Talk:List of fatal dog attacks in the United Kingdom

Let's try to collect and list all the fatalities
Someone was generous yesterday and dumped a bunch of raw URLs for citations onto the article (25 new links). Someone else dumped even more this morning (9 links). I sifted through each one of them and put them into the list (table). We have now gone from 3 fatalities to 35 fatalities. My estimate is that we have about HALF of them (more or less). If anyone is interested in doing some research to uncover more reports of fatalities, please do so. Even if you don't insert them into the table, it's still helpful. Nomopbs (talk) 17:48, 25 June 2019 (UTC)


 * Nomopbs I checked tbe majority of the mainstream newspapers and news sites in the UK (see below) for fatal dog attacks I am struggling to find stuff from the 1980s to 1990s. Maybe you might have better luck. Dwanyewest (talk) 13:28, 1 July 2019 (UTC)
 * https://www.dailymail.co.uk/home/index.html
 * https://www.express.co.uk/search?s=
 * https://www.mirror.co.uk/all-about/dangerous-dogs
 * https://www.theguardian.com/uk/dangerous-dogs
 * https://www.thesun.co.uk/?s=
 * https://www.thetimes.co.uk/
 * https://www.telegraph.co.uk/
 * https://www.bbc.co.uk/news
 * https://www.independent.co.uk/
 * https://www.heraldscotland.com/


 * Thanks for looking. A lot of the online news articles are being taken down (expired) after about two years. If we had any old links to news articles, we might be able to find them in the Wayback Machine. In lieu of an article for a particular death, any article, book or study even "mentioning" individual deaths would even be good (giving a name and basic date). — Nomopbs (talk) 16:55, 1 July 2019 (UTC)

Change name of page to "Fatal dog attacks in England,United Kingdom.
Does anyone know how we could change the name of the page to "Fatal dog attacks in the United Kingdom"? I know we could create a re-direct from "Fatal dog attacks in the United Kingdom" to "Fatal dog attacks in England and Wales", but it would be better if the page name was for "UK" and not the limited "England and Wales", and the re-direct went the other way.

Historically, the page used to be a simple list of death counts from a single publication from the Office for National Statistics, which is preserved in the current article under the heading "Number of fatalities each year (England and Wales 1981-2015)". 

In April 2019, I co-opted the England & Wales page and started to add fatalites into a table in a similar manner as the articles Fatal dog attacks in the United States and Fatal dog attacks in Canada. Recently, I created a master article Fatal dog attacks which lists other fatalities globally that don't have another place to be posted, and which lists/links all the other country pages in one place.


 * —Nomopbs (talk) 18:21, 25 June 2019 (UTC)
 * You can request a move at WP:MOVE. I agree England and Wales seems a bit arbitrary and excludes, among other places, Scotland and North Ireland, which I would think would be significant population omissions. PearlSt82 (talk) 23:55, 25 June 2019 (UTC)
 * Thanks for the tip. I just used WP:MOVE to change the name of the article. Nomopbs (talk) 02:41, 26 June 2019 (UTC)

A note on scientific findings of breed is needed
With bias forming against breeds and a focus not placed towards owners, this article needs some balancing science, and perhaps a link to overall dog bite statistics - assuming the goal is to provide rounded information without agenda and inform. 31.54.32.209 (talk) 02:39, 28 September 2023 (UTC)


 * It would seem that this needs to be adjusted with the current introductory paragraph which is significantly inflammatory. The sourcing regarding the statement of "american bully XLs" being the major statistic, but not sourcing reference to this excepting newspaper articles is obtuse. The subsequence reference to UK GOV website is nonsensical and should be rectified.
 * Sourcing should be reliable, including Police reported data, rather than news media where there is significant bias when reporting, and untrained individuals making assessments on type. 217.155.201.0 (talk) 16:34, 16 January 2024 (UTC)

The Belfast Case
Should the incident in Belfast where a pregnant woman was attacked by an XL Bully and forced into early labour, resulting in the death of the baby, be added to this list? Hashbrown3839 (talk) 19:32, 12 December 2023 (UTC)
 * Please link the case here, but unfortunately I do not think it counts as a fatal dog attack as the dog did not attack the baby. Feudonym (talk) 01:35, 20 March 2024 (UTC)
 * I think I would count it. The indirect causation doesn't seem disqualifying to me; if a dog attacked Bob and caused him to stumble and bump into Charles, knocking Charles off a cliff to his death, I would still call that a "fatal dog attack", personally. ExplodingCabbage (talk) 10:03, 16 June 2024 (UTC)

