User talk:Gtoffoletto/WIKILegal

• The combination of no boomerang with an edit restriction on both parties to only take part in this would make for a troll fest. However daft the case, any troll or sockpuppet can hobble a productive wikipedia just by filing a case against them. When people are in a dispute, "no boomerang" give a perverse incentive to avoid trying to be reasonable and instead be the one who filed the ANI case - however egregious your behaviour, if you take the case to AN/I you are off scot free. Improving AN/I is important, but not all changes are improvements. BTW, user RFCs were discontinued not because they were inefficient, but because they were an unreasonable thing to put a fellow volunteer through. Wikipedia is a better place without them. ϢereSpielChequers 17:22, 26 May 2020 (UTC)

  @WereSpielChequers: No boomerang and edit restrictions are some of the more "radical" ideas here and I think they merit discussion. On the whole I think:

  • no boomerang keeps the process more linear. You can "counter sue" if necessary.
  • the edit restriction is placed to put a "cost" on filing the report for both the accuser and the accused and avoid frivolous reports. The "clerks" can dismiss cases if they appear malformed or totally unsubstantiated so it should avoid easy trolling. We may also impose some restrictions on who uses the system (only registered users with certain parameters for example) to avoid abuse.

  I never knew RFC/Us existed. I've reviewed them and they seem a very complex and ineffective system. I think we should use them as an example of what not to repeat. -- {{u|Gtoffoletto}}  ^talk 22:46, 26 May 2020 (UTC)

  Glad you don't want to simply reinstate RFC/U. Keeping the process linear may not be desirable, and while people could "counter sue", the point of a boomerang is that it often isn't the original target countersuing, often it is the people looking at the case. Moving to a more bureaucratic system with an edit limit as the price for filing would dissuade many from filing a case. There is already a cost on filing frivolous reports, a reputation cost for established editors, and for the truly egregious a risk of a boomerang. ϢereSpielChequers 15:24, 27 May 2020 (UTC)

  @WereSpielChequers: We should maybe clear that up: this would not substitute normal admin activity. Any admin could impose sanctions through normal processes. A truly egregious case (no proper documentation for example and unfounded accusations) would be dismissed and could be handled on AN/I by admins. Do you think that makes sense in your experience? The only important point here is that the voting by peers in this process will be on one single issue: the sanction requested for that user and not an open discussion.

  Regarding RFC/Us I never used them in practice. From my understanding this proposal is quite different (e.g. this would be binding). Or do you see it falling into some of the same traps? Thanks -- {{u|Gtoffoletto}}  ^talk 16:52, 27 May 2020 (UTC)

Sometimes is a high standard for marginal decisions

When it comes to a decision making system involving edge cases and contentious decisions, someone who only thinks it sometimes makes the wrong call should be counted as satisfied with the system. When it comes to the much larger number of routine decisions like the more than 99% of blocks and deletions that don't come to AN/I, there I would expect a higher standard. It should be rare for a routine block or speedy deletion to be the wrong call. I'm pretty sure it is very rare for a vandalism block to be unjustified, I have more concerns with parts of the deletion process. But AN/I? if the typical case had 90% agreement with the result, you'd expect pretty much every regular to consider that it sometimes gets it wrong, if only because of the proportion of cases where they were in the dissenting 10% ϢereSpielChequers 20:42, 26 May 2020 (UTC)

We can discuss about individual data points but by looking at the report as a whole I think the picture that emerges is pretty damning. Also you don't need to infer satisfaction. The survey directly asked participants if they were satisfied with AN/I:

• 1 – Dissatisfied (24.63%)
• 2 – (19.40%)
• 3 – (29.10%)
• 4 – (21.64%)
• 5 – Very satisfied (5.22%)

Only 27% of respondents indicated they were satisfied with the way that AN/I cases are handled. That's an average vote of 2,6

-- {{u|Gtoffoletto}}  ^talk 22:55, 26 May 2020 (UTC)

Basic legal rights must be respected

WP:FREESPEECH (a clueful essay) and WP:NOTCOURT (policy) may be relevant: editing is not a right but a privilege and Wikipedia articles are not experiments in democracy. It's closer to a meritocracy and appealing to reliable sources and policies is necessary (see "Serious encyclopedias" at WP:ARBPS and of course: WP:NOT, WP:PSCI, WP:GEVAL, WP:YESPOV, WP:RS, even WP:PARITY, etc). We also have WP:DR and WP:ARBCOM for more structured processes. —PaleoNeonate – 21:24, 26 May 2020 (UTC)

@PaleoNeonate: I think WP:NOTCOURT is crucial. That's why we have judgement by peers here. It will not be judgment by bureaucrats/judges strictly adhering to the rule of law. The "clerks" I propose are just like the "mediators" in WP:DRN. They organise the process but they don't decide the outcome. Also they may not be necessary (though I think they would help a lot).

