffolk BUAs which shows the status. If anything I'd include all BUAs and BUASDs not just those with a certain population. Crouch, Swale (
talk) 11:07, 23 January 2022 (UTC)reply
Keep all It looks like it's 'throwing baby out with the bath water' time again, folks. The deletion argument is flawed. These are list articles and, taking just
List of settlements in Derbyshire by population as an example: all entries are notable and wikilinked. All entries are referenced. Citations to detail such as population counts would, by their very nature, be expected to be based upon primary sources (i.e. census data). The two columns already allow a user to determine population change over a decade, and once further census data is published, would allow a 2-decade population change to be viewed, assuming they were updated. My only concern is that some of the cited sources are not currently reachable, but
WP:DEADLINKS have never been a rationale for deletion as far as I am aware. In what realm are these pages not informative or relevant to an encyclopaedia? So this is a strong keep from me, and my only criticism is that the population columns are not themselves sortable, which would render them even more useful across different time periods.
Nick Moyes (
talk) 14:27, 24 January 2022 (UTC)reply
Looking at the Derbyshire article's notes columns every settlement is using a different definition, some are parishes, some are groups of wards, some are built-up areas, some are districts. Its not comparing like with like, it's almost a text book example of
Wikipedia:Synthesis. I've made the population columns sortable for you.
Eopsid (
talk) 19:30, 24 January 2022 (UTC)reply
County/district/parish as public entities and wards of those are ultimately legal terms only to describe the governance type of a given area. The individual census base elements that are added up to form the geography and population of those different entities are all based on the same methodology. --
The Equalizer (
talk) 15:57, 25 January 2022 (UTC)reply
The base elements aren't the problem. Its the way they are being combined and compared. It needs to be using a consistent definition, that article's just a mixture of different population statistics.
Eopsid (
talk) 19:16, 25 January 2022 (UTC)reply
But districts can contain parishes and parishes can contain wards and wards can contain parishes so they can't be compared.
Eopsid (
talk) 08:57, 26 January 2022 (UTC)reply
I'd just stick to using BUA/BUASD populations but parish populations could be mentioned as well if needed but this is clearly a settlement list so settlement population should take precedence over parish population and for consistency/to avoid confusion I'd just use the BUA/BUASD as the determining factor for where on the list the population rank is but that's not a problem for AFD. Crouch, Swale (
talk) 22:20, 26 January 2022 (UTC)reply
I disagree
Eopsid. Counties comprise a mix of districts, parishes and wards etc, all layered atop each other.
A big town or city which is a single urban area might therefore be its own district or if important enough, a
Unitary authority. So such a district we take into account.
Everywhere else say, is parished (but in the districts we are ignoring) - and parishes roughly correspond to a single settlement
Then that single urban district and all the parishes in the county will add up to the total county population. And we have ~tada~ a list of biggest to smallest settlements
Now, if an urban area in the county is not its own district or parish like that outlying
Kingswood, South Gloucestershire by Bristol suburb, then how is that counted?
County population then = Own district(s) + parishes + unparished urban area
There are a number of lower census population stats that together can tabulate that space - wards, MSOA, SOAs etc
Councils do sometimes list what wards represent a suburb.
As long as that remaining area hasn't be accounted for in the own district and parishes counts, that remaining area population can be tabulated from lower geographies covering the same space.
And that is why the tables can be a list of "council" areas (when it's not but gives the locals a general understanding of population for their area as many relate to parishes as a key village identifier) - OR - alternatively, a list of BUA/SDs. Both have their pros and cons.
This is why
human geography is very perplexing! There are complications (eg parishes do not always have one settlement, settlements straddles boundaries, BUASDs cross several parishes/districts/counties) but the stats given enough calculation can be worked out.
But we shouldn't be attempting to calculate them, that's up to reliable sources to do, otherwise its just original research.
Eopsid (
talk) 09:01, 27 January 2022 (UTC)reply
Find the comment with the Wiki policy (not guidance!) link further back up the messages that does state routine calculations are fine.
The Equalizer (
talk) 09:56, 27 January 2022 (UTC)reply
These are more than just routine calculations and what you wrote "but the stats given enough calculation can be worked out." implies that.
Eopsid (
talk) 11:10, 27 January 2022 (UTC)reply
You're now guessing. Please provide an example of this supposed complexity, the learned amongst us here could explain where you are going wrong.
