Talk:Fatal dog attacks in the United States/Archive 14

Discussion regarding inclusion of breed information
This may be jumping the gun since there is an ongoing AfD for this article, but I would like to start this conversation regardless. One issue that has come up multiple times in the AfD, and that certainly has the most total words discussing it, is the inclusion of breed information in this article. To be very clear, the discussion about whether that information warrants deleting the article belongs at AfD; here I would like to have a discussion about what to do with this article if it is not deleted. In my view there are two main questions that need to be discussed:


 * Is there evidence in reliable sources that breed identification is wrong often enough to matter? A 2% error rate is irrelevant. A 98% error rate is a huge problem. Where do reliable sources actually put this?
 * If there is, then is there evidence in reliable sources that breed identification is often wrong in a way that is material to this page? Assuming near-misidentifications matter less than egregious misidentifications, how badly are breeds being misidentified? Are these instances of Alaskan Malamutes and Siberian Huskys not being distinguished, or is this calling any dog with a blocky head a pit bull?

What do do with breed information in the article depends greatly on the answers to those questions, but there are three options I would like to put out there for discussion:

1. Keep the article as it is.

2. Remove the 'Category of Dog' column. This option would retain breed information within the 'Circumstances' column but would diminish its overall prominence within the article.

3. Remove all breed information with the exception of a) reported DNA test results or b) the dog is identified in a RS by an expert, e.g., a veterinarian, and potentially other criteria that there may be consensus for.

There are a lot of strong feelings surrounding this so please, assume good faith, be civil, stay on topic.  Pais  a re pa  05:34, 20 September 2020 (UTC)


 * Yes, this is jumping the gun, because I'm sure we're all jaded with the lengthy AfD. And since it's still undecided, why are we having this conversation? Couldn't you have waited until the AfD was completed?


 * DNA testing was NOT available for the majority of earlier fatalities. DNA testing IS NOT available for most of the dogs involved in a fatal attack. DNA testing has been unreliable when it has been available in studies, because most of those studies were to prove that "you can't tell a dog is a pit bull" and the only testing laboratory that has been used has been MARS Wisdom and it doesn't even test for the breed American Pit Bull Terrier.


 * Did you read the information at the top of this Talk page in the pink? The guidelines for this page on listing breed were established long before I came along (I joined about two years ago).


 * Yes, keep the breed column. Everyone always wants to know what breed it is. Breed reported by media is usually obtained from the police blotter. Police get it from their own observations, what the owner said it was, what Animal Control said it was, etc. Why would we second guess them. Ultimately, does it really matter what breed is reported or if it's occasionally wrong?


 * — Normal Op (talk) 05:53, 20 September 2020 (UTC)
 * As I said originally, please assume good faith. All the same, thank you for taking the time to provide your opinion.


 * I started this conversation while the AfD is ongoing for two reasons: 1. Many of the arguments being made there are not, in my opinion, directly relevant to deletion but are absolutely relevant talk page discussions. 2. It appears likely to me that it will close as 'no consensus'. 3. This is an opportunity to have a constructive conversation while a lot of interest has been created by the AfD. If others share your feelings of being jaded then this discussion won't go anywhere and that's okay too.


 * Keep in mind that consensus is fluid and is overturned simply by finding new consensus. The existence of a guideline does not in any way preclude a conversation about changing it.  Pais  a re pa  06:33, 20 September 2020 (UTC)
 * To reply to your statement of Ultimately, does it really matter what breed is reported; isn't that an argument for removing the breed information? If it doesn't matter what breed is reported, why does it matter if it is a prominent aspect of this article? I appreciate your other input about the bias of studies, the lack of testing for American Pit Bull Terrier, etc., but I would like to see those allegations supported by reliable sources.  Pais  a re pa  06:42, 20 September 2020 (UTC)


 * The AfD is over now (no consensus). I have repeatedly answered up about the lack of testing for APBT, but I will gather and post that here for you. (I see Astro$01 has answered, too, while I was drafting this response to you, and I will put my study information under his later.) To clarify better about my "Ultimately, does it really matter what breed is reported" comment: if the police, the owner and the ACO report the dog as Breed X, does it really matter if DNA would have said it was something different. There are many reliable sources that state that when a mixed breed dog presents (phenotype) predominantly as a particular breed, it tends to behave similar to the breed it most looks like. I see no reason why not to continue presenting the breed designations as reported in secondary sources such as media (who got it from the police blotter) or police press releases.


 * The studies commissioned by the pro pit bull contingent (aka pit bull lobby) are all designed to conclude that no one can identify a mixed breed dog's composite breeds... with the ultimate goal of presenting the argument that a particular breed (pit bull breeds in particular) has no specific "set of behaviors" upon which buyers/adopters can rely (even though "we know them by their actions"). Every other breed presents their dogs as having a set of common behaviors and traits... except the pit bull breeds who present only the positive traits; negative traits are explained away with "no one can identify a dog's breed by its looks" and "it must have been trained to do that". Of note is that a pit bull's looks are dominantly stamped on its offspring and a 50/50 mixed breed dog (half of which is a pit bull breed) will most likely still look like a pit bull. Drop to 25% and a pit bull's looks may still shine through, though less frequently than if 50/50.