Remove DEFRA summary count
Propose to remove the DEFRA summary counts - this is a very short list with only 13 years listed ending in 2017. It also conflicts with the more comprehensive ONS statistics, with much lower figures, and hence potentially downplays the number of fatalities. It does however closely match the numbers from the article, which is noted as incomplete, and is there fore duplicated anyway. Feudonym (talk) 01:40, 20 March 2024 (UTC)


 * I think this is a good idea, as the information value of the DEFRA summary is low. Wikigrund (talk) 09:14, 21 May 2024 (UTC)

AfD:List of fatal dog attacks in the United States
has been nominated for deletion, project members are invited to comment on the the AfD discussion page. ~ Veritas Aeterna (talk) 21:09, 29 March 2024 (UTC)

Restoring a couple of Daily Mail links...
Revision https://en.wikipedia.org/w/index.php?title=List_of_fatal_dog_attacks_in_the_United_Kingdom&diff=prev&oldid=1225852588 removes a couple of citations of the Daily Mail that I added. I'm about to revert it, but think it's worth outlining my reasoning here so we can discuss.

In both cases, there ARE other media outlets - ones that do not appear in Deprecated sources - that reported the same facts as the Daily Mail article I'm citing. The claims about Archie-Lee Hirst, for which I cited https://www.dailymail.co.uk/news/article-1048192/Rottweiler-mauled-baby-death-walked-months.html, were previously sourced from an Evening Standard article, https://www.standard.co.uk/hp/front/rottweiler-which-mauled-baby-to-death-had-not-been-walked-for-five-months-6883866.html. Lesley Banks, on the other hand, was also "written" about in the Metro (https://metro.co.uk/2012/08/10/woman-dies-from-rottweiler-bite-after-she-refused-to-report-her-beloved-pet-530924/) and the Mirror (https://www.mirror.co.uk/news/uk-news/landlady-died-from-dog-bite-fearing-1246224).

One might thus think that, to comply with the Deprecated Sources policy, we should cite those other newspapers instead. However, I am extremely reluctant to do so, because all of those articles appear to have simply been plagiarised from the Mail. The Standard article about Hirst we were previously citing is literally a verbatim copy and paste of Mail article published years earlier. The Metro and Mirror articles about Banks, meanwhile, are close paraphrases of the earlier Mail article, with paragraphs lifted and then a few words tweaked.

This kind of "churnalism" was a common feature of the British press from at least the 90s until fairly recently; for most of my life, most of our newspapers had no compunctions at all about blatantly copying stories from competitor papers, either verbatim or with a few words changed. It leaves me rather unsure, in general, how to deal with the Mail being on the Deprecated Sources list, because it is frequently the Mail's original reporting that I would see being plagiarised by the rest of the press, as is the case here. But what I think surely can't be the correct thing to do is to cite the plagiarists as a source for claims that originate in the Daily Mail. If we do that, we're still relying on the Mail as our source of information, but laundering the citation through a plagiarist at another newspaper who has almost certainly done no additional investigation themselves; that surely cannot be better than directly citing the Mail!

I note furthermore that the 2017 RFC about the Mail specifically notes that their older reporting might be reliable, and that discussion about deprecated sources in that RFC and elsewhere notes that the deprecations are not absolute prohibitions and encourages using common sense. Since the articles here are 5 years or more before 2017, and the messy situation above exists where non-deprecated sources exist but it seems inappropriate to cite them, it seems to me that this is a situation where citing the Mail is justified.

If we don't do that, though, then surely we should actually remove the factual claims that we're relying on Daily Mail reporting for, rather than keeping the claims but removing the citation! The purpose of deprecating sources is to avoid false or unreliable information getting added to Wikipedia, not to weaken the status or page-rank of the Daily Mail; removing citations of the Mail without also removing any facts for which the Mail is our only source seems wholly wrong to me; if we think the factual claims we've taken from the Mail are untrustworthy, then removing the Mail citation but leaving the claims is just hiding the problem, not fixing it, and makes it more difficult for future editors to detect and address the problem.