The Equalizer (
talk) 17:59, 27 January 2022 (UTC)reply
Keep all. In most cases, population changes will not be huge. Give priority to the update of
List of settlements in Bedfordshire by population as only data from 2001. The others I sampled had data (also) from 2011.
gidonb (
talk) 05:22, 27 January 2022 (UTC)reply
Relisted to generate a more thorough discussion and clearer consensus. Please add new comments below this notice. Thanks, Sandstein 06:56, 29 January 2022 (UTC)reply
Keep all, but they must be improved. The deletion arguments by the nominator seems flawed, but I can see why they'd have considered it in the first place. The articles need some work done for consistency and quality purposes. Contributors from the UK geo portal may be able to help out further. --
Jf81 (
talk) 14:16, 4 February 2022 (UTC)reply
Relisted to generate a more thorough discussion and clearer consensus. Please add new comments below this notice. Thanks,
SpartazHumbug! 22:51, 6 February 2022 (UTC)reply
*Keep all - it seems that the nom did not understand the nature of the sources.
Ingratis (
talk) 04:44, 8 February 2022 (UTC)reply
OK, sorry - I should have taken the time to read through more carefully, although it's difficult to credit that it's not possible for these lists to be made accurate. No !vote.
Ingratis (
talk) 04:03, 9 February 2022 (UTC)reply
"...where they don't match expected settlement boundaries", Whose expectations are we talking about here? Yours? You have made your arguments and they have been discussed (adequately, in my view) there is no need to keep repeating yourself over and over. It could be construed as
badgering or an attempt to
make your arguments seem more plausible through assertion.--
Ykraps (
talk) 07:11, 11 February 2022 (UTC)reply
Obviously not my expectations, the expectations of whoever wrote the article. To give an example from
List of settlements in Gloucestershire by population, it uses built-up area subdivisions but the Bristol one crosses the county boundary into Gloucestershire, so to give a population figure for Kingswood, which is the Gloucestershire part of the Bristol built-up area subdivision, they article uses a parliamentary constituency. A ranked settlement population list using multiple definitions of settlements is clearly
WP:Synthesis. There are examples in the other articles too, I've listed all of them in another discussion about cleaning these articles up.
Eopsid (
talk) 09:57, 11 February 2022 (UTC)reply
I really don’t think your qualified to decide what others expect. If you have issues with particular articles, the correct thing to do is raise them on the corresponding talk page; not nominate multiple articles for deletion, any number of which, may or may not have those issues.--
Ykraps (
talk) 18:23, 11 February 2022 (UTC)reply
The correct thing to do would have been to nominate these articles for deletion just after the England and UK wides ones were deleted instead of waiting years.
Eopsid (
talk) 22:17, 11 February 2022 (UTC)reply
I agree, per
Wikipedia:Deletion is not cleanup if there's a problem with the articles that needs to be fixed outside AFD, as suggested all of the articles should probably use the ONS BUAs and BUASDs not other things at least not for the 1st figure. Crouch, Swale (
talk) 11:28, 12 February 2022 (UTC)reply
Keep (procedural), a ridiculously large multiafd with 36 articles, lets say a perfunctory 5minute search for
references to see whether each/any list meets
WP:NLIST and we're looking at around 3hours, suggest that nominator could have listed say a group of 6 at a time, that would have been more manageable.
Coolabahapple (
talk) 09:09, 12 February 2022 (UTC)reply
I voted delete, but I'd support a procedural keep for the reason you've given. 36 articles in one AfD is rather ridiculous. It's possible a few of them are notable but won't be kept because people are assuming none of them are based on only search for half of them or however many. That aside, it's always better to nominate a few articles of the same type at a time to establish a precedent for deleting the rest. So I think procedurally keeping them and re-nominating them in much small groups would be good. --
Adamant1 (
talk) 06:05, 14 February 2022 (UTC)reply
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's
talk page or in a
deletion review). No further edits should be made to this page.
The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's
talk page or in a
deletion review). No further edits should be made to this page.
There used to be these two now deleted articles
Wikipedia:Articles for deletion/List of localities in England by population and
Wikipedia:Articles for deletion/List of largest United Kingdom settlements by population. There are also articles for each ceremonial county in England on the same subject based on the same data set and they have the same problems. They have no sources outside the primary ones. The source of which doesnt even split them out by county (many of the areas cross county boundaries) so any county splitting is Original Research. They also use the word settlements which the source doesnt mention at all referring to them as built-up subdivisons. These aricles are misleading.