 * The only "breed" for which counter arguments are presented is 'pit bull'. No other breed club, association, society or registry is moaning about one of theirs killing someone. Even when the owner whose pit bull killed someone says "no, it's a Boxer mix!" we don't see the Boxer associations bitching about how it was NOT a boxer. There is an entire industry surrounding the saving of pit bulls from euthanasia (no kill, retail rescue, rescues/donations, transport, etc.). No one else is trying to 'save' dogs that kill people. It's a big money-making operation and a lot of agencies have jumped on the band wagon. I cannot do anything about the propaganda that is presented outside of Wikipedia, but I vehemently object to those same agents using Wikipedia to forward their agenda. And since "pit bull" is the most frequently occuring breed designation on the List of fatal dog attacks in the United States the article itself has (unfortunately) become part of the pro-pit/anti-pit battleground. Anyone arguing for keeping the breeds is labelled "anti pit bull" by those wanting the breed information removed. Normal Op (talk) 14:59, 20 September 2020 (UTC)


 * I realize that this is largely an argument around pit bulls. However, I saw arguments of bias on both sides. I am familiar with the arguments; what I'm not familiar with is the evidence in reliable sources, which is what matters on Wikipedia.  Pais  a re pa  15:50, 20 September 2020 (UTC)
 * I agree sources are needed. In particular, I would like to see reliable sources for these claims: There are many reliable sources that state that when a mixed breed dog presents (phenotype) predominantly as a particular breed, it tends to behave similar to the breed it most looks like. and The studies commissioned by the pro pit bull contingent (aka pit bull lobby) are all designed to conclude that no one can identify a mixed breed dog's composite breeds. PearlSt82 (talk) 16:07, 20 September 2020 (UTC)


 * The Journal of the American Veterinary Medical Association published an article in 2013 that looked at the accuracy of breed reporting in public media, a point that was raised in the AfD but which may have been lost in all the turbulence. The digital original is behind a paywall, but a PDF of a photocopy of the print article is posted here: Co-occurrence of potentially preventable factors in 256 dog bite-related fatalities in the United States.
 * The authors reviewed dog bite fatalities from 2000 through 2007 and compared the dog's breed as reported in media with the reports of investigating law enforcement personnel to see where, in the authors' view, media accounts diverged ("were discordant") with each other and where they diverged with law enforcement recollection and records. A "strict definition" required the different sources to use the same words to be in agreement, e.g., "Labrador Retriever" would not be concordant with "Labrador Retriever mix", whereas an "expanded definition" would treat "Labrador Retriever" and "Labrador Retriever mix" as being concordant. The study found:
 * "For single dog incidents (148 incidents), on the basis of the strict definition (exact match), breed descriptors in media reports were discordant for 32 of 148 (21.6%) dogs; animal control or local law enforcement assessment of breed differed from the media account for 45 of 129 (34.9%) dogs."
 * "On the basis of the expanded definition (any agreement between alleged breeds and mixes), breed descriptors among media reports were discordant for 19 of 148 (12.8%) dogs; animal control or local law enforcement assessment of breed differed from the media account for 18 of 129 (14.0%) dogs."
 * The bottom line was that there was an exact match in ~65% of the cases and general agreement in 86% of the cases.
 * From a practical standpoint, breed identification is always about the subjective evaluation of a dog's appearance because breed evaluation evolves around how closely a particular dog adheres to a published breed standard. Law enforcement tests dog DNA in dog attack cases when there is a question whether a particular dog were involved pre- or postmortem, but that's about it. Given the subjective nature of breed identification, even by "experts" such as veterinarians and dog show judges, I look to the field of business intelligence, where an 85% level of confidence is generally considered "high confidence" (99% confidence is "near-certainty"). I therefore think the answer to the first question is, "The evidence is that breed identification is not wrong enough to matter" because we can have 'high confidence' that the WP:RS reporting is in fact 'reliable' when it comes to identifying dog breeds."
 * I think the "Category" or "Type of Dog" should remain as its own column. Law enforcement does not care about the breed of dog, but the population as a whole generally does. The reason media reporting includes the type or breed of dog is because one of the first questions people ask is, "What kind of dog was it?" It seems to me the list is more useful because high-interest information is displayed prominently rather than being lumped into the "Circumstances" column, so I would "Keep as is" Astro$01 (talk) 14:41, 20 September 2020 (UTC)
 * The upper level findings of this study state: For 401 dogs described in various media accounts, reported breed differed for 124 (30.9%); for 346 dogs with both media and animal control breed reports, breed differed for 139 (40.2%). Valid breed determination was possible for only 45 (17.6%) DBRFs - this (in my opinion anyways) doesn't give me high confidence, especially when there are 400+ dogs involved in fatalities over a 10 year period. PearlSt82 (talk) 15:01, 20 September 2020 (UTC)
 * Which raises the question, "Why are the top level conclusions not supported by the actual data they presented?" Astro$01 (talk) 15:16, 20 September 2020 (UTC)
 * They are - the numbers you posted are from the "single dog incidents" dataset, which makes up only 148 incidents, as other incidents involved multiple dogs. Even still, assuming an upper bound of an 85% success rate, 85% of 400 dogs is equal to 60 misidentified dogs over 10 years. Again, in my opinion, way too large of a margin of error. PearlSt82 (talk) 15:18, 20 September 2020 (UTC)
 * Actually, it doesn't amount to "misidentified dogs", it amounts to "disagreement over a subjective evaluation standard." To assert otherwise is to to assert that 60 to 70% of the population can't tell the difference between a Chihuahua and a Great Dane. And this is the basis for all this ruckus? Seriously? Astro$01 (talk) 02:25, 21 September 2020 (UTC)

