Wikipedia:Arbitration/Requests/Case/RHaworth/Evidence

The possibility that RHaworth makes decisions by "fuzzy logic"
I was invited to post here by User:CodeLyoko 03:11, 31 December 2019

The block and unblock details:
 * 00:11, 18 December 2019 blocked  with an expiration time of indefinite (account creation blocked) 
 * 11:56, 18 December 2019 unblocked   (see user talk page)

There is some concern that the block was overuse of the log tag " ", that is, it was unjustified over an ordinary block. This concern implies that the policy written at CheckUser should be reviewed.

It is possible that RHaworth considered this block as an ordinary block, no critical private information involved, and acted accordingly. This decision is at odds with the text at CheckUser.

It is also possible that RHaworth did not assess all the details, but instead applied fuzzy logic.

My "evidence" is my statement as follows, as a DRV reviewer:


 * This possibility of RHaworth making decisions by "fuzzy logic" is my impression from reviewing his deletions that are brought to DRV. I have concluded many times that he has misapplied CSD policy to delete a page that needed deletion but via a different CSD criterion or different deletion process.  Also, the use of non-rigorous fuzzy logic is consistent with the speed of his deletions.

--SmokeyJoe (talk) 04:45, 31 December 2019 (UTC)

Velanatti used multiple accounts in violation of policy
Below is a selected timeline of edits by the two sockpuppet accounts:
 * 1) 10:25, 8 Dec 2019: user Ponni Concessao creates Ponni concessao (it has also lived at Ponni Concessao and is now at Draft:Ponni Concessao).
 * 2) 11:35: RHaworth tags the article for A7 (no significance) speedy deletion.
 * 14:36: Bbb23 deletes the page, citing A7 and G11 (unambiguous promotion).
 * 8:07, 10 Dec 2019: user Velanatti begins recreating the deleted article in their sandbox.
 * 1) 10:32, 17 December 2019‎: Velanatti copies their sandbox to Ponni Concessao. It is strongly similar to the version earlier tagged for deletion (comparison).

Bbb23 blocked both accounts and tagged them as "confirmed", meaning in checkuser jargon that the evidence is indisputable that one person is operating both accounts. The user's actions constitute sockpuppetry policy violations, including circumventing policies by using a second account to restore deleted creations of the first; contributing to the same page or discussion with multiple accounts (see the history of Draft:Ponni Concessao, User:Ponni Concessao, User:Velanatti/sandbox); and avoiding scrutiny by creating commercial pages under several accounts and titles.

RHaworth acted on Fram's advice
See User talk:RHaworth/2019 Dec 21, wherein Fram contested RHaworth's deletion of Ponni Concessao and also opined on Bbb23's block, suggesting that Velanatti's use of multiple accounts was not prohibited by policy. Administrators are responsible for their own actions, but the participation in this instance of a well-known and well-respected former administrator should be noted, especially when it comes to making pronouncements about policies.

RHaworth reversed a clearly marked Checkuser block without consulting with Checkusers
See SmokeyJoe's block log evidence above. Note as well:
 * Bbb23 was not pinged to the unblock discussion on Velanatti's talk page (User talk:Velanatti)
 * RHaworth has only one edit on Bbb23's talk page, from 2014
 * Bbb23 posted on RHaworth's talk page inquiring why they were not consulted

Administrators "should not" review or reverse checkuser actions
The checkuser policy and blocking policy both make strong recommendations that administrators not act against clearly-marked checkuser actions without first consulting with the checkuser team. Both policies use the language "administrators should not undo or [loosen/alter]" (emphasis added). That particular language appears based on the Arbitration Committee's 2010 statement on checkuser blocks. This language, along with WP:IAR, leaves open the interpretation that an administrator who strongly believes in good faith that reversing a checkuser action is the correct course of action may do so without violating the relevant policies. This, I believe, is not the intent of the community.

Evidence presented by SandyGeorgia
I saw notice of this case at User talk:Barkeep49 and User talk:ComplexRational.

RHaworth may need a break
RHaworth's "making an opera over an edit" post to ComplexRational, with "Please may I be forgiven?" sounded like a plea from a burnt-out sysop. RHaworth has been an administrator since 2005; he seems to "work" almost 'round the clock, and has never taken an extended break. His mainspace article contributions do not indicate that he edits content for enjoyment, apart from sysop duties. This is not a healthy combination of factors.

Behaviors described in other evidence are not new
Barkeep's analysis of multiple instances in only 17 days of confusing, inaccurate, scolding or snippy responses led me to check my intersecting contribs, where I was reminded of an encounter with RHaworth at the biography for Karel Styblo, an article I found at New Page Patrol about a scientist described as "the father of modern tuberculosis control". I wrote the article in 2012 after seeing behaviors from RHaworth similar to those described by Barkeep49.