@Nikkimaria, I'd welcome your commentary here on all the above! ExplodingCabbage (talk) 16:55, 15 June 2024 (UTC)


 * My understanding of the discussion is that "older" was in reference to much older content, as in pre-internet, rather than just preceding the RfC. Given that and your findings, I'd suggest removing the content entirely. Nikkimaria (talk) 17:15, 15 June 2024 (UTC)
 * Noted, but my personal judgement is still that we should keep them. I don't really see much reason to doubt the specific facts we're citing the Mail for in either case.
 * In the case of Lee-Hirst, the one fact we get from the Mail article that isn't in the other sources we cite is that the dog that killed her had not been walked for months. If the Mail were citing an unnamed neighbour as the source, I might worry about whether they'd made that claim up. But instead they're citing the testimony of a named expert witness at the baby's inquest. It seems very unlikely to me that the Mail would've manufactured that testimony out of whole cloth!
 * In the case of Lesley Banks, the claim we're getting from the Mail (and the plagiarised articles in other papers) is that her decision not to seek medical treatment was motivated by fear that her dog, which had previously saved her life, would be put down. It appears to be implicit in the article that this claim once again originates from an inquest. Most of this, we can corroborate from other sources. The fact of her not seeking treatment is corroborated by the RSCPA source we're citing, and that the dog had previously saved her life is corroborated by local news (https://www.shropshirestar.com/news/2010/01/14/pet-rottweiler-saved-owners-life/). The one thing that is not explicitly stated elsewhere is that her motivation for not seeking treatment was fear of the consequences for her dog. But given that we can confirm the Mail got most of the details right, it seems overwhelmingly more likely to me that they accurately reported what was said at the inquest, rather than that they made something up that miraculously is correct in every detail we're able to corroborate against other sources but is wrong about the details we can't.
 * I would not be massively upset about being overruled, though. I'll wait and see what others think, if anyone else decides to weigh in! ExplodingCabbage (talk) 17:40, 15 June 2024 (UTC)
 * I would suggest to always take the most reliable source. Even when it means that sometimes we can't add some information.
 * Please do not add questionable sources (especially if other sources are available), some people only look for excuses to delete the whole page. Wikigrund (talk) 11:23, 22 June 2024 (UTC)
 * Looks like purging the information that originates in the Daily Mail is the emerging consensus view of those who have actually read this discussion and voiced an opinion. I'm happy with that.
 * This may yet prove controversial. There is perhaps a competing view - expressed implicitly by the edits of @David_Gerard and @FlightTime, and by a scolding @FlightTime left on my Talk page for reverting them - that the correct thing to do is to include the information that the Daily Mail reported, but not cite the source we got it from. I don't see how that can possibly be the right thing to do, though, and tentatively assume that neither user has really looked carefully at the article and that this implied viewpoint surely cannot be the considered opinion of either of them.
 * I'll therefore go ahead and make the edit, albeit at risk of further conflict from the apparent "trust the Daily Mail's reporting, but don't link to them" faction. ExplodingCabbage (talk) 11:58, 22 June 2024 (UTC)
 * Actually, after re-scrutinising the Mirror article about Lesley Banks, I see that - although some sentences and paragraphs are basically identical to ones from the Daily Mail - there are quotes from the coroner in the Mirror article that are not included in the Mail one. This makes me a little unsure about my "plagiarism" diagnosis in the case of Banks; perhaps instead they simply copied and pasted some stuff from the same press release, or else plagiarised a bit from the Mail while also including some of their own reporting. It's certainly not as clear-cut as the Lee-Hirst case where the Standard simply copied and pasted a Daily Mail article verbatim and republished it.
 * I'm going to therefore make the following edits:
 * change the citation for the Lesley Banks case to the Mirror, instead of the Mail, and keep all the information currently included there
 * remove the Mail citation from the Lee-Hirst case along with the claim that the dog's owners never walked it
 * ExplodingCabbage (talk) 12:18, 22 June 2024 (UTC)
 * Boom - I found a non-deprecated source reporting on the Lee-Hirst inquest, including the key claim the Daily Mail made about the dog not being walked for months! http://news.bbc.co.uk/1/hi/england/bradford/7576074.stm
 * It seems we can actually have the best of all worlds - cite non-plagiarised, non-deprecated sources, and keep basically all the claims we were previously getting from the Daily Mail. We'll still need to tweak the article a little bit - the point about the lack of exercise leading to "agitation" of the dog is not explicitly included in the BBC article - but we get to keep almost everything this way. ExplodingCabbage (talk) 12:29, 22 June 2024 (UTC)
 * All done. I think this surely gets us the article a state that everyone can live with. ExplodingCabbage (talk) 12:40, 22 June 2024 (UTC)