Pinging user:Eddie891, user:Dream Focus, user:AleatoryPonderings, user:Rhododendrites, user:Hog_Farm, user:Deacon_Vorbis, user:Berrely, user:Onel5969, user:Lightburst, user:Gleeanon409, user:Michael2468b, user:Casliber, user:Ajf773, user:GreenC, user:SilverTiger12, user:Mariolovr, user:ImTheIP, user:7&6=thirteen, user:Red_Rock_Canyon, user:Guerillero, user:Atsme, user:power~enwiki, user:Stifle, user:Cavalryman, user:Pawnkingthree, user:Megalibrarygirl, user:Oeoi, user:Usedtobecool, user:Newshunter12, user:Kind_Tennis_Fan, user:StellarHalo. This should be every editor who participated in the last AfD.  Pais  a re pa  16:07, 20 September 2020 (UTC)
 * My final comment here and I'll begin by asking all of you a question from a BLP policy POV: What if the name of your child or family member was included on this list - how would you feel? How would you like to see your dead baby's name and the cause of death published for all the world to see with the description of the mauling and the circumstances? What do you think that does to a parent? If this isn't a BLP vio, nothing is. Newspapers publish news and then the story goes away - encyclopedias are supposed to publish lasting, notable & ACCURATE encyclopedic information that is corroborated and verifiable, and does not violate any of our polices, particularly BLP and our 3 core content policies. We are NOT a memorial, not a statistical list or indiscriminate collection of information, not news, not a platform for advocacies or necrophiliacs and the like, and we should not be publishing material that is inaccurate or cannot be verified/corroborated or that is based on ignorance/emotion/misinterpretations/failed id. I've already mentioned the other policies in the AfD. Neither news orgs, police reports nor forensic reports can accurately identify the dog(s) - it is strictly based on looks and even that has failed - the closest we can get to accurate dog IDs is via genetic testing. I strongly oppose this list for all the reasons I've mentioned. I feel the same about the other indiscriminate death lists of non-notable people - they violate our policies. That's all I'm going to say, so please don't ping me anymore. This is page is off my watchlist. Atsme Talk 📧 16:36, 20 September 2020 (UTC)

Your entire argument is an appeal to emotion and is based on entirely flawed comparisons. A dog breed is in now way comparable to the name of a deceased human. Your hypothetical list would probably list a country of citizenship which happens to be the closest analogue for humans. It truly is shocking the levels of argument and appeals to emotion that the pro pit bull lobby uses to wipe the record of the horrific actions of this breed - actions proven time and again by verifiable statistics. — Preceding unsigned comment added by Eldrichr (talk • contribs) 19:18, 28 September 2020 (UTC)


 * I agree with Atsme's comment. Honestly Wikipedia is not an indiscriminate collection of information and I still don't believe this article should be kept. Also, I suggest not doing mass pings like this. Most people would really not be interested. — Yours, Berrely  • Talk∕Contribs 16:47, 20 September 2020 (UTC)
 * I appreciate both of your input that the article should have been deleted, but (for now at least) it is staying so it seems reasonable to try to find consensus on the concerns raised in the AfD. Many people were already having this conversation there where it won't actually affect the page when deletion fails -- it seemed obvious to move it here where it can actually change the content of the page. I chose to discuss breed to avoid having too broad of a topic, but there is absolutely an argument to remove individuals' names from the article as well and only list "adult male", "adult female, 55" etc. It isn't deleting the article, but consensus is never all-or-nothing. I was unaware that being pinged is considered an inconvenience, so I apologize for that. Many people were clearly interested and I wasn't about to selectively ping only a few people in the AfD to avoid the appearance of canvasing. I assumed others were like me and just ignore pings if it happens to be something they don't care about, so again, I apologize.  Pais  a re pa  01:51, 21 September 2020 (UTC)


 * Keep as is We all know what a Pitbull and other breeds look like. I am sure dog owners have also contributed to the reports by confirming the involved breeds. The differences between breeds is apparent and significant. Finally removing the category would greatly diminish the list. Lightburst (talk) 16:38, 20 September 2020 (UTC)
 * Option 2, with attribution. PearlSt82 (talk) 16:53, 20 September 2020 (UTC)