There was much back-and-forth with an inexperienced editor, Anderton, who was struggling to understand copyright and sourcing, to write an article on the obviously notable Styblo. Anderton was getting advice from Dennisthe2, who also acknowledged Styblo's apparent notability. Anderton did not seem to be a disruptive editor, rather confused, inexperienced, but trying to do the right thing.

After the initial copyvio deletion, RHaworth's first post to Anderton deleted a small portion of Anderton'a writing ("Hello all"), contained a snippy edit summary ("no copyvios, no CVs") and gave the incomplete and contradictory advice: As to the CV, Wikipedia is not for CVs. Please use the CV to create a proper article in your own words.. In a subsequent response, RHaworth gave very confusing advice: The copyright material simply was not appropriate for Wikipedia. By the time a proper article has been written, it will cease to be a copyvio. (I don't know what that means.)

Anderton then re-wrote the bio from the CV in his own words, just as RHaworth had advised. 

In spite of the obvious notability of the subject, and admittedly Anderton had not supplied a source (easily found), RHaworth next added a prod tag when the article was in this state, on the grounds of no evidence of notability. By that point, I was so put off by the confusing information given to Anderton, that I wrote the bio myself, based on easily found reliable sources. 

RHaworth has given years of service to Wikipedia, but Barkeep's diffs indicate that he seems at this point to be overtaxed to a significant degree, and behaviors that have been in evidence before are surfacing. His service as an admin is appreciated, but it appears that he needs a break and to be able to find a less taxing way to contribute to Wikipedia.

Sandy Georgia (Talk)  18:53, 1 January 2020 (UTC)

RHaworth has applied CSD criteria inappropriately
I am pulling these examples from past discussions at ANI, because the number of RHaworth's deletions is too large to sort through them manually.
 * CSD A10: page had a broader scope than the one it was supposedly a copy of; A10 inapplicable.
 * CSD G4: no previous deletion discussion had taken place; G4 inapplicable.
 * CSD G4: (please ctrl-f "RHaworth") article had been substantively rewritten, G4 inapplicable.
 * CSD G11: ; article was a textbook stub for a politician meeting WP:NPOL but without substantive coverage; not remotely promotional.
 * CSD G11: ; article, as above, was ineligible for G11.
 * CSD G1: ; subject was clearly identifiable.
 * CSD A7: subject met WP:NPOL, and article was therefore clearly ineligible for A7.

RHaworth has been dismissive when queried about deletions
RHaworth has been dismissive and rude when new users, especially IPs, have questioned his deletions (or other actions). There are far too many instances of this for them to be listed exhaustively; here's the first few I found, showing that this problem is common, long-standing, and persistent. Visible in many of these, in addition to the problems related to WP:BITE, WP:ADMINACCT and WP:ADMINCOND, is the basic misunderstanding that en.wp requires English sources for notability and verifiability (giving newbies this impression, when we're trying to write a global encyclopdia, is a serious problem).
 * December 2019:, , ,.
 * May 2019:,.
 * May 2018:, ,.
 * September 2017:.
 * January 2015:.
 * September 2014:.
 * May 2012:.

RHaworth has intentionally misrepresented OS policy
RHaworth has taken exception to how OS works on Wikipedia, and has given misleading advice, and been unnecessarily rude, when dealing with OS-related matters. Evidence includes the diffs below, but also an email thread (linking to OS otrs tickets) which I will forward to ARBCOM.
 * ,

RHaworth has been unable to understand the problems with their actions
Even on the occasions on which he has admitted error, RHaworth has been unable to fully understand what they did wrong. This may be seen at ANI discussions on the subject, , (please search for their username), their statement at the case request page, and at whatever the heck this was.

Evidence presented by Barkeep49
While I might or might not have the time to do a deeper dive into RHaworth's deletions (to try and ascertain just what his "miss" percentage might be) I do want to note one deletion, done while the case request was active, which clearly is against policy. Draft:Bergen Fitzsimmons was deleted as WP:A7. A series speedy deletion criteria cannot be used in draft space. If this had been in article space the A7 would have been, in my view, appropriate. And further it's not clear that this topic could have ever been turned into something useful or would have survived MfD so arguably it could have been a case of IAR. Or it could have been a case of RHaworth not realizing it was a draft (or even, and I think this unlikely but not impossible even given the inciting event, that they didn't know A series criteria couldn't be applied in draft space). I do think the context in which some of these errors have taken place is important. Is there a large number of errors because of the volume of deletion or something else? I hope some other editor who can look at deleted content will make this assessment if I find I don't have the time to do so. Best, Barkeep49 (talk) 18:37, 1 January 2020 (UTC)

Re:DGG
, you probably have already seen it but this was the focus of a large part of what I submitted when it was at the case request. What I didn't include there was just how much the intereaction RHaworth and I had upset me, enough that when this case request came a little while later I was motivated to enough to go through the version of his talk page at that time to see what I found. And what I found was him being rude to any number of established editors who did not phrase their question in a manner that pleased him. This is beyond any snippiness towards those he percieves as having a COI.

