Wikipedia talk:Requests for arbitration/Archive 17

This page is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

A question on recusals

Two current arbitrators were recused from the Giano motion, Dmcdevit and James F. (A third recusal was Mindspillage who has since left the ArbCom). Indeed, James F. was a party to the arbitration and indeed received a caution to "maintain the decorum appropriate for an arbitrator" due to a "you are all idiots" posting. Why are they now voting on this case? Sjakkalle (Check!) 07:14, 2 January 2007 (UTC)

Answer, "because they can". The issue you raise has everything to do with ethics. Many people think that rules of the ethical behavior are rather obvious and should be simply implied in everyone's behavior. Perhaps that's why they where never written into a Wikipedia policy. Indeed, these two arbitrators should not have been involved in this motion in the first place as it is started in connection to the ArbCom case where these to Arbitrators were recused. More, one of the arbitrators was a part (!) of the case on whose motion he now still votes all right. The other Arbitrator found it fit even to start a motion on his own initiative (!) to reopen the case in which he was recused. Seems bizarre to you? So it seems to me and many others.

More strange things happened on the very onset of this motion. The support votes were cast within minutes after the motion was submitted while routinely open and shut motions stay unvoted for days because no one bothers.

The issue of the backroom coordination was already raised in this respect. But there is more to it. Both the original case (called "Giano-case") and this motion were not mostly about Giano but about the culture of backroom activity followed by the implementation of such deals onwiki.

We may discuss whether the arbitrators having a history of involvement exactly in such backstage deals aimed at "ridding" from editors who call for transparency and accountability should be impeached and that would be a separate discussion. But that the recusals in the particular case are surely warranted in the very least. --Irpen 08:31, 2 January 2007 (UTC)

The reason for the apparent block voting is that Dmcdevit mentioned that he had made the motion on the Arbcom IRC channel, no more to it than that. I think it is a good motion, but as you can see from the voting, discussion continues. However Giano is gone. Most action from now on will be with respect to the problems this matter raised. General nastyness on the admin IRC and how much slack to cut especially good contributors when they get nasty, and perhaps are provoked. I say none, but opinions differ. As to the failure to recuse, I'm not sure, but if they recused in the first case, probably should remain recused. Have to look at their explanations. Fred Bauder 13:52, 2 January 2007 (UTC)

I would like to thank Fred for showing a good example of arbitrator's decorum here. There are many reasons why Arbitrator's IRC should be private and I don't ever recall any serios calls to change that. At the same time, the fact that the private discussion indeed took place is publicly acknowledged. This way, both the necessary privacy and the needed degree of transparency are both achieved.

I have to disagree with the remainder of Fred's explanation. The core of the original case was not Giano's incivility but specifically the community reaction to the backroom deals aimed to satisfy the few who self-appointed themselves ar the "bedrock of Wikipedia". The original falldown, to remind you, was caused by the massive community outrage of how Carnildo was promoted to Adminship against the editor's consensus purely due to the backstage maneuvers of some of his supporters among both the bureaucrats and arbitrators. At the time many voiced the outrage and Tony Sidaway, who was at the time the most vigorous enforcer of the backroom decisions, acted ruthlessly to suppress the dissent through outside-of-policy blocks as well as horrific remarks posted to several threads. In the middle of the crisis a little known account, but obviously a sock or reincarnation of an existing user, concocted a haste ArbCom case which was instantly accepted.

This fallout is nothing but the continuation of this old controversy reiterated by a new series of the outside of the policy blocks and extreme arrogance of self-professed "elite" (these events should and likely will[1][2] become a separate ArbCom case). Back to the recusals, the particular arbitrators in question are involved fully in the secretive plottings over IRC aimed at specifically driving out the editors who most vigorously oppose to the opaque and mysterious politicking at Wikipedia. And they have no moral right to arbitrate the case where such politicking is the core controversy.