 * *Option 2 or 3 due to unreliable breed information in media reports on dog bites per multiple studies already cited and discussed here. For example, per Inconsistent identification of pit bull-type dogs by shelter staff, even though the study lumped all of the various pitbull-type breeds into one "pitbull-type" category (theoretically making it easier to identify a pitbull-type dog accurately), the study still found that DNA breed signatures identified only 25 dogs (21%) as pit bull-type, shelter staff collectively identified 62 (52%) dogs as pit bull-type. In other words, 37 of the 62 dogs that were visually identified to be "pitbull-type" did not have any DNA from any of the "pitbull-type" breeds included in the study -- that is an error rate of ~60% (37÷62). Again, multiple studies have found that casual visual breed identification techniques (matching physical attributes to breed) are unreliable with various rates of error, depending on the study. Therefore, when an owner/shelter/animal control officer (which is usually the primary source of breed information in most media reports on dog bites) provide breed information, there’s a good chance (per studies) that the visual breed assignment/label is not accurate. Again, I don’t think anyone is mistaking Chihuahuas for Bulldogs or Dachshunds for German Shepherds - the challenge is with specific unique breeds within breed types/groups/categories.. e.g. within the diverse bully-type (bulldog-type) group such as the various pit bull-type breeds and other similar breeds and mixes such as Dogo Argentino, American Bulldog, Cane Corso, Boxer dog, and many others. And unsurprisingly (per the DNA studies) there is a higher level of agreement/accuracy for purebred dogs than for mixed breed dogs. However, accuracy declines significantly when attempting to accurately identify, for example, an American Staffordshire Terrier mix vs a Dogo Argentino mix vs an American Bulldog mix vs a Cane Corso mix. As the majority of dogs in the U.S. are mixed breed dogs per  genetic analyses identified 125 distinct breeds with 91 breeds present at both shelters, and 4.9% of the dogs identified as purebreds, the accuracy (or lack thereof) of the identification of mixed breed dogs is an important point for this topic. For those that disagree, respectfully I'm not going to revisit back-and-forth discussions/arguments on the validity of these studies (while they are enjoyable as they are intellectually engaging, I've already provided my points in previous and recent discussions on these disagreements) -- so I'm only providing my feedback here for which options I find to be most appropriate for the main page. Thanks, Michael2468b (talk) 01:59, 21 September 2020 (UTC)
 * - Opinion stricken. Michael2468b was a confirmed sock puppet. Astro$01 (talk) 03:41, 28 September 2020 (UTC)
 * So this isn't about dogs in general - it's only about pit bulls - which aren't even a breed but a general type based on generally shared attributes and appearance. Good grief. Tell you what: we can put some text up at the front that says something about pit bulls being a type of dogs with shared physical attributes, and that when you read "pit bull" it could be a purebred or mix of several named types of dogs, such as the ones you named. That should do it. Astro$01 (talk) 02:33, 21 September 2020 (UTC)
 * That wasn't the intent of my comment, the referenced studies include discrepancies for many breeds, not just the pit bull-type breeds -- I only referenced the discrepancies for pitbull-type breeds as an example since this appears to be a larger point of contention in this discussion. There are many other breeds and mixes that get misidentified in the studies. Also, I don't think that including, for example, a Dogo Argentino (either purebred or a mix) in the "pitbull type" category would be a solution (even with a disclaimer of sorts) as this is a unique/specific breed that as far as I know, is not included (by law or by any canine registries) as a pitbull-type breed. But to your point "pitbull" is not a breed, so why then is it listed on the main page along with other specific breeds? Instead of attempting to list specific breeds in the main article (which is prone to inaccuracies/errors and disagreements, as discussed), listing the major breed types/categories instead and consolidating the breeds into these broader categories may be a better and more accurate approach -- so for example, "Category of Dog" would include "Shepherd-type" (includes German Shepherd, Belgian Malinois, etc.), "Hound-type" (Beagle, Coonhound, etc.), "Bulldog-type" (Dogo Argentino, American Bulldog, Boxer, the various breeds classified as pit bulls, etc.), "Sled-type" (Malamute, Husky, Chinook, etc.). Otherwise, it's the equivalent of listing car incidents/statistics as: Honda Civic, Ford Bronco, Toyota Camry, four-door sedan, Volkswagen Jetta, etc. (in this example, the specific car models would be specific breeds like "German Shepherd" and "pitbull" would be four-door sedan) -- it's not appropriate to mix and intermingle types/categories of dogs along with specific breeds. So instead, why not just list the broader higher-level categories (Shepherd-type, Bulldog-type, Hound-type, etc.) which at least would generally have more agreement/consensus vs. listing specific breeds. I think moving from specific breeds to the broader categories of dogs/breeds would resolve most of the contention/disagreements being discussed here. Michael2468b (talk) 03:11, 21 September 2020 (UTC)
 * - Opinion stricken. More sockpuppetry from Michael2468b. Astro$01 (talk) 21:21, 28 September 2020 (UTC)
 * I did some extensive reading from the breeders/founders of the Dogo Argentino. They most definitely did include pit bulls in the mix to create the breed, and not a small portion. In some countries, Dogos are included in lists of prohibited or restricted breeds under the presumption that they are too dangerous to have "as pets". Normal Op (talk) 03:25, 21 September 2020 (UTC)
 * While the foundation stock of Dogo Argentinos does share some similarities to the foundation stock for pit bull-type breeds (e.g. the Bull Terrier), their foundation stock includes 10+ breeds and they are considered unique breeds by all canine registries/organizations and by their DNA -- ergo, it would be inaccurate to simply classify all bulldog-type breeds (such as the Dogo Argentino) as pit bull-type breeds. Yes, Dogo Argentinos are on some breed-restricted lists; however, that point is mute as this has no bearing on the fact that they are not pit bulls. Great Danes, German Shepherds, Rottweilers, Dobermans, Chow Chows, and many other large or strong breeds are also on some breed-restricted lists -- however, that doesn't make them all "pitbulls" nor does it make them all inappropriate to own "as pets" (as with anything, safety relies on the human). Also, to clarify, per the Wikipedia pages for Dogo Argentino and American Pit Bull Terrier, the foundation stock of Dogo Argentinos includes 10 breeds and the foundation stock of the APBT includes only 3 breeds - the only one breed they share in common is the Bull Terrier. If we're going to be breed-specific, then we need to be breed-specific. Michael2468b (talk) 03:37, 21 September 2020 (UTC)
 * Yet more sockpuppetry from Michael2468b. Astro$01 (talk) 21:21, 28 September 2020 (UTC)
 * I'm sorry, but changing a WP:RS source that says "pit bull" to a Wikipedia entry that says "pit bull-type" (or "Coonhound" to "hound-type", etc.) violates WP:OR, because you're substituting what you wish the WP:RS had said instead of what it actually said. Astro$01 (talk) 05:00, 21 September 2020 (UTC)
 * Understood; however, it's not about me wishing to change what a WP:RS has published; but instead, I was just providing suggestions as possible solutions to some of the disagreements discussed here -- if the suggestions (such as categorizing the various individual breeds by categories/types based on the breed information provided by the WP:RS) violate WP:OR then of course I agree they shouldn't be considered. However, I still stand by my original reasons for selecting options 2 or 3. Thanks, Michael2468b (talk) 05:18, 21 September 2020 (UTC)
 * The Michael2468b sock puppet show has ended. Astro$01 (talk) 21:21, 28 September 2020 (UTC)

Arbitrary break

 * Here is the info you requested. I found two studies for you that disclose the limitations of the Mars Wisdom DNA testing. Both studies make conclusions about the difficulty of pit bull identification by observers while not being able to test for 'pit bulls' or the most common breed of pit bull, the American Pit Bull Terrier (APBT).


 * 1. Inconsistent identification of pit bull-type dogs by shelter staff (Olson, 2015). To wit: "Mars Veterinary contributed DNA testing services for breed identification." and "Samples were shipped to a commercial DNA analysis laboratory (Wisdom Panel Professional Canine Genetic Analysis, Mars Veterinary)" and "Limitations of our study include unknown sensitivity and specificity of the DNA breed testing and lack of a DNA test for American pit bull terrier." and "the largest testing service does not offer a DNA test for identification of American pit bull terriers . Additionally, it does not provide a test for ‘pit bulls’, since the term variously refers to a loose collection of breeds and their mixes or to dogs with similar morphology rather than a group of dogs with a controlled gene pool." and "For the purposes of this study, the terms American pit bull terrier, American Staffordshire terrier, Staffordshire bull terrier, pit bull, and their mixes were included in the study definition of ‘pit bull-type breeds’ because these terms are frequently included in laws regulating dog ownership based on breed or phenotype."