If you want more evidence from his previous talk archive I see similar issues with Buffs and Billinghurst. In looking at the two archives before that I see less of this snippiness towards established editors (biggest offender) though also things like, , none of which are damning on their own but all part of a larger pattern. Barkeep49 (talk) 05:00, 13 January 2020 (UTC)

Too quick to pull the trigger for deletions of Drafts that have potential, given a little time/TLC
Key: D - Deletion(s) of draft(s)/original, A - AFD/DRV resulting in keep
 * Evan Nagao D A
 * Shefali Rana Dx3 in rapid succession A
 * Securities and Exchange Commission v. Electric Bond and Share company D A

Biting the noobs/uncivil remarks/Dismissive remarks to legitimate questions

 * Learn to provide a link whenever you talk about a page.
 * Count yourself lucky that I am talking to an IP address who could not provide a link
 * Kindly have the decency to wait until someone who a) has no CoI and b) knows what an article here should look like, thinks your company is notable and writes about it here.
 * I can forgive you forgetting what you had done half an hour earlier...But I cannot forgive your apparent inability to look at your own contributions history.
 * You are hopelessly long-winded...Text emailed. Go to DRV if you are not satisfied
 * When asked why a page was deleted, he responded Please keep your fantasies on your own website
 * Dismissive edit summary of responses & literally deleting requests for assistance: nothing really needing any reply
 * Hostile remarks to Charles Okorobo: "I have moved your latest edit to draft space...turn your rubbisch {sic} into a Wikipedia article." link
 * When asked for restoration of an article, he replied: "You do not deserve it...I find your use of the word "hurt" very curious. Experienced Wikipedia editors do not feel hurt: For those of us who work in the deletions area the only emotion is a resigned acceptance that we are going to see the same inappropriate rubbish for ever." only to backtrack seconds later and make the article a draft.
 * Editing others' user pages because he "didn't like it"
 * A request to send the text of a deletion was met with "How do {you} expect me to find the text if you don't provide a link?".
 * Editing the remarks of others on his talk page
 * In general, a perplexing archiving system which seems to be designed to obfuscate efforts to review entries on his talk page
 * Since this case has started "kindly have the decency to wait until someone with no CoI thinks you are notable and writes about you here". Despite recommendations on how to handle COI, there is no requirement in WP:COI to NOT write about yourself. To the contrary, COI explains how to do so ethically. To unilaterally imply such actions "indecent" is unnecessarily inflammatory. I would contend that, from this response and others, he is demonstrating that he has not learned from previous mistakes and will continue such behavior.

RHaworth's error rate on speedy deletions that require analysis is too high
Some context; no-one expects administrators to be perfect. However; RHaworth makes many deletions a day, but a lot of those are not contentious and very unlikely to be wrong anyway. For example, G8 deletions are usually the talkpage of an article you've just deleted; this is an automatic "next step" when you delete an article which has a talk page. There are a lot of those, obviously, which we can discount. The same goes for G13 abandoned drafts; these are tagged by HasteurBot as soon as they have not been edited for six months, so they are foolproof. And so on.

The problem comes with deletions that are likely to be contentious; of these, A7 and G11 are the most at issue because they actually require the admin to analyse whether they do fall into those categories. And this is where problems arise. The completely ludicrous one that started this whole issue off was this tagging of the article Draft:Ponni_Concessao as an A7 speedy by RHaworth. Now bear in mind this is an admin who does a huge number of speedy deletions, yet clearly, judging by this example, has no idea what A7 actually means (an article about a real person that does not credibly indicate the importance or significance of the subject). There is clear credible importance there, not to mention some decent sources including a full page interview in The Hindu. This was simply ridiculous.

When the case was requested, I decided to look a bit closer at what RHaworth was doing. I looked at two days worth of speedy deletions which he performed while this case was in the request stage (19th-20th December). They included;
 * PKH DAS. Not a G11.  A slightly flowery but pretty-much factual article on a film director who still has articles on two films he directed.
 * Eris Lifesciences wasn't an A7 (it might not pass an AfD or even a PROD, but it wasn't an A7: companies who spend $13m buying a trademark - this was sourced - aren't A7 material)
 * Ultimate Custom Night - Deleted as a G1?? Ludicrous. The article was crap, but it was in English and even had sources. A PROD would have done the trick.
 * Cricingif wasn't an A7 (one of the most notable cricketers ever is promoting a cricket app? That's not your run-of-the-mill bit of software). It's not a G11 either - purely factual. Again - would it pass AfD?  It doesn't matter.
 * Oke Maduewesi wasn't an A7 or a G11. Luke moir wasn't either.  Again, their notability isn't relevant here.  Quite often - especially with articles where there may be many more citations in a language other than English, we end up at PROD or AfD and someone will find extra sources and fix them.   Or even the original author will do so.  When they're speedy-nuked, they don't get the chance.  No-one is saying we shouldn't use A7 where it's correct, but if it isn't, PROD and AfD are there for a reason, and one of those reasons is to not scare away newbies who have made a good-faith attempt at writing an article only to have it binned straight away.
 * Energy Institute - deleted as an A7 on this version (it obviously isn't one) and then taken to DRV here. RHaworth says "I think total absence of references is pretty good grounds for deletion".  Not for an A7 that has made a credible claim of importance, it isn't.