How to deal with cases where it's ambiguous if they were really "fatal attacks"?
There are two recurring kinds of story that fit into this category:


 * 1) A dog gives their owner a small nip, perhaps accidentally. The wound gets infected and the owner dies from the infection. Examples: Gary Dickinson (not yet included, but listed at https://committees.parliament.uk/writtenevidence/92337/html/), Lesley Banks, Damian Holden.
 * 2) A person collapses due to some other cause, then their pet dog interacts with them in such a manner that it is ambiguous and up to interpretation whether it was "attacking" them, merely unsentimentally snacking on their body in the belief that they are already dead, or actively trying to help them. Examples: Kirsty Ross and George Dinham (both of whose dogs attempted to help after their owner had an epileptic seizure, and ended up killing them instead), James Rehill (Rottweiler ripped chunks of flesh from the owner's face after the owner fell unconscious in the street from a stroke; this behaviour was described by some witnesses as the dog attempting to help the owner, and the coroner ruled the death was due to "natural causes").

Should these appear in the article, or should we leave them out on the grounds that they're arguably not "attacks" at all? ExplodingCabbage (talk) 17:18, 15 June 2024 (UTC)


 * Dogs don't have reasoning thought and wikipedia should not be reproducing media tidbits where someone uttered that a dog was trying to help their owner. It is implausible, and stems from a revulsion of the incident and an attempt to explain unnatural or deviant behaviors. Those people would not have died had a dog not bitten them or snacked on their unconscious bodies. It is not normal for dogs to eat their unconcious or dead human masters except perhaps in a survival situation such as being long confined in a house with their dead and decomposing owner and the dog is wasting away from starvation. Even then, the average domesticated dog won't eat a human body. Not having been "fed for roughly 48 hours" is not starvation, such as is unnecessarily mentioned in the Clifford Clarke death.


 * Any bite could be categorized as an attack. Surely one doesn't need to have a full-on violent mauling to qualify for inclusion here. I think you are quibbling over the word "attack". IMO, you could omit those deaths where someone died "because of" a dog, but which didn't involve biting, scratching or other touch behavior, such as being chased, tripping and bashing one's head on a rock. Going into premature labor (as mentioned above) doesn't count as a fatal dog attack, though snacking on a pregnant belly probably would. The cliff scenario, possibly.
 * Wikianon3770617 (talk) 18:06, 16 June 2024 (UTC)
 * I think that every case where a dog has led to the death of a human should be added to the list.
 * This includes cases of transmitted diseases (such as rabies), sepsis and also injuries from falls due to a dog's behavior. The official cause of death category in statistics is often "Bitten or struck by a dog". There is space in the "Circumstances" column to explain what happened and to add the info that it was most likely an unfortunate accident.
 * I do not think we should exclude cases where "the dog did not want to harm or kill" the person. Wikigrund (talk) 11:08, 22 June 2024 (UTC)
 * I think I basically agree with you. Pedantically, "every case where a dog has led to the death of a human" is too broad - this would include e.g. cases where someone drowns trying to rescue their pet dog, or where someone dies due to a dog allergy, or trips over a dog and falls off a cliff - but I assume you don't mean to suggest including those. I think (and expect you'll agree) that we should include every case where:
 * there is a plausible possibility, reported on by some reliable source, that a dog took some action that caused harm to the victim, even if it's controversial whether it really happened...
 * ... and that action could reasonably be characterised as an "attack", even if that characterisation is debatable...
 * ... and it's plausible (and has been suggested, either explicitly or implicitly, by a reliable source) that this caused, or contributed to, the victim's death, even if this is also controversial
 * Then we can explain the details in the "Circumstances" column, as you say. Inevitably this approach will lead to some dubious inclusions, but the reader can judge them for themselves, and at least this way gives an objective standard and doesn't require us to reach a semantic/philosophical editorial position on precisely what constitutes an "attack", nor to have philosophical arguments about causation, nor to try to divine what will often be fundamentally unknowable information about whether a dog bite contributed to an already-sick person's death or whether the death was coincidental. ExplodingCabbage (talk) 11:28, 22 June 2024 (UTC)

We're inconsistent about whether the "Date" is date of attack or date of death
Somebody could clean this up, if they were minded to, be going over every case in the table. Would be good to get a consensus here first on which date is preferable to use. ExplodingCabbage (talk) 17:51, 16 June 2024 (UTC)