Eopsid (
talk) 21:22, 21 January 2022 (UTC)reply
Delete all. Beyond the sourcing question, these articles have generally not been substantially updated for several years, and the information in them therefore can not possibly be accurate, since population fluctuates year after year. There is no current prospect that these can be kept up to date for the indefinite future.
BD2412T 03:45, 22 January 2022 (UTC)reply
Fix and keep. I created a number of these originally. Those sources at the time did forward to a county specific list of settlements. What's not being realised is that there have been changes on the source website since and so now the data is not on the landing page anymore. But if you drill down into the website, the data can still be found. The banner above states we consider alternatives to deletion, simply locating and updating the references will correct the problem.
On this page it does refer to the table as a list of Settlements, and there is a status column explaining whether they are
BUAs or sub divisions.
As to
BD2412's concern about outdated stats, the census is done every ten years which are official counts, and anything outside of those is an estimate. 2021 was the latest census and so up-to-date figures will be imminent.
The fact that interim data would be "an estimate" doesn't fix the speed with which periodic census data becomes outdated. If the updated title reflected "as of 2021" this would be more accurate, but I would still question the encyclopedic value of such a snapshot in time.
BD2412T 05:49, 22 January 2022 (UTC)reply
Eopsid, CityPopulation.de was used at the time because Nomis didn't have direct links to census area data in those days and the Neighbourhood Statistics census site which did was decommissioned. That London link seems to have been superseded by the agglomeration list
https://www.citypopulation.de/en/uk/agglo/E34004707A__london so again I think site pages have been 'improved'. The data is now better accessible through a primary source, and
Template:NOMIS2001 /
Template:NOMIS2011 have been developed to take advantage of that, so it would be a matter of updating the refs.
BD2412, usually the article refers to the shortcoming of the census data in the prose of the articles, and estimates are really only given for main settlements (because the data is only provided to district level, but not towns/parishes/settlements as it's never been that granular. There is talk of using other gov data sources to create future censuses because of the expense
https://www.bbc.co.uk/news/uk-51468919, that methodology is more in line with how estimates are created, but until that is accepted practice I personally think that decade long difference gives the best insight into the growth or not of a settlement and smoothes out any irregularities. But the articles are definitely notable as there will always be interest in what are the largest settlements.
The NOMIS data doesnt split it out by county. So splitting it by county would be Original Research. The NOMIS data also doesnt refer to them as settlements but as built-up area subdivisons.
Eopsid (
talk) 14:27, 22 January 2022 (UTC)reply
That link doesnt mention Built-up area subdivisions at all. I will have a deeper look into the NOMIS data to see if it does show counties.
Eopsid (
talk) 19:35, 24 January 2022 (UTC)reply
The maps are all based on Open Street Maps which isnt a reliable source, its basically another Wiki.
Eopsid (
talk) 19:16, 25 January 2022 (UTC)reply
Yes Wiki too then - nominally unreliable as it too can be changed easily. Yet we are here fighting to make sure it's relevant and the best it can be. So too at OSM I bet. You haven't kept apace with developments, as well as the Nomis academics and ONS who I would expect to be very fussy how the census data is shown there, Wikipedia uses it as a base for their
mapping service, and many well-known online aggregators and services use its geolocation or mapping service -
OpenStreetMap#Popular_services. The bit that actually matters, the county boundary datasets is from the Ordnance Survey and regularly updated -
https://wiki.openstreetmap.org/wiki/Ordnance_Survey_OpenData. Simply put those businesses wouldn't be risking their reputation on a rubbish mapping product.
Keep - Here are the
reasons for deletion. Neither 'outdated' nor 'needing improvement' are on the list. The sources used are not primary, they're secondary and tertiary, and routine calculations (ie adding output areas) are perfectly acceptable under
WP:Synth. --
Ykraps (
talk) 09:45, 22 January 2022 (UTC)reply
The secondary source some of them use is CityPopulation.de which I dont think is a reliable source. I dont think there are any reliable sources for settlements split out by county.
Eopsid (
talk) 14:30, 22 January 2022 (UTC)reply
Eopsid, at risk of repetiton Nomis does show what counties the built up areas and subdivisions are in.