Often, these deletions are made at high speed. I also think there is a tendency to "believe" the CSD tagger without checking for themselves (although there is also the issue that they don't seem to have a great grasp on the rules for A7 and G11 especially.) I am not suggesting any particular remedy here, but something needs to change, for the reasons I mentioned above.

Volume of deletions
RHaworth has deleted over a half million pages, more than any other human by a large margin (WP:ADMINSTATS; rank #1 was a bot running on an admin account). It is my experience that the overwhelming majority of these were speedy deletions; that's hard to analyze definitively, since he tends to use the template-generated default deletion summary (extreme example, as discussed at DRV).

RHaworth is also one of the most frequent undeleters (per WP:ADMINSTATS again), ranking #6 among all accounts. Of particular note is that most of the undeletions by ranks #1, #2, and #4 are from history merges, not reversals of page deletions per se, and that rank #3 is a bot. It is my assumption that the majority of his undeletions stem from self-reversals, since RHaworth is not active either at WP:REFUND nor with history merges as are most of the other most-frequent undeleters.

History at WP:Deletion review
(Methodology: query/41073 is essentially a Special:Whatlinkshere of all DRV daily log pages, filtered to show only pages RHaworth has ever deleted. It contains many, many false positives - about 80% for 2019 DRVs - and so isn't itself useful without human analysis.  I've omitted deletions where RHaworth was not the most recent deleter, except where I saw some mention and substantive review of RHaworth's prior deletion as well.  I've also omitted G4s, G6:consensus-at-TFD's, and G8s unless I saw some mention of the speedy itself during the deletion review, as opposed to a review of the originating discussion/deletion.  Otherwise, I believe this to be comprehensive. )

2019
Worth individual attention: WP:Deletion review/Log/2019 December 26, WP:Deletion review/Log/2019 December 3, WP:Deletion review/Log/2019 August 31, WP:Deletion review/Log/2019 July 9, WP:Deletion review/Log/2019 May 16

Speedy deletion declines
In response to Thryduulf's. query/41353 shows RHaworth's edits in 2019 containing "unspeedy" in the edit summary; there are 144 such. This is obviously not intended to be exhaustive (though he has used that word in every speedy deletion I've seen him decline), and there may be false positives too (such as after an undeletion), but it does show that the blanket assertion that he doesn't check speedies at all is untrue. 144 is appallingly lower than his 52882 deletions in 2019, of course. —Cryptic 16:49, 14 January 2020 (UTC)

RHaworth's A7 speedy deletions are erratic
Many editors have brought evidence of isolated problematic speedy deletions, usually A7 or G11, but the proportion is unclear. As I specialise in processing A7s, I checked 50 consecutive A7 deletions (excluding db-band, which I don't work on) from RHaworth's deletion log, working back from 1 October 2019. I would have deleted around half under A7 myself. Many of the remainder did merit some form of deletion, but were not (unambiguous) A7 material, in my opinion:
 * A few were articles on a topic not eligible for A7: eg YEET theorem (probably a joke), a couple on software (an explicit exclusion in the rubric) eg WebWork time tracker(software)
 * A couple were not in English: কে এম তাসলিমুল হাসান, Mahavir jadhav
 * One was a duplicate article (joke?): Hoff Productions Inc, (Hoff Productions)
 * One was contested by another admin: UNITER (changed to prod by prior to deletion)
 * Several were long-standing articles with many edits: eg CinemaWorld (2012 article with 100 edits on pay tv channel operating in multiple countries), Halid Muslimović (2009 article with 82 edits on Bosnian musician with sales/awards claims and Bosnian sources) and a couple of others with less of a claim
 * A couple have been resurrected successfully: Imogen Waterhouse, Inge Koch (Statistician)
 * The remainder had some form of explicit or implicit claim: eg CP Kukreja Architects (notable buildings and a look at refs/quick Google search suggests definitely notable), Lukman Lukku (appeared in 16 blue-linked films), Blue Revolution (creates broadcasted radio shows), Pop Diaries (moderate Indian Alexa ranking), Makro South Africa (founded 1971, 7k employees & 21 locations), and ~10–15 others for which there is no space.
 * Complete list

RHaworth sometimes salts titles prematurely
Salting is designed for persistent recreations of something that is never going to be notable. In this set of 50, I noticed several cases where a person/company that could be notable if the article were developed or could become notable in future were salted eg Sandeep Vyas. This is off-putting to subsequent good-faith editors working on the topic.