WP:FREESPEECH is also important. Editing is not a right and may be revoked through this process. But that doesn't mean one should not be treated fairly. Fair treatment should always be a right on Wikipedia and our processes should guarantee that.

WP:DR specifically states that user behaviour disputes should be reported to AN/I which we are trying to improve here. WP:ARBCOM should be the last resort and not the norm. The process would be: Editor Dispute Resolution, Appeal (still through EDR), ARBCOM (final level). -- {{u|Gtoffoletto}}  ^talk 23:08, 26 May 2020 (UTC)

I've added references to WP:FREESPEECH and WP:NOTCOURT in the proposal. Let me know what you think. -- {{u|Gtoffoletto}}  ^talk 23:36, 26 May 2020 (UTC)

There has been a longstanding tension on this site between natural justice and rough justice. Many of the most upset banned editors over the years have been aggrieved on this point, a bouncer dispensing natural justice will bar someone from a pub or nightclub on a lower standard of evidence than a court should use when giving someone a criminal record. There is also the difference between the standard of civil evidence "balance of probability", and that of criminal evidence with its presumption of innocent unless proven guilty. I am pretty sure that the community is divided as to both of these issues. My own preference, and by bizarre coincidence, I think the current aspiration of the WMF, is that at least in some areas we raise our standards to that of the magistrate rather than the bouncer, but we have yet to work out just how much efficiency we would lose by moving to a more legalistic approach, and especially where it comes to returning trolls, racists and other very problematic banned editors. This is a difficult circle to square, but some of the tensions in the community and grievances of the disaffected are symptoms of a certain fudge in our governance as to when and whether we go with standard of justice expected of lawyers and the standard expected of bouncers. For 99% of cases we have this about right. Admins make uncontentious decisions to block or delete and there are appeal processes which a small minority use, often to plead extenuating circumstances and for an extra chance rather than to claim a miscarriage of justice. Of course AN/I is for a contentious subset of the <1% of decisions that are contentious. If we move to a default of "Basic legal rights must be respected" the implications are not trivial. ϢereSpielChequers 16:21, 29 May 2020 (UTC)

@WereSpielChequers: very interesting point. I think we should certainly strive for more than the "bouncer" level. I've spent enough time in clubs to say that bouncers are usually a good example of unfair dispute resolution where manipulation and bias is rampant :-) We can do better. But probably we should not aim to reach a level of a court of justice. It would be excessive and inappropriate given the subject matter and nature of Wikipedia's community. However, this proposal is already a balance between the two. I think in this analogy the proposal would be best described as a "council of club goers". Basically it would be judgement by peers so it would be like asking the other club members to review a video of the fight and letting them decide who to kick out. I think it would result in a more fair process with only a slight loss in efficiency as today we already have a "council of bouncers" and not a single bouncer so discussion and time loss is already abundant. -- {{u|Gtoffoletto}}  ^talk 21:58, 29 May 2020 (UTC)

But are those people appraising the case on the basis of a balance of probabilities, or on the basis of innocent unless proven guilty? To reduce the tension in AN/I and elsewhere in the community one of the things you need to do is get an agreement on the standard of justice that applies and that should be expected. Creating new structures without resolving dissonance is like rearranging the deckchairs on the Titanic. ϢereSpielChequers 06:53, 30 May 2020 (UTC)

@WereSpielChequers: Yes this is something that requires clarity. But do we really need to impose a standard? I think it should be based on the accuser's case. He has the onus to propose the sanction based on evidence(some kind of evidence via DIFF is required always or the case is dismissed) of the accused's behaviour. The editors will then vote based on his accusation, proposed sanction and on the editor's defence.