 * 2. Is That Dog a Pit Bull? A Cross-Country Comparison of Perceptions of Shelter Workers Regarding Breed Identification (Hoffman, 2014). To wit: "the validity of genetic testing to determine pit bull ancestry is also deemed questionable due to the pit bull’s nature of being a “type” that may source from a range of purebred bull breeds (Wisdom Panel, 2013). Furthermore, although the genetic tests recognize some bully-type breeds, they do not recognize the American pit bull terrier ."


 * So let's take these two studies' disclosures about Mars Wisdom Panel DNA test and expand outwards to those studies that relied on Mars tests to identify pit bull DNA and come to some conclusion about breed identification or pit bull behavior. Both Voith 2009 and Voith 2013 also relied on Mars Wisdom DNA testing, did NOT provide any disclosure of the limitations of the testing, and made conclusions about pit bull identification. (Voith incidentally is listed as an expert for NCRC on NCRC's website which is wholly owned by American Farm Foundation whose goal is admittedly "To secure equal treatment and opportunity for 'pit bull' dogs.")


 * Further studies relied on (built upon the conclusions of) these earlier studies and their faulty conclusions. In all, I don't think there's a single study that used DNA that didn't use Mars testing service. And like I said before, no other breed group is trying to prove that you cannot identify a dog's breed by looking at it (it's only the pro pit bull groups/lobbies/whatevers) so all of the studies leading to this conclusion were done by and for the pit bull people. This is why the Mars DNA exclusion of American Pit Bull Terrier is so important in the evaluation of this matter.


 * — Normal Op (talk) 17:17, 20 September 2020 (UTC)


 * Keep as is In addition to governments (which have made breed specific prohibitions), insurance companies have made similar judgments and refused to insure homeowners who harbor certain breeds.  Relevant information.  Reliable.  Follow the money.  7&amp;6=thirteen (☎) 17:27, 20 September 2020 (UTC)
 * Now I will show you that even the legal profession won't use DNA evidence because it's unreliable!
 * Attorney Adam Karp (who purports to be a legal defender in Dangerous Dog Classifications ) penned a paper in 2014 called "DOWN TO A SCIENCE: Combating Breed Discriminatory Litigation with Frye, Daubert, and Rule 702". This paper has become the "road map" to defending your dog in a dangerous dog hearing and in getting breed-specific legislation repealed or overturned. In this paper Karp mentions the case where someone tried to get Mars Wisdom to come to court and testify about their DNA tests. Mars refused! Let's take a closer look at what bits-and-bobs are in this paper about DNA and its reliability.


 * Footnote on page 19 says: "The Wisdom Panel 2.0 does not, at this time, detect the American Pit Bull Terrier (“APBT”), placing in doubt the test’s probity where BSL excludes or includes the APBT. For instance, if excluded from the legislation, the subject dog’s genetic composition might be mistakenly matched by the Wisdom Panel to other BSL-prohibited breeds for which Mars does test (e.g., Staffordshire Bull Terrier (“SBT”)). The risk of a false positive for SBT might increase by the absence of APBT in the breed detection database. Accordingly, reliance on the MARS test may violate equal protection since its database excludes APBT genetic markers. Such a dog simply cannot be tested reliably, and would be erroneously reported as genetically comprised of other breeds . Further, MARS sends a form letter explaining that different specimens from the same animal may produce different results."


 * This brings in doubt all of the studies using MARS to detect or not-detect "pit bull" and make conclusions based on these DNA test results. Mr. Karp goes on to write on page 17: "On par, is it strategically preferable for the dog owner to object to the use of any form of identification (visual or DNA) as untenable under Frye, Daubert, and ER 702, thereby preventing the BDL [BSL] enforcer from satisfying the threshold burden of production?" and "But such DNA-based evidence relied upon by, say, a veterinarian or an animal control officer results in an opinion that is only as reliable as its source. If the foundation remains impervious to critical scrutiny due to Mars’s refusal to share its proprietary methodology, then one must indirectly challenge the professional qualifications of the expert who purports to testify."


 * Which leads me to point out that these studies all point to the conclusion "you cannot identify any dog breed visually" when they are relying on "DNA evidence" that is, IN FACT, unreliable and prone to false results!


 * Karp's suggested strategy is at play here in Wikipedia discussions and content-wars, and has been the preferred strategy of the pro-pitbull contingent since he wrote this in 2014. You simply shoot down the credentials of anyone who is "testifying", you present whatever evidence you want, but the SENIOR DATUM that is presented over and over and over is that "no one can correctly identify a dog's breed visually"... extrapolated into "the police, animal control officer, dog's owner, and the media can't correctly identify a dog's breed without a DNA test, therefore every single entry in List of fatal dog attacks in the United States is suspected as being incorrect".


 * That statement is false, because visual identification is done every day with good results. The thing that can't be correctly identified is a "pit bull" dog's breed by using Mars Wisdom DNA testing... BECAUSE they don't test for American Pit Bull Terrier and APBTs are the most prevalent "breed" within the pit bull group.


 * I can assure you, no one at animal control or police are doing DNA tests on dogs that have fatally attacked a human in order to "determine breed". They may do one to determine if a particular dog's DNA matches the saliva found on the victim in order to include or exclude a dog as being a perpetrator (such as in a multiple dog scene to determine which dog did NOT also attack) but that sort of test does not also determine breed (different test). And if neither the police are testing the dogs for breed and the lawyers can't use Mars DNA results in a case, why are we — here in Wikipedia — relying on study results that rely on the same unreliable DNA test? Why are we wasting our time arguing should we require DNA results in order to label a dog's breed in a fatality incident? The DNA results aren't available for these dogs, and if it was available then the results would be unreliable. Visual identification by persons on the scene (including the owner) are far more reliable than any nitpicking we can do here as Wikipedia editors.