RHaworth makes errors with other speedy criteria requiring admin judgement
Several examples have been noted by others, but in this time period I happened to notice Castleman Disease Collaborative Network, deleted A3 (lacking meaningful, substantive content), despite a long, independently sourced version in the history that had been turned into a redirect.

RHaworth occasionally deletes articles out of process
Also within the same period, I noticed the troubling case of Bell Dance, an article dating from 2013, which RHaworth deleted spontaneously because "The only source there is is blogpost, obviously not a reliable source", despite having previously declined a prod, thus mandating a deletion discussion. In my opinion, unchecked deletion of articles by admins is an emergency resort that should be reserved for the most problematic cases that require immediate deletion, or for the most uncontroversial of housekeeping.

RHaworth's user-talk edit notice is offputting
See MfD. Espresso Addict (talk) 06:45, 9 January 2020 (UTC)

RHaworth is unwilling or unable to change

 * On 24 April 2019, RHaworth said he would change from saying "kindly have the decency to wait until someone with no CoI thinks you are notable and writes about you here" on the advice of . Spin forwards four months and he's still doing it! And he's just done it again!


 * Today I received an off-wiki complaint about Ultimate Custom Night, which was deleted by RHaworth two weeks ago with the rationale WP:G1 : "Incredibly devoid of quality and even the most basic of markup. It's been 18 months, nothing's getting better". The criteria states "In short, if it is understandable, G1 does not apply" (let alone if the article cites PC Gamer and Rock, Paper, Shotgun)


 * RHaworth has been regularly incivil to anonymous editors purely for not having an account, often saying "Count yourself lucky that I am talking to an IP address" (eg:, , , , , , , ).

Ritchie333 (talk) (cont)  12:11, 3 January 2020 (UTC)

RHaworth's conduct can be unhelpful

 * RHaworth's conduct doesn't drop to the level of WP:NPA but his replies can be curt and unhelpful, which WP:BITE recommends against. I was working with, who was distressed to see an article about themselves on Wikipedia that they did not create, and wanted it removed (which WP:BLP / WP:BLPREQUESTDELETE is sympathetic towards). In that context, RHaworth's conduct, adding "I hope the result [of an AfD] will be "speedy keep"" (the actual result was an unambiguous "delete") is particularly poor form. Ritchie333 (talk) (cont)  17:07, 6 January 2020 (UTC)
 * In May 2019; having already created two lengthy threads on RHaworth's conduct, I concluded to that it was not in the project's interest for me to try and seek any further formal action against him, regardless of any evidence that I could or should. RHaworth's response seems rather odd. "I take the view that the bold, revert, discuss cycle is just as applicable to admin actions as to editorial ones." is not backed up with policy; indeed WP:RAAA  (and, to an extent, WP:WHEEL) says the precise opposite: "Administrators may disagree, but administrative actions should not be reversed without good cause, careful thought, and (if likely to be objected to), where the administrator is presently available, a brief discussion with the administrator whose action is challenged." Ritchie333 (talk)  (cont)  18:25, 8 January 2020 (UTC)

RHaworth's talk page archiving is unorthodox

 * RHaworth archives his talk page by moving User talk:RHaworth to a new page, recreating the page, and full protecting the original. This is not disruptive per se but obfuscates the history of the talk page and frustrates editors (eg: ). Unless there is evidence of RHaworth's talk page archives being regularly and persistently vandalised, the latter action seems contrary to the protection policy - "Applying page protection as a preemptive measure is contrary to the open nature of Wikipedia and is generally not allowed if applied for these reasons." Ritchie333 (talk) (cont)  13:24, 10 January 2020 (UTC)

RHaworth has an unacceptably high speedy deletion error rate
See User:Thryduulf/RHCSD /Analysis of speedy deletions by Thryduulf for full details.

I have spent, so far, 90 minutes reviewing the deletions RHaworth did on the 12 of November. In the first 5 minutes of deletion activity that day he deleted 10 pages:
 * 3 were unambiguously correct (but 1 of them should have been referred to Oversight)
 * 4 were unambiguously incorrect (3 of them should be able to be speedily deleted but currently do not meet any crtierion)
 * 2 were deleted under a criterion that did not apply but a different criterion did.
 * 1 was unclear, but was likely either incorrect or deleted under the wrong criterion.