But there is more to it. This "motion" should be considered either as a reopening of the old case or as a separate case. If this is the former, it is plainly clear that the recused arbitrators cannot just reenter the case on the whim. Besides, in the old case such motion was studied and rejected specifically because it was not warranted by evidence. If this motion is to be considered independently of the old case, it should be presented as such. In this case we need evidence and discussion, not a in-haste move to quickly vote and shut this down. However, as per both the policy and the common sense, the case should be presented to the ArbCom first. This cannot be a legitimate case if it is being started by the Arbcom member on his own and, even worse, the initiator of the case being a judge too. If Dmcdevit wishes to submit the case to ArbCom, he should recuse and desist from all closed channels as far as this case is related. He can be either a party of the case (perhaps initiator too) or the judge but not both. --Irpen 18:54, 2 January 2007 (UTC)

A question on recusals 2

Is there an impeachment process? That would be, um, messy. Carcharoth 12:31, 2 January 2007 (UTC)

We serve at Jimbo's pleasure, so he's the one to talk to. Fred Bauder 13:52, 2 January 2007 (UTC)

I raised this question with both of these arbitrators on their respective talk pages, so you can find there their responses to this issue. Newyorkbrad 12:37, 2 January 2007 (UTC)

They give reasonable explanations for not recusing. Fred Bauder 13:59, 2 January 2007 (UTC)

I disagree, with respect to JDF. Even if his reasons for recusing himself last time do not apply this time, there is another reason which is that he has expressed (on and off wiki) what appears to be a personal bias against Giano. Even if he has no bias against Giano, Giano has grounds to believe that he does. Given that the cost of recursing is only that another arbitrator has to be found, his repeated refusal to recuse in the face of repeated requests from uninvolved parties is puzzling. I'd rather hear from JD before approaching Jimbo, but given that JD seems to have taken a wikibreak, it's an option I'm considering. Regards, Ben Aveling 10:56, 3 January 2007 (UTC)

My recusal is noted here. [3] FloNight 14:40, 2 January 2007 (UTC)

Actually, at first glance that recusal looks a bit strange because of the dates confusion. You might want to explain (for future readers) that the recusal came after the case was accepted and after it closed, and only relates to future motions (because you weren't elected until after the case closed!). I'm sure the ArbCom clerks might have something to say as well about how best to note a recusal to a motion made after a case is closed. Carcharoth 17:46, 2 January 2007 (UTC)

Thanks, I don't think there would have been many eyebrows raised if you had declined to recuse but your forthrightness in mentioning your past issues with this case is welcomed. Regarding Fred Bauder's comments above, I have read the reasons given. Dmcdevit has an adequate reason for unrecusing himself from this case, his reason to recuse is unrelated to this motion. I cannot agree with that James should be unrecused however. He made a posting on the AN which provoked a lot of the reaction and flamethrowing which led to the arbitration case being filed, and he was listed as a party by the person who filed the request. My opinion is that he should not be voting on anything related to this case, period. Sjakkalle (Check!) 15:18, 2 January 2007 (UTC)

Is it possible to come up with a system other than voluntary recusal? I suppose an arbitrator's record of recusal plays a part in their re-election, but that doesn't help if an arb still has a couple of years to go, or if they don't intend to seek re-election. I am force to conclude, like others before me, that it is only considered and forceful discussion from the sidelines, plus maybe internal discussion in the ArbCom, that can keep arbs on the straight and narrow as far as recusal is concerned. At the very least, would ArbCom be able to come up with something more concrete to covere cases where an arb is a named party, and ArbCom agree that the arb should be a named party? That would seem to be a clear case of recusal for the case and future motions. Carcharoth 17:42, 2 January 2007 (UTC)

The question of whether an arbitrator's declining to recuse himself or herself from a given case is reviewable by the full committee was raised in a case a few weeks ago and gave rise to this discussion, although no conclusion was reached. One arbitrator appears to believe that the issue can be raised by motion of a party while at least one other has voted that this procedure is not permissible because whether to recuse rests in each arbitrator's own individual discretion. No comment here on which view is preferable; real-world analogs are discussed in the mainspace article on recusal. Newyorkbrad 17:57, 2 January 2007 (UTC)

I note that – as with all issues on Wikipedia – Jimbo and the Board could pull an Arb from a case by decree. I expect that if the question of an Arb's recusal comes up, it might be discussed privately by the ArbCom and Jimbo; a reluctant Arb could face strong pressure from that quarter to recuse without necessarily requiring a polarizing, public motion. TenOfAllTrades(talk) 18:26, 2 January 2007 (UTC)