For example an accuser may point to some evidence and propose a preemptive ban based on probability of future behaviour: "this user cannot be trusted in the future with X given Y". The other editors can then vote if that "preemptive" ban is appropriate or not. In another case the evidence may already be there so "innocent until proven guilty" applies. This is how it works today, with the problem that there is no structure to provide clear voting, defence of the accused (usually if the accused participates in any way it results in self incrimination as it is seen as defensive), focused voting and discussion etc. In short: it all depends on the case brought forth by the accuser. The community will judge every case no matter what standard is applied. Consensus will ensure fairness. Does it make sense to you? -- {{u|Gtoffoletto}}  ^talk 12:02, 30 May 2020 (UTC)

I've tweaked a little the Judgment section. I've added this sentence The key principle during judgement is to Wikipedia:Assume good faith. The accused has the onus of presenting his case and to convince other editors that a sanction is necessary based on evidence (WP:DIFFS). What do you think? -- {{u|Gtoffoletto}}  ^talk 12:29, 30 May 2020 (UTC)

You have exactly two enforceable rights on Wikipedia: the right to fork, and the right to leave. Wikipedia is a private property owned by the Wikimedia Foundation, and the Foundation has plenary authority to ban, which it may delegate. Guy (help!) 16:49, 15 July 2020 (UTC)

Update some terms for clarity based on feedback

I've edited some details to better address the comments received here and elsewhere. See history for a complete diff. I've also included a short "proposal goal" section with a very short summary to make this more easily understandable at a glance. -- {{u|Gtoffoletto}}  ^talk 12:19, 30 May 2020 (UTC)

With the accused and the accuser both temporarily blocked from editing

I don't think that makes sense. If someone must be blocked it's either after the result (in some cases before but only if necessary). A productive editor would not file complaints if it wastes that much of their time. Even ANI that is designed for quick action can have threads lasting a week before closure. Blocking non-offenders would also pollute the block log that normally only includes valid blocks (with some exceptions). —PaleoNeonate – 17:49, 2 June 2020 (UTC)

@PaleoNeonate: I'll repost here part of my reply to Nosebagbear (who made several interesting points) in the Village pump [1]: Your "optional" edit ban makes no sense. this is one of the most "exotic" options. My point is to minimise frivolous case requests (like you state) and to ensure this is a "last resort" for editors. This process should not substitute attempts at resolving conflict through other means. It is for sure a harsh solution. But the bilateral ban is exactly to avoid "first mover advantage". The accused is allowed to "counter-sue" and there is no limit to the number of cases (however they must be complete with evidence and accepted). Do you still feel this introduces more problems than it solves? An option in line with the ones above would be to allow voters to vote for "boomerang" and not just yes/no to avoid the proliferation of counter suing. The defendant may even request it in his defence.

Basically an alternative approach to avoid frivolous reports would be to maintain the "boomerang effect". The defendant may even ask it in his defence. The voters may therefore vote Yes, No, or Boomerang and the sanction would be imposed on the accuser. It may maintain the "linearity" of the process while reducing frivolous reports. I think we may need to remove the Yes/No voting restriction in general (though keep the limit that no reasoning can be added to avoid manipulation and influence). So that both this problem and the problem of selecting the right "sanction" is resolved (see discussion at the Village pump). Users may vote Support/Oppose/Boomerang + supported sanction (if different from the proposed).

Example: A user accuses another of X and proposed a topic ban. Users may think the accused deserves a harsher Indefinite ban. So they will vote: "Support – indefinite ban". If the consensus exists for an indefinite ban then the sanction will be raised to an Indef. If they like the sanction proposed they will vote "Support – proposed sanction". On the contrary the users may vote: "Oppose – boomerang topic ban". And if the consensus was in that direction the accuser would be topic banned. -- {{u|Gtoffoletto}}  ^talk 19:28, 2 June 2020 (UTC)

Gtoffoletto, needless bureaucracy. It's normally pretty clear how long a restricvtion should be, and all it needs is a closure template that includes the expiry and earliest appeal date. Guy (help!) 19:20, 15 July 2020 (UTC)