The Equalizer (
talk) 10:20, 23 January 2022 (UTC)reply
Eopsid"Some of them use..." You cannot mass delete articles because some of them use! Request closure of this and reopen deletion requests for the some of them. --
Ykraps (
talk) 09:20, 24 January 2022 (UTC)reply
Sorry let met clarify. There are two sources these articles use. One is the census data on built-up area subdivsions itself (or urban area subdivisions which was they were called prior in the 2001 census data) and the other is CityPopulation.de which is based off the same data.
Eopsid (
talk) 19:30, 24 January 2022 (UTC)reply
You seem to think that the ONS stats are a primary source. They are not. The census forms are the primary source, which are collated, analysed and commented on by the ONS. This makes the ONS a secondary source. There also appears to be some confusion over original research. WP:OR is the synthesis of multiple sources, by Wikipedians. If a further source, not a Wikipedian, does this, it is not OR. If you think the source is unreliable, tag it and state your reasons (ie self-published). There are other sources available so there is no reason to delete all these articles.--
Ykraps (
talk) 07:57, 26 January 2022 (UTC)reply
Ok my mention of primary sources was irrelevant. The articles sources are using something called Built-up area subdivsions (which they've misleadingly referred to as settlements). When there are anomalies in the subdivisions Original Research has been used to mask it such as in
List of settlements in Gloucestershire by population where they are using the population of a parliamentary constituency when the built-up area subdivision fails to provide population for a town. The UK wide version of these articles (
Wikipedia:Articles for deletion/List of largest United Kingdom settlements by population) was deleted for similar reasons 9 years ago. The main original research masking back then was that there was no subdivision for London, so a number of them were combined together. I will open a discussion on the Reliable Sources noticeboard about the potentially unreliable source CityPopulation.de.
Eopsid (
talk) 21:00, 26 January 2022 (UTC)reply
Eopsid, utterly ignoring that ONS do link BUASDs to settlements, for which I provided the ONS methodology notes elsewhere in the discussion. Urban = built up = settled by residents. My response at the base of the page explains the logics of the mix of geographies - so find the relevant data and fix it.
The Equalizer (
talk) 00:52, 27 January 2022 (UTC)reply
Urban = built-up= settlements doesnt imply that subdivisions of said built up areas are settlements and anywhere that isnt a subdivision isnt a settlement.
Eopsid (
talk) 09:01, 27 January 2022 (UTC)reply
I'm not convinced that having (
Wikipedia:Articles for deletion/List of largest United Kingdom settlements by population) deleted was wise, and I certainly don't consider it a valid argument for starting a mass deletion campaign against any article with a similar title. Your concerns regarding built up areas, which I don't endorse, are not relevant to every article and, as I said earlier, there are sources other than www.citypopulation.de available. These are articles that need improving, not deleting.--
Ykraps (
talk) 06:58, 27 January 2022 (UTC)reply
There are no other sources which aren't flawed thats why the UK and England wide ones were deleted
Eopsid (
talk) 11:10, 27 January 2022 (UTC)reply
Yes there are! As has already been explained, the ONS figures are a reliable secondary source and routine calculations are permissible. In addition, local governments tend to produce their own figures. Here are the ones for Dorset, for example [
[1]]. Regarding your concerns about BUAs/SDs, you can request moving the articles to Built up ares in..., or similar, but this really is semantics because a built up area is a settlement. Also, as I keep mentioning, your arguments are not relevant to every single article. If you wish to delete them en masse, you ought to have a reason that is common to all. I am not going to be drawn into an argument about the rights and wrongs of a previous deletion, and I don't think this discussion is going anywhere productive. We all seem to be repeating ourselves now so why don't we let this request run its course? --
Ykraps (
talk) 06:30, 28 January 2022 (UTC)reply
But that Dorset page doesnt use the word settlements at all. Only parishes, towns, wards and unitary authorities, it would be useful for a List of Civil Parishes in Dorset by population or List of Towns in Dorset by population (but it doesnt include Bournemouth so big flaw there) but not List of Settlements in Dorset by population
Eopsid (
talk) 09:00, 28 January 2022 (UTC)reply
Are you proposing we delete these articles because you don't know what a settlement is?! A settlement is a built up area with a population. It saves having to put, list of hamlets, villages, town and cities in.... The term chosen by the ONS is built up areas. As I said above, if you want to propose a move to something else, then that is your perogative but you are just arguing semantics! The figures for Bournemouth are kept by the local government responsible (obviously, I would have thought). [
[2]] I appreciate that you are now heavily invested in this discussion and perhaps don't know how to bow out gracefully but, as I said earlier, this isn't going anywhere productive. Nothing you have said, or are likely to say, is going to make me change my stance!--
Ykraps (
talk) 06:41, 29 January 2022 (UTC)reply
But the link you gave me earlier wasnt about built up areas but civil parishes which can have multiple built up areas or only parts of one. These articles arent using built up areas (except for Cornwall and Bedfordshire which are using them and civil parishes and should probably be split into two articles) but for the most part are using built-up area subdivisions and the consensus from the UK and England wide discussions was that labelling these subdivisions of built-up areas as settlements was misleading.