 * One current example of each: Ken McCall's "Date" is given as 18 November 2015 (when he died, several days after being attacked), but Damian Holden's is given as 18 June 2009 (when he was bitten, months before he actually died). ExplodingCabbage (talk) 17:53, 16 June 2024 (UTC)


 * Date of attack, because the page is list of attacks, not list of deaths. Put date of death in the notes, or next to the date of attack.
 * Wikianon3770617 (talk) 18:06, 16 June 2024 (UTC)
 * I would suggest putting the date of the attack in the column and the date of death either below that date (with a death symbol) or to put it in the description of the case. ("The victim died 3 weeks later on ..."). Wikigrund (talk) 10:35, 22 June 2024 (UTC)
 * Do you have a suggestion for a "death symbol" to use, and is there precedent of it being used on Wikipedia? Sticking something like a grinning skull-and-crossbones emoji next to an account of e.g. some toddler getting mauled to death by a dog, especially if their parents may still be alive to see it, feels crass and disrespectful; I'm inclined not to do that!
 * I agree with the general idea, though. Just think I'd rather want to go for something more like "Died: 1 Jan 1990" instead of using a death symbol, unless there's a symbol whose use has precedent. ExplodingCabbage (talk) 11:15, 22 June 2024 (UTC)
 * August 17, 2007
 * August 26, 2007 †
 * You can find an example where it was used like this here: Fatalities in France
 * This symbol is often used next to the year or date of death on Wikipedia (pages about people). Wikigrund (talk) 11:37, 22 June 2024 (UTC)
 * Nice. Yeah, that's much more dignified than the grinning skull I imagined when you said "death symbol". I'll make the change. ExplodingCabbage (talk) 11:42, 22 June 2024 (UTC)
 * I'm still planning to finish this off at some point, but it's blocked by https://en.wikipedia.org/wiki/Talk:List_of_fatal_dog_attacks_in_the_United_Kingdom#What_to_do_about_the_now-uncited_material_from_the_Daily_Mail_in_the_Hirst_and_Banks_cases?; I can't continue this without adding a citation to the Lesley Banks case, and I don't want to do that without clear approval from others on this Talk page, because it's part of what got me blocked recently. ExplodingCabbage (talk) 22:28, 15 July 2024 (UTC)

Summary counts
I've removed the DEFRA and "this article" summary counts for the following reasons: - DEFRA only has a small selection of years (2005-2017), with no further numbers since then - DEFRA figures contradict ONS which is more widely used and trusted globally - DEFRA figures are lower in almost every year, giving an inaccurate picture - Article list is not exhaustive, particularly figures from early 2000s - ONS summary count is sufficient Feudonym (talk) 00:57, 20 June 2024 (UTC)

What to do about the now-uncited material from the Daily Mail in the Hirst and Banks cases?
The consensus reached in a recent discussion on this Talk page was to remove or re-source any information we are currently getting from the Daily Mail (proposed by @Nikkimaria & @Wikigrund in https://en.wikipedia.org/wiki/Talk:List_of_fatal_dog_attacks_in_the_United_Kingdom#Restoring_a_couple_of_Daily_Mail_links...)

I did so, but my work was first reverted on the ostensible grounds of being "OR", and then, when I had a second go with an extra source added, reverted again on the grounds of being "POV trolling". Both charges seem clearly spurious to me, but since I also got a two-week block for my troubles after the other editor did their second revert, I am hesitant - purely for reasons of self-preservation - to reimplement the changes without having unambiguous approval on the Talk page first.

Unfortunately the editor who reverted them has made very clear that they consider my attempts to discuss the matter with them to be harassment, so I will not be attempting further discussion with them and we cannot expect them to explain themselves here. I therefore would appreciate others chiming in.

An overview of the current situation and my proposed changes:

Archie Hirst

With the Daily Mail citation (https://www.dailymail.co.uk/news/article-1048192/Rottweiler-mauled-baby-death-walked-months.html) removed from the article, there is no longer any citation of a source that covers the inquest, and therefore also no source that covers the dog's lack of exercise (which emerged at the inquest) nor any expert commentary about the effect of this on the dog's mental state. An alternative source exists, http://news.bbc.co.uk/1/hi/england/bradford/7576074.stm, which covers broadly the same material, albeit with minor differences - the BBC article quotes Mullen as saying the dog had insufficient "mental stimulation" whereas the Mail article claims the lack of exercise made it "neurotic".