 * — Normal Op (talk) 01:18, 21 September 2020 (UTC)


 * Option 2, to give my own opinion.  Pais  a re pa  03:01, 21 September 2020 (UTC)
 * Option 1, to formally state my vote. (For those voting option 2, I can't even begin to imagine who might do the work to move the breed from the breed column to the circumstances column for hundreds of entries.) Normal Op (talk) 03:30, 21 September 2020 (UTC)
 * Comment User:Normal Op I agree with all of your observations. I have tried cases involving expert testimony about the reliability of DNA evidence; but that is not the issue here.  Indeed, I would venture that it is almost never used to determine the heritage of dogs involved in attacks; this is just a Straw man irrelevancy.
 * That insurance companies (follow the money) and governments (follow the law) make these imprecise judgments is an undeniable fact.
 * The more telling fact is that you don't have to be an expert to come with (perhaps ill-founded) rough guesses as to the breed of a dog. It is done every day, and if it is good enough to be reliabily sourced, it is good enough for Wikipedia. Verifiability not WP:Truth.  7&amp;6=thirteen (☎) 15:28, 21 September 2020 (UTC)
 * I think you're on to something, 7&6=thirteen. What I see here in these discussions on Wikipedia (about inclusion/exclusion of breed info in fatal dog attacks) is the presentation of a minority viewpoint — that dog breed cannot be identified visually or without DNA testing — which is an advocate's position. No one who knows anything about dogs really believes this. (I'm not talking about 3- and 4-way crosses or Heinz 57 dogs, which everyone agrees aren't readily discernable.) If the law, insurance, police, public and media don't see a problem with it, why is it that we're arguing about an advocacy position here in Wikipedia? Why would we second-guess what has been presented in reliable source news? This is not really what we're encouraged to do as editors in Wikipedia (see SYNTH, OR, and others). Yes evaluate sources for reliability, but no to (allegedly) evaluating the science behind our reliable sources in order to discredit what has been published. How did we get so far off the mark? Normal Op (talk) 16:11, 21 September 2020 (UTC)
 * Second guessing the sources is WP:OR. And it would be more unreliable ... Verifiability is the simplest benchmark.  Cheers.  7&amp;6=thirteen (☎) 17:30, 21 September 2020 (UTC)


 * First option, keep as is. The article for pit bulls have the reference fact: Independent organizations have published statistics based on hospital records showing pit bulls are responsible for more than half of dog bite incidents among all breeds despite comprising 6% of pet dogs. Reliable sources keep reporting this, that's what we go by, not those who like this sort of dog and are upset people might think poorly of them if they read something bad about them on Wikipedia.   D r e a m Focus  04:09, 22 September 2020 (UTC)
 * There are similar data quality problems on the majority of Wikipedia's list pages. For example, media reports of ages of victims of terrorist attacks are stunningly often wrong. And the only you can know if a report is wrong is by observing multiple reports claiming different ages. But it's not guaranteed that the age is right just because all reports claim the same age. They could have gotten the age from the same faulty "upstream" source or perhaps they copied it from each other. So is the solution to stop publishing the age of victims of terrorist attacks on Wikipedia? I don't think so. Im The IP  (talk) 00:40, 25 September 2020 (UTC)
 * Remove breed type, the names of victims and circumstances - per WP:REDFLAG, WP:RSBREAKING, WP:NOTNEWS, and WP:BLP. After reviewing the cited sources for this list, the majority are cited to local breaking news (published or televised, and are already archived). Breaking news is considered questionable or unreliable because of errors made in haste, including unverifiable circumstances based on anecdotal information. Different considerations would apply if the material is sourced to multiple high-quality RS with verifiable breed ID and circumstances, not guesswork. Visual IDs of breed type are proven to be inaccurate in many instances (including IDs by experts). Atsme  Talk 📧 10:46, 5 October 2020 (UTC)

Page size, entry style, referencing, etc.
At the beginning of the day, this article appeared at #4 at Special:LongPages. I suspect that some quantity of pageviews and deletion attempts are as a result of it having appeared there. I have split out the 2010s into its own page, and when I did it was still at #12 on LongPages. I have cut about 10K from 2010 but that page is still at #14. Many of the flaws appear to have been caused by incremental editing in real-time. Some observations/suggestions for the regular editors to hopefully avoid this and related pages hitting AFD again, particularly since almost all of the attacks are pretty much WP:RUNOFTHEMILL (...WP:CSC#2): ~Hydronium~Hydroxide~(Talk)~ 06:38, 18 October 2020 (UTC)
 * Immediately use inline tags such as update inline or disputed inline where there is uncertainty. For instance, where someone has been charged but the trial is yet to occur, or news reports conflict in their details. Update them when sources becomes available. It is unlikely that what a sheriff would do remains unclear seven years on. An incident is unlikely to still be under investigation 9 months on.
 * When there is certainty/expertise, use that and remove the uncertainty. For instance, if initial reports are that they were pit bulls, but animal control states that it was a mixed breed dog, then use the latter and don't mention the former unless the mis-identification actually has real significance (eg: State legislation is launched to control breed X when the dog was breed Y).
 * In general there's no need to include information that's already covered elsewhere in the row.
 * Review and update sources once they become available. There's no reason to have three 2010 duplicative or overlapping sources if a single 2011 source covers all required information. Cite overload can occur due to dealing with things incrementally in real time.
 * (Re)check facts. Early news reports can be particularly wrong. Interpretations of news reports can be wrong. Opinions and allegations need to be labelled as such.
 * Don't include trivia/material not related to the death/attack (eg: victim had cancer). Avoid what should be obvious, ref Principle of Some Astonishment (the police shot the dog so we can tell from context that they were probably called by someone; the dog was euthanized after the attack, because it is unlikely that this was done before or during). These would help in avoiding WP:NOTNEWS and WP:NOTMEMORIAL.
 * Though these articles are US-centric, enwiki is international. Avoid "called 911").
 * Use a standard "house style" for list entries (The "Killed by..." type was relatively common and brief), and keep it to the bare essentials of attack, forewarnings, environment, and aftermath.