Eopsid (
talk) 09:13, 29 January 2022 (UTC)reply
Oh dear. No! The link I provided earlier [
[3]] gives population figures for towns, wards, and parishes! More information than is needed! It seems to me, the problem is that you simply don't understand the terminology being used. Is that the case?--
Ykraps (
talk) 16:30, 29 January 2022 (UTC)reply
Keep The reasons given for deletion are absurd, and defy basic
WP:Common Sense. Some of them need a bit of improvement undoubtedly, such as consistent sourcing, and a consistent definition being used. But deleting them would not improve the encyclopedia.
G-13114 (
talk) 10:19, 22 January 2022 (UTC)reply
I don't understand how the reasons are absurd. The England and UK wide versions were deleted for the same reasons.
Eopsid (
talk) 14:27, 22 January 2022 (UTC)reply
The problem with the sourcing is that its a misleading use of it. The source doesnt refer to the areas as settlements but as built-up area subdivisions. I'd support moving all the articles to List of Built-up area subdivisons of X county.
Eopsid (
talk) 14:27, 22 January 2022 (UTC)reply
The source also doesnt split by county so they'd need merging by region to avoid Original Research.
Eopsid (
talk) 14:47, 22 January 2022 (UTC)reply
Eopsid I think there are bigger battles out there to fight. It's known a given built up area consists of one or more settlements, the vast majority of BUAs are within a given county, only a handful straddle or extend across county boundaries. And the counties of where the BUAs are within is visible in Nomis. It's not a problem to explain in the table that this is the case.
The Equalizer (
talk) 15:04, 22 January 2022 (UTC)reply
These aren't BUAs but BUASDs i.e. Built-up area subdivisions. Please dont confuse the two.
Eopsid (
talk) 15:23, 22 January 2022 (UTC)reply
There are quite a few that cross county boundaries, looking at
List of settlements in Gloucestershire by population which contains the largest cross boundary BUASD which is Bristol's. The anomaly has just been covered up with original research. The articles are full of original research where people can't see their town because its not in the source and then add it using a completely different definition to the main one.
Eopsid (
talk) 15:30, 22 January 2022 (UTC)reply
Eopsid, as mentioned by an earlier post extrapolating stats from the OAs to cover that overlapping area is not OR, so fix the problem as much of the list is fine. The articles should really be semi-protected and only changed on consensus as yes I have seen the same issue with editors not realising a certain neighbourhood is already within a BUASD.
The Equalizer (
talk) 10:20, 23 January 2022 (UTC)reply
Delete all due to the problems identified by other people with the sourcing and the fact that the only way to resolve said issues is by relying on OR. Which is clearly a bad way to maintain articles. There's also already consensus that this isn't the way to do things because the England and UK wide versions of similar articles were deleted for the same reasons. It's absurd to ignore prior consensus and rely on OR to justify keeping these instead of just deleting them. --
Adamant1 (
talk) 18:30, 22 January 2022 (UTC)reply
*Delete all Lacks sources establishing that
WP:NLIST is met. "List of Towns in XX" would be better achieved via categories.
MrsSnoozyTurtle 02:04, 23 January 2022 (UTC)reply
The source that was used CityPopulation.de did divide them by county so the original use of that as a reference was sound. The suggested category method does not gather population stats in one place so that they can be sorted by size, which the tables do.