I wonder if it might be appropriate to ask the Arbitration Committee to confer amongst themselves and produce a voluntary "standards of conduct" in regards to recusals which they could all agree too. --Barberio 18:54, 2 January 2007 (UTC)

Few months ago (how naive I was then!) I made a what seemed to me obvious and non-controversial suggestion to the policy that would have addressed that matter. Several of then current arbitrators disagreed and Fred, the only arbitrator with real life legal background, saw the benefits of my proposal at the time but said that this is perhaps to be addressed when the new ArbCom convenes. Now, perhaps, it's the right right for that. --Irpen 19:06, 2 January 2007 (UTC)

Lots of instruction creep and written standards of conduct are not necessarily going to help here, because it's impossible to foresee every possible situation that might arise, plus this is informal wiki arbitration and not an actual legal proceeding or anything like that.

But for what it's worth, I happen to have read much of the U.S. literature on recusal standards in the real world, and the best articulation of a standard that I've ever seen is this one sentence in the Model Code of Judicial Conduct: "A judge shall disqualify himself or herself in any proceeding in which the judge's impartiality might reasonably be questioned." Newyorkbrad 22:10, 2 January 2007 (UTC)

I think we all agree with that standard, but the problem in this particular case seems to be a disagreement over what constitutes a "reasonable question" of impartiality. Also, it remains to be seen what happens when an Arbitrator deliberately flaunts his or her authority over recusals and fails to recuse from a case where there is an obvious conflict of interest. Under the current system, it seems that nothing can be done about this. Of course, similar thought-battles have been waged since the ratification of the United States Constitution over the idea of judicial independence. --causa sui ^talk 22:58, 2 January 2007 (UTC)

There's an extra layer to this, which is that Giano was saying, before his blocks, that J.Forrester was out to get him. For that reason alone, it would make sense for J.Forrester to stand down from voting on a sanction for Giano. People like myself, who (apart from believing resignedly that all human groups tend to evolve a degree of speaking behind each other's backs and politicking) haven't the faintest idea whether Giano had a point or not, might at least be able to unfurrow their brow on the objectivity of this voting. qp10qp 23:27, 2 January 2007 (UTC)

Strong reasons for editors to deliberate carefully during ArbCom elections. This time I saw excellent questions, including on this very issue. And I think all the newly-appointed arbitrators answered the recusal questions well.

But the decision to recuse should be the arbitrator's alone, using some sort of guideline (Newyorkbrad's is direct and clear). And if he doesn't? No matter. One voice in 15 will not be decisive. We may be offended, but in the long run the arbitrator is casting one of 15 votes in return for losing community trust. For that matter, arbitrators should be prepared to recuse when there is even the appearance of conflict of interest. No one's individual voice is so important that the project will be lost without it. Jd2718 23:31, 2 January 2007 (UTC)

Newyorkbrad quotation on recusal standard is the best we can do indeed. We can't be any more specific than the rule that recusal is warranted when the "impartiality might reasonably be questioned." Reminding you that the issue here is not Giano per se but serious and justified suspicions of secretive attempts to "run Wikipedia" by expunging the editors who happen to undermine the ability of self-appointed "elite" and the #admin IRC having been widely perceived as the main place where such activity happens, the two arbitrators known for their wide involvement in this channel can be and are reasonably questioned on their impartiallity on that basis alone. That one of them was part of the original case and the other initiated this one should by itself be sufficient grounds to question the ability to be impartial.

On another note, it is true that merely someone's saying that others are "out to get him" would not have been enough. Otherwise, this would have become a bulletproof defense for all sorts of trolls and troublemakers. But it is obvious that we are having a different case here. Not only Giano is not a troll out to damage the Wikipedia (much to the contrary) but that a narrow circle concentrated around #admin is indeed out not just to get Giano but to shut down the community outrage caused by their actions is a wide sentiment here.

