Wikipedia talk:Arbitration Committee

Evidence age limitations?

Hi! This message is intended for Wikipedia talk:Arbitration/Policy, which redirects here.

I'm floating the idea of whether we could have a cutoff date for age of evidence submitted for cases. I saw in the recent Scottywong and AlisonW cases, some very old evidence was accepted and considered, and while I have no comment on the decision of those cases, the age of some of the evidence felt unfair to me as a bystander. The current "Admins sharing an IP and potentially misusing an off-wiki connexion" case request seems on the verge of being declined, but some very old evidence is being presented there as well.

I think I'd like it if a sentence were added to WP:ARBPOL somewhere to the effect of Evidence older than seven years at the date of submission is not admissible. Seven years is already a long time in the context of a human lifespan, and extra long in internet time or Wikipedia time, which I think runs on cat years. People can change a lot in seven years, and if evidence older than seven years needs to be considered in order to establish a pattern of poor behaviour, I think it is fair to state that no such pattern can reasonably be established.

The exact number is just a starting point for this thought, and I don't remember the exact age of the evidence in the two cases that spurred this message. Maybe some of the older evidence was less than seven years old. I do think it would be beneficial to establish some sort of limitation on age of evidence, and the age can be tweaked / hammered out / bikeshedded if enough other editors think this is a good idea. Or maybe it's a terrible stupid idea, like many that cross my brain from time to time. How do people feel about this? Folly Mox (talk) 17:11, 16 September 2023 (UTC)

I agree, though only to a point. I think that if the evidence shows an ongoing pattern of behaviour, then it should probably be allowed, if only for context.

"ongoing" seeming to be the key word.

That said, there is (and has been) a lot and varied Dispute resolution going on Wikipedia, so I think others might also have insight on this as well. - jc37 18:15, 16 September 2023 (UTC)

• I don't think there can be a hard-and-fast rule. Older evidence is less meaningful in and of itself, and in particular people are unlikely to be directly sanctioned for older misconduct, but old evidence can nonetheless help establish a pattern or lay down vital context. It's not that uncommon for ArbCom cases to involve long-running issues that go back a long, long time (even more than seven years), and in that case arbs will want the details of that history. For example, in the recent World War II and the history of Jews in Poland case, ArbCom cited evidence in its decisions that went back to 2007 - although largely to establish history and context, the fact that there was a 2009 case involving several of the same people in this one (and that there had been regular problems since then) certainly played a role on how hard ArbCom came down on them. They even used the FoF noting that there is no limits to the age of evidence, which implies that they felt it was important to include evidence establishing those parts of the history. --Aquillion (talk) 18:30, 16 September 2023 (UTC)

  I entirely agree with Aquillion.  — SMcCandlish ☏ ¢ 😼  21:04, 16 September 2023 (UTC)

  As do I. Thryduulf (talk) 22:37, 16 September 2023 (UTC)

• Old sins cast long shadows. SN54129 18:45, 16 September 2023 (UTC)
• I'd suggest that it not be a matter of the policy, but rather, it can be (and already is) something that the drafting arbs can specify on a case-by-case basis. --Tryptofish (talk) 21:16, 16 September 2023 (UTC)
• I think there's a flaw in the statement that

  very old evidence was accepted and considered

  . The committee generally lets anyone submit any (on-wiki) evidence they want. This arguably makes the committee's job harder, in that we quite often get duplicated or irelevant submissions, but I think it's the right approach. We do not have formal evidence rules and do not formally "accept" submisssions, they are submissions, that's it. I have, as a filing party, submitted old evidence alongside new evidence precisely to make the case that there was apattern of disruptive behavior. I'd agree that if someone submits evidence that someone did something one or two times over seven years ago that is not part of an ongoing pattern, the committee should and almost certainly would simply ignore it in the final decision. Irelevant evidence of any age gets submitted regularly at nearly every single case. However, ArbCom is not a court of law, and I don't believe a statue of limitations would accomplish much besdies wiki-lawyering. Beeblebrox (talk) 23:20, 16 September 2023 (UTC)

My thinking has been a procedure should be adopted around old evidence similar to how we handle private evidence. So evidence older than X years (I'd go 10) can be considered but only if the committee explicitly gives permission. Barkeep49 (talk) 00:00, 17 September 2023 (UTC)

• Somewhat speaking with my ex-arbitrator hat on, I didn't assign much weight to evidence of misdeeds from long ago compared to more recent evidence. To me, it wasn't that something old was entirely irrelevant, but more so that it was more useful in terms of providing context. I'm not in favour of a black-and-white "older than X years is inadmissible" rule, between such evidence being potentially useful, and that I don't think such evidence forms any significant portion of less-than-useful evidence in general. Maxim (talk) 00:38, 17 September 2023 (UTC)
• Whether or not there's a "pattern of behaviour" is regularly (and rightly) a deciding factor in user conduct cases. It's always struck me as inherently self-contradictory that arbs also often tell people off for mentioning things that happened more than a few of years ago. – Joe (talk) 09:05, 17 September 2023 (UTC)