The Equalizer (
talk) 10:20, 23 January 2022 (UTC)reply
Weak keep Thank you,
The Equalizer. I have changed my view of the topic, thanks to your reply and the discussions below.
MrsSnoozyTurtle 21:33, 25 January 2022 (UTC)reply
Weak keep seems like a valid list with a reasonably clear criteria which is that used by the ONS which is essentially independent of the settlements. I have some lists like
User:Crouch, Swale/Suffolk BUAs which shows the status. If anything I'd include all BUAs and BUASDs not just those with a certain population. Crouch, Swale (
talk) 11:07, 23 January 2022 (UTC)reply
Keep all It looks like it's 'throwing baby out with the bath water' time again, folks. The deletion argument is flawed. These are list articles and, taking just
List of settlements in Derbyshire by population as an example: all entries are notable and wikilinked. All entries are referenced. Citations to detail such as population counts would, by their very nature, be expected to be based upon primary sources (i.e. census data). The two columns already allow a user to determine population change over a decade, and once further census data is published, would allow a 2-decade population change to be viewed, assuming they were updated. My only concern is that some of the cited sources are not currently reachable, but
WP:DEADLINKS have never been a rationale for deletion as far as I am aware. In what realm are these pages not informative or relevant to an encyclopaedia? So this is a strong keep from me, and my only criticism is that the population columns are not themselves sortable, which would render them even more useful across different time periods.
Nick Moyes (
talk) 14:27, 24 January 2022 (UTC)reply
Looking at the Derbyshire article's notes columns every settlement is using a different definition, some are parishes, some are groups of wards, some are built-up areas, some are districts. Its not comparing like with like, it's almost a text book example of
Wikipedia:Synthesis. I've made the population columns sortable for you.
Eopsid (
talk) 19:30, 24 January 2022 (UTC)reply
County/district/parish as public entities and wards of those are ultimately legal terms only to describe the governance type of a given area. The individual census base elements that are added up to form the geography and population of those different entities are all based on the same methodology. --
The Equalizer (
talk) 15:57, 25 January 2022 (UTC)reply
The base elements aren't the problem. Its the way they are being combined and compared. It needs to be using a consistent definition, that article's just a mixture of different population statistics.
Eopsid (
talk) 19:16, 25 January 2022 (UTC)reply
But districts can contain parishes and parishes can contain wards and wards can contain parishes so they can't be compared.
Eopsid (
talk) 08:57, 26 January 2022 (UTC)reply
I'd just stick to using BUA/BUASD populations but parish populations could be mentioned as well if needed but this is clearly a settlement list so settlement population should take precedence over parish population and for consistency/to avoid confusion I'd just use the BUA/BUASD as the determining factor for where on the list the population rank is but that's not a problem for AFD. Crouch, Swale (
talk) 22:20, 26 January 2022 (UTC)reply
I disagree
Eopsid. Counties comprise a mix of districts, parishes and wards etc, all layered atop each other.
A big town or city which is a single urban area might therefore be its own district or if important enough, a
Unitary authority. So such a district we take into account.
Everywhere else say, is parished (but in the districts we are ignoring) - and parishes roughly correspond to a single settlement
Then that single urban district and all the parishes in the county will add up to the total county population. And we have ~tada~ a list of biggest to smallest settlements
Now, if an urban area in the county is not its own district or parish like that outlying
Kingswood, South Gloucestershire by Bristol suburb, then how is that counted?
County population then = Own district(s) + parishes + unparished urban area
There are a number of lower census population stats that together can tabulate that space - wards, MSOA, SOAs etc
Councils do sometimes list what wards represent a suburb.
As long as that remaining area hasn't be accounted for in the own district and parishes counts, that remaining area population can be tabulated from lower geographies covering the same space.
And that is why the tables can be a list of "council" areas (when it's not but gives the locals a general understanding of population for their area as many relate to parishes as a key village identifier) - OR - alternatively, a list of BUA/SDs. Both have their pros and cons.
This is why
human geography is very perplexing! There are complications (eg parishes do not always have one settlement, settlements straddles boundaries, BUASDs cross several parishes/districts/counties) but the stats given enough calculation can be worked out.
But we shouldn't be attempting to calculate them, that's up to reliable sources to do, otherwise its just original research.
Eopsid (
talk) 09:01, 27 January 2022 (UTC)reply
Find the comment with the Wiki policy (not guidance!) link further back up the messages that does state routine calculations are fine.