 * I'm not sure what you mean by runofthemill. 'Dog bites man' is run of the mill; 'dog KILLS man' is extraordinary. In such instances people want to know first how it happened, second what kind of dog it was, third what happened to the dog, and fourth did anyone get arrested for it. Rarely do we find media coverage on aftermath of dog owners (such as arrest, indictment, trial, conviction) as it becomes "old news" and isn't often covered by news outlets.
 * Most of your early edits seem perfectly reasonable, but your edit-scalpel started making deeper cuts... in particular the removal of citations. Stop removing details because of some arbitrary ideas you have invented. If the article is still too big, then just split it again — because losing details is NOT the best way to reduce this page's size. Your suggestion of "if initial reports are that they were pit bulls, but animal control states that it was a mixed breed dog, then use the latter and don't mention the former" is poor advice and doesn't lead to more correct information. In the upswing of the popularity of the No kill movement, animal shelters (frequently operated by humane societies contracted by municipalities) stopped using "breed labels" on ALL dogs, and went to a blanket "everything is a mixed-breed and we know nothing" method of handling intakes (because they believed that no one would adopt a pit bull if it was labelled a pit bull). It was a deliberate policy decision based on marketing needs, not for a lack of expertise in pinpointing dog breeds. Initial information reported by media are usually from police blotter information which was obtained by interviews with victims, owners, neighbors, and witnesses, including police officers and attending animal control officers. Any later press releases by animal shelter personnel are useless if they reduce the accuracy of a breed designation.
 * And... and I cannot stress this enough... this page has had a long-standing policy about how breed labels are placed. Often, two citations are given per event — one which has more incident details, and one which has a more precise breed designation. If you are removing one of two (and I noticed you've been deleting one of two on several entries), you may be deleting the one with the breed designation, but at least you are losing half of the information provided. A quick search might well find another article with a breed designation that is later still (to fit your arbitrary "later is better" idea). But the "latest" of several articles doesn't make it more precise. Tomorrow, I will be checking every single entry you modified. I already see several that need fixing. Your edit was definitely bold, but some of it is incorrect and will be undone. I'm not worried about some of the minor trimming you did, but loss of citations and changing of breed is a problem. Normal Op (talk) 07:25, 18 October 2020 (UTC)
 * Thanks for letting me check each entry. I only made changes to 4 (in 2010). Thank you for all your work. Normal Op (talk) 04:39, 19 October 2020 (UTC)

Extremely incorrect number of deaths
First sentence content extremely incorrect information about deaths "Fatal dog attacks in the United States cause the deaths of about 3000 to 5000 people in the US each year,[1]"

link [1] number of deaths estimated 30 to 50 people "However, it is essential to know that, on average, about 30 to 50 people die from dog bites each year" https://www.ncbi.nlm.nih.gov/books/NBK430852/

link [2] average of approximately 19 deaths - "This study utilized the compressed mortality files from CDC WONDER to evaluate deaths from dog attacks over the 27-year period 1979–2005. Results: An average of approximately 19 deaths was reported annually from dog attacks during this time period." https://www.wemjournal.org/article/S1080-6032(09)70079-1/fulltext\


 * Hi! The 3000 to 5000 number was the result of vandalism. It was reverted to 30 to 50 consistent with the cited source. Astro$01 (talk) 04:25, 11 November 2021 (UTC)

Is the list of fatalities even encyclopedically relevant?
I can't find a single dog attack death here that can be said to be encyclopedically relevant / notable. The inclusion of a list of fatalities seems to violate WP:INDISCRIMINATE. Keeping commentary about dog attacks (eg. statistics) is perfectly valid but a list of individual people who have died because of it seems inappropriate for wikipedia. It is also extremely long, and even when they have been split up by decade the articles are also longer than necessary. Perhaps the fatalities list can be moved to wikidata but I cannot see a way that it can belong on wikipedia. Please discuss? QueensanditsCrazy (talk) 15:33, 17 March 2022 (UTC)


 * @QueensanditsCrazy there are many similar lists, such as "cougar attacks in the United States", therefore a larger discussion has to happen. 71.187.245.87 (talk) 20:56, 28 April 2022 (UTC)
 * Good point. IS there a more appropriate place to discuss these issues? I am on the side that that page is also INDISCRIMINATE and so should be deleted or moved to wikidata or something of the sort. Btw do you have a username? QueensanditsCrazy (talk) 21:22, 28 April 2022 (UTC)
 * There have been three debates on deleting this list, most recently about 18 months ago. You may want to review the discussions posted at the top of the talk page. Astro$01 (talk) 16:21, 30 April 2022 (UTC)

Dogbitelaw.com and Dogsbite.org as reliable sources
The use of these sites as references for the number of children killed each year appear to violate SELFPUBLISHED guidelines. I recommend removing this material or finding a better source. Astro$01 (talk) 12:03, 1 June 2022 (UTC)
 * Aside from that, is there anything in particular about it that you object to? – Illegitimate Barrister (talk • contribs), 18:24, 1 June 2022 (UTC)
 * The only thing I'm really concerned about is the source, as the information itself seems appropriate to me in the context of the rest of the introduction. Astro$01 (talk) 01:39, 2 June 2022 (UTC)
 * I have deleted the information from Self-published sources identified since there doesn't appear to have been any effort to find WP:RS for this information. Astro$01 (talk) 03:00, 16 June 2022 (UTC)

This page is absolutely, undeniably used for anti-pitbull advocacy. I believe that is its purpose.
There is a thread today in an active anti-pitbull subreddit with more than 50,000 users decrying the deletion of old data while advocating users to make their own edits. This is a subreddit that, if monitored because moderators delete evidence hourly, you can continually see users advocate putting all pit bulls to death and discussing the graphic ways they personally want to harm pit bull type dogs. There is no question Wikipedia is being weaponized by these groups. 97.113.92.117 (talk) 16:20, 23 July 2022 (UTC)