Based on this very small sample, the error rate is >50%. I will, when I get time, continue my review. Thryduulf (talk) 17:40, 5 January 2020 (UTC)

I have just analysed another 10 (details added at User:Thryduulf/RHCSD /Analysis of speedy deletions by Thryduulf), of those:
 * 8 were unambiguously incorrect (7 are examples of content that should be speedily deletable)
 * 1 was unambiguously correct.
 * 1 was unclear, but most likely either incorrect or deleted using the wrong criterion. Thryduulf (talk) 15:17, 6 January 2020 (UTC)

Another 10 analysed although not all are consecutive this time (with details added to User:Thryduulf/RHCSD /Analysis of speedy deletions by Thryduulf), of those:
 * 4 were unambiguously incorrect
 * 4 were correct but with caveats
 * 2 were unambiguously correct. Thryduulf (talk) 16:18, 14 January 2020 (UTC)

RHaworth does not verify speedy deletion nominations are correct
There is an almost perfect correlation between whether the speedy deletion tag applied (by someone else) is correct or incorrect and whether the speedy deletion is correct or incorrect. This, combined with the speed of the deletions, means there is absolutely no evidence that RHaworth is doing any verification that the speedy deletion nomination is correct before pushing the delete button. Administrators are required to use the tools with care and judgement, and this includes verifying that all pages they speedy delete meet the criteria for speedy deletion. Doing otherwise effectively grants permission to speedy delete to non-administrators placing speedy deletion tags.

For comparison, analysing 25 of the first 43 deletions RHaworth did on that day has taken me several hours. All 43 were performed in just under 20 minutes. Thryduulf (talk) 16:18, 14 January 2020 (UTC)
 * In response to Cryptic's comments, OK he does decline some speedies but not that many and there isn't evidence of significant anyalysis:


 * Interval means the time between the end of the previous continuous period of deletions and the start of this one.

This is all just indicative of course. For example it is very likely that at least some of the interval between the second and third periods was spent collating the set of 143 pages to be deleted. Thryduulf (talk) 19:56, 14 January 2020 (UTC)

RHaworth "does not talk to" IP editors
On multiple occasions, RHaworth has refused to answer questions asked by IP editors or has otherwise dismissed their concerns. Examples include: When challenged by Deryck Chan (in 2011) they responded: "As to IP addresses, I consider it perfectly reasonable for me to refuse to talk to them - [...] But if you wish to suggest an alternative wording to "I don't talk to IP addresses" which indicates that I will respond if they repeat the message logged in, then I will use it". Thryduulf (talk) 20:09, 9 January 2020 (UTC)
 * User talk:RHaworth/2019 Mar 30: "I don't talk to IP addresses but fortunately #54129 is willing to do so."
 * User talk:RHaworth/2017 Aug 15: "I have very little time for IP addresses. Feel free to take it to DRV."

Misused and misapplied CSD
I can't begin to contribute the levels of evidence that others have, but for my part, here are three instances I was involved in:
 * G10: User talk:RHaworth/2019 Mar 06
 * G13: User talk:RHaworth/2019 Mar 06
 * R3: User talk:RHaworth/2019 Mar 30

These are just what I have personally engaged him with about on his talkpage; I should note that tracking down specific conversations can be tricky given his a-typical archiving method – RHaworth moves the page history then fully-protects it – and occasional renaming of sections. ~ Amory <small style="color:#555"> (u • t • c) 20:31, 5 January 2020 (UTC)

Number of deletions
This builds off of Cryptic's deletions graph, but I thought it might be helpful to note specific historical numbers with easy-to-follow links. The table below takes its data from the XTools Adminstats tool, with each link to the entire project's totals from 1 January to 31 December of that year (this case was opened 19 December 2019, so the last fortnight of 2019 data might be unreliable). For brevity, only page deletions are included here.

Alternatively, if on-wiki data is preferred:

Community discussions in recent years
I include a selection of links to community discussions where the main topic was the actions of RHaworth. If I have time I will provide a short analysis of each discussion and its outcome.
 * Administrators'_noticeboard/IncidentArchive753, initiated by User:Andy Dingley in May 2012.
 * Administrators'_noticeboard/IncidentArchive831, initiated by User:Magnolia677 in February 2014.
 * Administrators'_noticeboard/IncidentArchive882, initiated by 103.6.156.167 in April 2015.
 * Administrators' noticeboard/IncidentArchive903, initiated by User:Rubbish computer in May 2015.
 * Administrators' noticeboard/IncidentArchive908, initiated by User:Duffbeerforme in December 2015.
 * Administrators' noticeboard/IncidentArchive980, initiated by User:Ritchie333 in April 2018.
 * Administrators%27_noticeboard/IncidentArchive988, initiated by User:Maile66 in July 2018.
 * Administrators' noticeboard/IncidentArchive1008, initiated by User:Ritchie333 in April 2019.
 * Administrators' noticeboard/IncidentArchive1017, initiated by User:Buffs in September 2019.
 * Administrators%27_noticeboard/Archive315, initiated by User:S. M. Nazmus Shakib in November 2019.