The Equalizer (
talk) 09:56, 27 January 2022 (UTC)reply
These are more than just routine calculations and what you wrote "but the stats given enough calculation can be worked out." implies that.
Eopsid (
talk) 11:10, 27 January 2022 (UTC)reply
You're now guessing. Please provide an example of this supposed complexity, the learned amongst us here could explain where you are going wrong.
The Equalizer (
talk) 17:59, 27 January 2022 (UTC)reply
Keep all. In most cases, population changes will not be huge. Give priority to the update of
List of settlements in Bedfordshire by population as only data from 2001. The others I sampled had data (also) from 2011.
gidonb (
talk) 05:22, 27 January 2022 (UTC)reply
Relisted to generate a more thorough discussion and clearer consensus. Please add new comments below this notice. Thanks, Sandstein 06:56, 29 January 2022 (UTC)reply
Keep all, but they must be improved. The deletion arguments by the nominator seems flawed, but I can see why they'd have considered it in the first place. The articles need some work done for consistency and quality purposes. Contributors from the UK geo portal may be able to help out further. --
Jf81 (
talk) 14:16, 4 February 2022 (UTC)reply
Relisted to generate a more thorough discussion and clearer consensus. Please add new comments below this notice. Thanks,
SpartazHumbug! 22:51, 6 February 2022 (UTC)reply
*Keep all - it seems that the nom did not understand the nature of the sources.
Ingratis (
talk) 04:44, 8 February 2022 (UTC)reply
OK, sorry - I should have taken the time to read through more carefully, although it's difficult to credit that it's not possible for these lists to be made accurate. No !vote.
Ingratis (
talk) 04:03, 9 February 2022 (UTC)reply
"...where they don't match expected settlement boundaries", Whose expectations are we talking about here? Yours? You have made your arguments and they have been discussed (adequately, in my view) there is no need to keep repeating yourself over and over. It could be construed as
badgering or an attempt to
make your arguments seem more plausible through assertion.--
Ykraps (
talk) 07:11, 11 February 2022 (UTC)reply
Obviously not my expectations, the expectations of whoever wrote the article. To give an example from
List of settlements in Gloucestershire by population, it uses built-up area subdivisions but the Bristol one crosses the county boundary into Gloucestershire, so to give a population figure for Kingswood, which is the Gloucestershire part of the Bristol built-up area subdivision, they article uses a parliamentary constituency. A ranked settlement population list using multiple definitions of settlements is clearly
WP:Synthesis. There are examples in the other articles too, I've listed all of them in another discussion about cleaning these articles up.
Eopsid (
talk) 09:57, 11 February 2022 (UTC)reply
I really don’t think your qualified to decide what others expect. If you have issues with particular articles, the correct thing to do is raise them on the corresponding talk page; not nominate multiple articles for deletion, any number of which, may or may not have those issues.--
Ykraps (
talk) 18:23, 11 February 2022 (UTC)reply
The correct thing to do would have been to nominate these articles for deletion just after the England and UK wides ones were deleted instead of waiting years.
Eopsid (
talk) 22:17, 11 February 2022 (UTC)reply
I agree, per
Wikipedia:Deletion is not cleanup if there's a problem with the articles that needs to be fixed outside AFD, as suggested all of the articles should probably use the ONS BUAs and BUASDs not other things at least not for the 1st figure. Crouch, Swale (
talk) 11:28, 12 February 2022 (UTC)reply
Keep (procedural), a ridiculously large multiafd with 36 articles, lets say a perfunctory 5minute search for
references to see whether each/any list meets
WP:NLIST and we're looking at around 3hours, suggest that nominator could have listed say a group of 6 at a time, that would have been more manageable.
Coolabahapple (
talk) 09:09, 12 February 2022 (UTC)reply
I voted delete, but I'd support a procedural keep for the reason you've given. 36 articles in one AfD is rather ridiculous. It's possible a few of them are notable but won't be kept because people are assuming none of them are based on only search for half of them or however many. That aside, it's always better to nominate a few articles of the same type at a time to establish a precedent for deleting the rest. So I think procedurally keeping them and re-nominating them in much small groups would be good. --
Adamant1 (
talk) 06:05, 14 February 2022 (UTC)reply
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's
talk page or in a
deletion review). No further edits should be made to this page.