The Wikipedia article currently states that the owners "were said" (by whom?) to "never" (hyperbolic!) exercise the dog, and that this added to its "agitation" (kinda supported by the Daily Mail article if we assume "neurotic" is close enough to "agitated", but only implicitly supported by the BBC article, which doesn't definitively state anything about the impact of the dog's lack of exercise).

To resolve the existing issues and adjust the prose to reflect the slightly different details reported by the BBC, I made the following (to date, still reverted) edit: https://en.wikipedia.org/w/index.php?title=List_of_fatal_dog_attacks_in_the_United_Kingdom&diff=prev&oldid=1230390472

I continue to think this edit should be made; right now, the current content is unsupported by the existing citations. Does anyone see any merit in the claim that this is either OR or POV? Do others approve of restoring that edit?

Lesley Banks

Almost all the details in the "Circumstances" column are unsupported as of the removal of the Daily Mail citation. In https://en.wikipedia.org/w/index.php?title=List_of_fatal_dog_attacks_in_the_United_Kingdom&diff=prev&oldid=1230388447, I proposed replacing that citation with a Daily Mirror citation, which supports all the facts already in the article; I note that the Mirror is not a deprecated source. (I previously was concerned that the Mirror article was simply plagiarised from the Mail, but I think I was wrong about this; the Mirror article contains lengthy quotes from the coroner that are absent from the Mail article, indicating that they did their own research or at worst were working from the same press release.) This too was reverted as part of the "OR" and "POV trolling" reverts (though it is hard to see how replacing a Mail citation with a Mirror citation could possibly be either of these things).

Can anyone see any objections to re-adding the Mirror citation, so that the current content relating to Banks is sourced? ExplodingCabbage (talk) 22:24, 15 July 2024 (UTC)


 * Thanks for the edits, @Wikigrund. Just two questions:
 * Why a Wayback Machine link for http://news.bbc.co.uk/2/hi/uk_news/england/bradford/7576074.stm? The article is still available on the web; I would've thought it'd be more normal to link to it directly.
 * Your edits eliminate any explicit mention of the dog's lack of exercise which seems relevant to me; any objection to changing "The grandparents had only acquired her 6 months prior" to "The grandparents had only acquired her 6 months prior and had not walked her for 5 months, according to testimony at the inquest." to include this detail?
 * ExplodingCabbage (talk) 11:28, 17 July 2024 (UTC)
 * The archived link is an additon, the link to the original is still there.
 * I removed it because I added "According to a consultant, the dog was kept in the back yard and did not have enough mental stimulation." For me, this indicates that the dog was not walked, so I have removed the part of the sentence. But I have no objection to phrasing it like you suggested. Wikigrund (talk) 13:31, 17 July 2024 (UTC)

Our citations give different annual totals for lots of years - and sometimes we have more than any cite does!
We cite three ONS sources:

A: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/adhocs/006077deathsfromdogbitesengland1981to2015

B: https://www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/deathsandhospitalisationsfollowingadogbitefrom1981to2019

C: https://www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/deathsbydogattackinenglandandwalesbyage1991to2023

The years they cover overlap, and they don't give the same number of deaths. The years on which they disagree are:

We currently go with source A over source C, but I'm not sure why, and I'm also not sure what explains the differences. I thought it might be to do with one source counting only dog BITE deaths while the other includes dog STRIKES, but that doesn't explain it because neither source is consistently higher than the other. Nor can counting by year of death vs year of registration explain it, since the missing deaths don't appear to consistently shift to an adjacent year.

There's also a year - 2012 - for which our sources agree there were only 3 deaths, but we've collected 4 in the article, proving the ONS is undercounting. @Nix D edited the table (https://en.wikipedia.org/w/index.php?title=List_of_fatal_dog_attacks_in_the_United_Kingdom&diff=prev&oldid=1234114725) to list the number of deaths in 2012 as 4, letting the count of our own collected cases overrule the ONS when it's higher. That might be a reasonable thing to do, except that right now we specifically claim that the summary counts we're giving are "From the Office for National Statistics", and as of Nix's edit that's not entirely true any more.

How should we unpick this? Why do we think we're seeing different numbers from different sources, which should we trust, and how should we tweak the presentation of the summary counts table so that we're not falsely attributing our own count to the ONS? ExplodingCabbage (talk) 20:36, 17 July 2024 (UTC)