My interactions with RHaworth
A few examples of my interactions with RHaworth:
 * User_talk:RHaworth/2016_Apr_18, incorrect application of G7, inappropriate response.
 * User_talk:RHaworth/2016_Nov_27, use of A7 in draft space.
 * User_talk:RHaworth/2016_Jan_13, incorrect use of G6, quote "I think that an unused redirect does not need to be dragged though any special process - just zap it."
 * Talk:Hermie, incorrect use of G13.

RHaworth sometimes deletes pages too quickly to be able to review them properly
In the one minute period from 09:31 on 19 December 2019, RHaworth deleted 49 pages. There are other similar periods of fast rate deletion.

Overturning CheckUser blocks
I want to elaborate on what said about overturning CU blocks. CheckUser states more completely: "[CU] blocks must not be reversed by non-checkusers. Administrators should not undo nor loosen any block that is specifically called a "CheckUser block" without first consulting a CheckUser." When the second sentence is taken in context with the first, I believe the policy is clear that no non-CU administrator may on their own unblock a CU-blocked user. However, I think the policy would be clearer if the second sentence used the word "must" instead of "should".

The companion language in WP:BLOCK always (twice) uses "should" instead of "must". Only the last sentence ("A reversal or alteration of such a block without prior consultation may result in removal of permissions.") implies "must". In my view, we should (sorry) come up with simple language that is identical in both policies.--Bbb23 (talk) 20:11, 7 January 2020 (UTC)

Evidence presented by Pawnkingthree
One incident of which I am aware may have happened a few years ago but I think shows RHaworth's misapplication of the A7 criteria, his condescending manner to fellow editors and his questionable grasp of policy.

RHaworth is too quick to delete
In July 2013 User:FourTildes, a colleague of mine at WP:WikiProject Sumo, created a new mainspace article, now at Musashigawa stable (2013). This is how it looked upon creation. It was nominated for speedy deletion via A7 by User:Ktr101 just seven minutes later, despite making a clear claim of significance - "This is the second stable in sumo history to be headed (and founded) by a foreign-born former sumo wrestler", and containing a reference from the Japan Times asserting that the stable was founded by a Yokozuna, the highest rank in sumo. Four Tildes asserted the significance of the article when he contested the deletion on the talk page, as the instructions on the speedy tag indicate. Despite this, RHaworth deleted the page anyway.

RHaworth is condescending towards fellow editors
FourTildes went to RHaworth's talk page to find out why the article was deleted, and the discussion is at User_talk:RHaworth/2013_Jul_22. RHaworth did restore the page to FourTildes' sandbox at User:FourTildes/sandbox 6 as FT requested, but could not resist petty sniping such as It's very important that you learn the difference between its and it's., and for no good reason collapsed FT's posts as if they were great walls of text. Here we have a productive user with thousands of edits who has never previously had a CSD tag on one of their articles and yet has followed the procedure for contesting the deletion correctly only to be told, Instead of writing a long "do not delete rationale" and the above message, your time would have been better spent adding better references to the article. As FT pointed out, "having only one reference" is most certainly not a speedy deletion criterion. Both the deletion and RHaworth's response to a legitimate enquiry leave a lot to be desired.

(This isn't quite the end of the story, however, as some three years later, presumably still having User:FourTildes/sandbox 6 on his watchlist, he nominated it for deletion when it looked like this, on the bizarre grounds that it was a violation of WP:WEBHOST, despite the fact that it would not take long to work out that the sandbox was being used to make regular updates to a mainspace article, List of active sumo wrestlers. The sandbox was of course, not deleted.)-- P-K3 (talk) 20:20, 8 January 2020 (UTC)

Deletions
My principle defence is that I am human and I make mistakes. I claim that the percentage of my admin actions which are mistakes is very low and acceptably low. Furthermore I always rectify my mistakes promptly or allow others to rectify them and I do so with good grace. Perfection is not required especially on Wikipedia which is a collaborative project - one user's mistakes can be corrected by others - usually calmly and without fuss.

Regarding Maya Rani Paul and Jogesh Chandra Barman, I forgot for a moment that all members of Lokh Sabah are notable and an article about one of them, however short it may be qualifies to be kept. The matter was rectified within half an hour and I accepted the rectification without demur, therefore I wonder why it should be classified as an "incident" and discussed at length?

Regarding the "Ponni Concessao incident, I forgot for a moment that it is set in stone that if a block has been done by someone with checkuser rights, that person must be consulted before the block is removed. The matter was been resolved to the satisfaction of all parties involved. Why did it need further discussion?

I could provide similar explanations for all my other "wrong" actions but it would be tedious and, I think, it would serve little purpose.