30+ years of human fatalities by dogs have been removed from this page. Who did that? Anti Pitbull people? I think not. You can’t hide the truth Helperofhumans (talk) 16:39, 25 September 2022 (UTC)

97.113.92.117 never claimed anti-pit bull people were responsible for the deletion of old data. That being said, if there is a concentrated movement of this sort, it's quite disturbing, and something that should be looked out for. Riffraff913 (talk) 01:17, 28 September 2022 (UTC)

Other decades
Were the articles covering the 2000s and 2010s deleted? FinnSoThin (talk) 21:28, 25 June 2022 (UTC)
 * Yes, they were deleted. Someone slipped a deletion request in under the radar for one page and the admin apparently decided to delete a few extra pages as well. See Articles for deletion/List of fatal dog attacks in the United States (2014) Astro$01 (talk) 11:32, 27 June 2022 (UTC)
 * It should have been merged, not deleted. The info should be recovered. Bad oversight by the admin imo. — AdrianHObradors (talk) 15:34, 9 October 2022 (UTC)
 * I agree that the information should be recovered Profpedia (talk) 12:02, 22 May 2023 (UTC)
 * The deleted information should be recovered and re-organized in a similar manner as other countries pages on Wikipedia about fatal dog attacks. For example; Canada's page is organized by before the 2000s and after 2000s and it has not been edited or deleted like the USA page has.
 * Fatal dog attacks are indeed more rare in Canada compared to the United States where it is much more common, but the deleted data can be re-organized by decade instead of by year.
 * The article as of now post deletion of pre-2010 attacks makes it look incomplete or that the editors and admins are purposely omitting data from before 2010 which is making it look biased to one side when it should stay neutral. 184.148.217.186 (talk) 16:36, 24 June 2023 (UTC)
 * Absolutely, I agree the information should be recovered. You can’t just delete pages Wikipedia pages in such an arbitrary manner. Veritas Aeterna (talk) 18:57, 9 May 2023 (UTC)

Possible Fabricated June 27th, 2021 Listing?
the June 27th, 2021 entry for a Cartyr Lou Evelyn Clinger seems to only exist documented on this Wikipedia page. further issues, like the breed of the dog being spelled in lower case as opposed to the Title Casing used throughout the article, ("pit bull," vs "Pit Bull,") and the description of the event being oddly detailed and story-like in contrast to the functional quality of other descriptions, suggest it's a fabricated event. i don't have the ability to edit this article, so wanted to flag. YellabelIy (talk) 15:55, 22 July 2023 (UTC)


 * Thank you for letting us know. I've removed the entry pending someone coming up with a reliable source for the event. — Trey Maturin™ 16:02, 22 July 2023 (UTC)

Problem article
The previous discussions on the talk page have valid points in their criticisms of this article.

First of all, dogsbite.org should not be considered a reliable source, nor should its owner, Colleen Lynn. At the most, it should simply be used as a guide to point to the more reputable sources (similar to how you should not cite Wikipedia, but the sources that Wikipedia itself cites). At the worst, its run by a biased individual who's been accused of presenting misleading statistics and information.

Second, it's very tenuous to list the breed of dog. According to the American Veterinary Society of Animal Behavior, "identifying a dog's breed accurately is difficult, even for professionals, and visual recognition is known to not always be reliable." I'm not sure what the policy is on this with regards to when news reports claim to know the breed involved, but at the very least, this should be added in the header.

Third, why is it only bites for the past few years?

This article is a mess, and these issues need to be addressed. Riffraff913 (talk) 19:58, 29 August 2022 (UTC)


 * I have moved items from this source to the talk page (below) for added research. Rublamb (talk) 03:55, 30 March 2024 (UTC)

Also interesting that the first paragraph makes reference to a study about pit bulls being responsible for more fatal dog attacks than other breeds, yet makes no mention of all the research that says otherwise, that breed is a poor indicator for such uses, or about how difficult it is to define a breed (particularly without genetic testing). Riffraff913 (talk) 20:23, 29 August 2022 (UTC)


 * That is because the multi-million dollar pit-bull lobby funded by Animal Farm Foundation (AFF), owned by Jane Berkey, with its own front organization to run biased studies, the National Canine Research Council (NCRC), which appear in AVMA, which is also pro-pit, produces the biased studies you have seen. Wikipedia is about truth, not biased studies. If the most common dog in rescues was Cane Corsos, and they were advocated as family-friendly dogs, with the horrors that currently result with pit bulls, we would be talking about them instead. Veritas Aeterna (talk) 15:55, 25 April 2023 (UTC)
 * The wiki on pit bull clearly explains that “pit bull” isn’t one particular breed, it is a list of breeds with a distinctive appearance that separate them from other breeds. So the dogs in question appear to be pit bull, they do not need a genetic test as proof, same as if I said a chihuahua attacked because it appeared to be a chihuahua.  The sources claiming the attacks are from pit bulls are from the dog owners, eye witnesses, police, etc.  If you have other sources disputing what those present at the time of the attacks saw, feel free to post those sources. Profpedia (talk) 12:12, 22 May 2023 (UTC)
 * If you have studies or research showing other breeds attack or kill, feel free to create a wiki pertaining to that breed and updating it with reputable sources showing fatal attacks. Profpedia (talk) 12:14, 22 May 2023 (UTC)
 * If you have other sources, please add this content to the article. Rublamb (talk) 03:56, 30 March 2024 (UTC)

Needs sources
I am removing the following section as it consists of items that need a source or better source. As noted by another editor: These cases need more research and better sources before entering the yearly sections. All information provided needs a reliable source.


 * I created that, but i'm not mad. CComp542Veraion19 (talk) 21:38, 10 April 2024 (UTC) — CComp542Veraion19 (talk&#32;• contribs) has made few or no other edits outside this topic.