On boomerangs

The term "boomerang" is one that has grown up over the years and I do not think you understand it. All it means is that anybody bringing a dispute to any forum opens the door to examination of the behaviour of all parties. That applies from ArbCom on down. And it always has. If you ban boomerangs you write into "wikilaw" the ability of anybody - troll, POV-pusher, sockpuppet or genuine complainant - to effectively frame a dispute in terms that paint their innocence as formally beyond question. The only response would be to open another thread about the same dispute. That's not efficient, and it makes no sense at all given that the term "boomerang" evolved long after the consensus view that any party bringing a complaint opens the door to investigation of their own actions. Guy (help!) 16:54, 15 July 2020 (UTC)

@JzG: this has been a common complaint. I think I agree. I'm working on a simpler V2 where the concept of boomerangs will be maintained. -- {{u|Gtoffoletto}}  ^talk 18:22, 15 July 2020 (UTC)

On efficiency

This is a volunteer project. If someone comes along three days into a discussion and finds new evidence that is either damning or exculpatory, it can and should be presented. ANI is not votes for banning, it's a discussion aimed at forming a consensus. We're expected to change our minds in response to new information, and preventing the presentation of new information stops that from happening. Guy (help!) 17:02, 15 July 2020 (UTC)

@JzG: is it efficient for all users to reinvestigate the matter forever until all evidence is found? There must be a limit/compromise at some point. The system I propose includes an information collection phase, where both parties (who would know the dispute better than them?) present the evidence, may defend themselves (this is crucial) and may even nominate witnesses (usually users that know the facts already). Once the info is out it is inefficient for all other editors to continue "wasting time" researching for further evidence. It's time for the voting/decision phase to get the issue out of the way and to serve a just "sentence" swiftly. Every system is a compromise and can't be perfect: so appealing is possible and arbcom comes next if necessary. All we need to do is lower the errors and general perception of unfairness/drama we have with the current system. What do you think? -- {{u|Gtoffoletto}}  ^talk 18:34, 15 July 2020 (UTC)

Gtoffoletto, I think you missed the point. Wikipedia is not a bureaucracy, we're a volunteer project, and if someone turns up at the last minute with something that changes the discussion substantially, erecting process barriers to introduction is a stupid idea.

To be honest I think you'd be nmuch better off addressing the causes of these disputes: a lack of effective content mediation mechanisms. Most of the contentious cases at ANI are people who have been warring over some content issue until either they or everyone else loses their temper. Guy (help!) 19:19, 15 July 2020 (UTC)

@JzG: I think I see what you mean. But what I usually see on AN/I is that after an initial confused phase of evidence gathering the conversation just degenerates into mayhem. It is very rare for "new information" to come up after a while. Especially if both parties have had their say in an orderly fashion. If we introduce some structure to the reporting and allow the accused to defend himself then we should capture most of the relevant information faster and with less disruption. Then people can decide based on the facts. Something was missed? Just appeal. We might want to create a slightly different reporting mask for appeals when one must specify the reason for the appeal (e.g. new evidence emerged).

I totally agree we should ALSO improve the mechanisms to avoid issues in the first place. But unfortunately reaching 0 personal disputes is impossible in such a complex system of people. We must also ensure that when things go wrong we can fix them in a fair way. This is in everybody's interest. Anyone could be the next one to go through AN/I's kangaroo court. I never even knew it existed in 10 years of editing and had my first taste only this year. Not pleasant :-) -- {{u|Gtoffoletto}}  ^talk 09:09, 16 July 2020 (UTC)

Gtoffoletto, you can't fix that other than by having clerks to remove trolls, gadflies and griefers. This is just how it is. Any ANI thread about any contentious topic (which, by now, encompasses pretty much every topic likely to end up there) will have a bunch of partisans seeking administrative actions to enforce their POV and remove their opponents. Guy (help!) 09:30, 16 July 2020 (UTC)

@JzG: Yes this is a major issue. And an existential one for Wikipedia. I think the only solution is to ensure that A LOT of users are on the platform. As many as possible to dilute any attempt at steering it in a particular direction. The more eyes and the bigger the community the less chances that it can be manipulated. This is why I think having a community that guarantees fairness, which is civil and polite, and which is welcoming and inclusive with new users is crucial. This is one piece of that puzzle and the reason I am dedicated to try to find a solution. A lot of good input so far. I think with a couple of iterations we might get to something good. Please join us if you wish! -- {{u|Gtoffoletto}}  ^talk 12:36, 16 July 2020 (UTC)