I insist that none of my mistakes have done permanent damage to the project or caused other editors to waste large amounts of time correcting them. Whilst I may have lapsed from following the strict letter of the wiki law, I insist that I have always followed the spirit of that law.

It has been pointed out that I do little actual article editing. Is there anything wrong in that? Many editors have very narrow areas of interest: mine is removing inappropriate material. Indeed I would say that my activities give me a much broader view of Wikipedia than someone who concentrates on one topic area.

If I am allowed to retain my admin rights, when doing deletions, I shall concentrate on obviously uncontroversial cases leaving other admins more time to deal with ones needing more detailed investigation. I shall also try to stay strictly within the letter of the wiki law even though it may result in any given "case" requiring more user time to resolve it. — RHaworth (talk · contribs) 16:54, 13 January 2020 (UTC)

Talk page
Regarding my talk page and other interactions with users, I will address one specific matter: the assertion that I intentionally misrepresented OS policy is a baseless libel. I stated here "I sent SPMCC88 here because I did not know where else. (But I have now discovered.)" Which was better: for me to say bluntly to SPMCC88 on my talk page "I don't know" or to give them an hint for further action? They followed my hint, were directed to the right place and got a reply? Why did the matter need further discussion?

Re messages in general, I admit that there is scope for improvement. My family say that I have Asperger syndrome and I firmly believe in the Quaker principle of plain speaking. I offer this as an explanation rather than an excuse.

I have had a metanoia: as I have explained here, I shall try and remember at all times when writing messages to avoid confrontational language because this is a collaborative project. But regarding my "kindly have the decency" mantra: I still think it is right ordering to tell blatant spammers that there is no place for them on Wikipedia. Indeed I consider it a kindness to do so - any further attempts to publish are likely to result in further rejections. But I am still trying to find an acceptable wording to tell them.

Specifically, I promise that I will never insist on links before I talk to someone. Nor will I refuse to talk to IP addresses. (But I will continue to point out the desirability of links and the advantages of creating an account.)

Both
Reader, if there is a specific incident which you really feel needs explanation, I am willing to answer - I presume the talk page of this page will remain open even if the evidence gathering proper has ceased. — RHaworth (talk · contribs) 16:54, 13 January 2020 (UTC)

RHaworth regularly deletes articles under A7 that do not qualify
The following is a short sample of improper A7 deletions within the last six months and by no means conclusive. I have filtered out examples others have already mentioned. Regards So  Why  19:49, 13 January 2020 (UTC)

Bitten with condescending insults
I approached RHaworth in August 2015 requesting help in understanding an old (his) Template deletion that ultimately took nearly two years (a weak delete) - I didn't know how to back-trace when redlinked. I initially used tps that he clearly resented, moving the enquiry to a new section: I apologised twice for any impropriety, explaining it was solely an expedient to get his name on my watchlist as I was short on time. I returned in September 2015 establishing a fresh enquiry, confirming difficulty in understanding the deletion timeline, having found it only through Google, thanking him. He retorted: There are other Bitey retorts at the same archive link: Kindly have the decency to wait..., I should simply intone my “kindly” mantra... and Right from the start you queered your pitch somewhat by showing your inability... with another slightly later complaint of Don't bite the newcomers. Clearly, he has long been allowed to get away with inappropriate, off-handed rebuffs, boasting of it from January 2011 (diff), proclaiming: I have a well-justified reputation for blunt speaking on talk pages. But such pages are not a vicar's tea party.
 * Why on earth did you place this message in the Aindra Dasa discussion - it seems to have no connection whatsoever? Please reply. This in turn means that is not needed - you are just a normal new contributor to this page. Your message is very difficult to understand but without links to the template/s under discussion it becomes meaningless waffle. If you want me to even express an opinion, I need links.— RHaworth
 * You give the impression of someone who wants to find a conspiracy where there is none. Also you are too long winded. Your use of Google is an insult to the MediaWiki software and a clear demonstration of your ignorance of it. Special:WhatLinksHere/Template:Infobox Isle of Man TT course tells you a lot more than your Google search found you. "I need to know if it was ever deleted soon after 28 July 2013 …" What does the deletion log entry say? Why do you not trust it? You speak of "an editor" - why not name him: **** *******. **** is a person who I have met "in the flesh". Like all good Wikipedians, he will be very willing to discuss his actions. So: make a date with ****, fly over to Birmingham, buy him a drink in the The Pickled Piglet and make your complaints direct. — RHaworth

As a non-newbie I was unconcerned/unimpressed, but the gratuitous condescending social commentary needs exposing. De-sysopped he could function with minimal human interaction and less opportunity for self-indulgent acerbic remarks - file mover? Not reviewer. My perception of WP is that when the required niceties of GF are abandoned, the perpetrator can expect reciprocation at similar level; this is about the person behind the edit.--Rocknrollmancer (talk) 22:49, 13 January 2020 (UTC)