Finally, as important as the standards for recusal are the standards of conduct for those who are recused. Jd2718 correctly points above that the arbitrator has only one vote. However, the primary influence the arbitrators have on the case is not through voting but through being able to argue their positions to other arbs in the backroom. It is high time to implement the policy that would require the recused arbitrators to discuss the case (if they have to) only publicly (except very few instances related to sensitive personal information). Real life rules of ethical conduct have similar provisions and I have read the mistrials declared when it was simply revealed that the judge played golf with one of the case parties (or its counselor). No way the recused judge can be allowed in the chambers of the judicial panel that is deliberating the case. Same rules should apply to ArbCom. Having eleven, twelve, thirteen or fourteen active members of an ArbCom provides sufficient guarantee that the case will be discussed by the diverse enough body. There is no need to add any cloud to the arbcom decisions that would be the case if the reasonable rules of both the recusals and the conduct of the recused are under doubts. --Irpen 00:08, 3 January 2007 (UTC)

We are discussing this matter, and hopefully a consensus will emerge that if you recuse, you ought not to be using the arbcom mailing list or IRC to advance your position. However, it remains up in the air. I automatically did this, from the beginning, based on my legal experience. Fred Bauder 00:16, 3 January 2007 (UTC)

Moot

James F. has now recused from hearing the Giano case. Thanks for doing so. :-) Sjakkalle (Check!) 14:19, 3 January 2007 (UTC)

But of course.

James F. (talk) 15:10, 3 January 2007 (UTC)

Case names

Since this was a more general discussion of case naming, I have moved it from Wikipedia:Requests for arbitration/Giano/Workshop. Thatcher131 21:07, 4 January 2007 (UTC)

The arbcom and clerks should avoid naming cases after a single person because that tends to give people the misconception that this single person was the cause and center of all problems. Most arbcom cases don't revolve around one person (indeed, if it were that obvious the person usually gets blocked without need for arbitration) but around a conflict between two or more. For instance, the "Giano" case has in fact nine other involved parties, and remedies against three others. >Radiant< 10:14, 3 January 2007 (UTC)

Hard to find an appropriate name, but I'll try. Not sure what that mess should have been called. Perhaps something about resyopping. Fred Bauder 14:57, 3 January 2007 (UTC)
As long as we are making suggestions about arbcom, I would like to see the new members try out less legalistic paradigms. WAS 4.250 11:01, 3 January 2007 (UTC)
"A rose, by any other name, would still smell as sweet", or perhaps in this case, "a rotting fish, by any other name, would still smell as foul". Yes, naming can be an issue, but really, it's a tangential thought to this whole debate. ArbCom is not on trial here (or anyone else, if we can help it), and while review of their methods is something they would happily undergo, this is not the proper forum. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 15:15, 3 January 2007 (UTC)
I think Radiant makes a good point here, and this seems like as good a place as any to make it. Paul August ☎ 18:30, 3 January 2007 (UTC)
How cases should be named has been discussed several times before on RfAr talk. The consensus was that article names rather than user names should be used whenever possible. In this case the controversy spanned various project pages so that wasn't really an option. An earlier format of adversarial names (X v. Y) has wisely been dropped, meaning that in some cases one or more usernames wind up being the case name. When this case was originally filed at RfAr by User:InkSplotch, it was captioned "Giano et al.", but when the case page was opened, someone dropped the "et al." Newyorkbrad 18:35, 3 January 2007 (UTC)
I think we should try to name the case after the issue. Usually, that's the page. Sometimes, it probably has to be the user, though that should be the last resort. In this case, the issues seem to be 1. The relationship between #admins and wikipedia; and 2. quantifying the extent to which provocation and a good record of contributions excuse incivility. Regards, Ben Aveling 07:00, 4 January 2007 (UTC)

Generally if the case relates to a specific article or issue we use the article or issue name (or we should, barring ocassional lapses). Many cases do focus on a specific editor, however, and unless we adopt some kind of numerical or date-based system, we're pretty much stuck with naming some cases after editors. (It's a heckuva lot easier to remember and find Wikipedia:Requests for arbitration/Zer0faults, which I'm dealing with now elsewhere, than Wikipedia:Requests for arbitration/26 July 2006.) If anyone has concerns about a current case, contact the clerk listed for the case; moving to a new name is trivial if the original name is a problem. The Giano case was exceptional in many ways and a better name could probably have been chosen, but I'm not sure changing it now would be a good idea, I suspect there are a lot of bookmarked links that would get messed up.

I'd say, don't use usernames period (unless the case clearly centres around one distinct user). If there's an article, subject, or issue involved fine - but on the odd case where there isn't, use something neutral: e.g. 'Case of 2nd Jan07' or even 'admin ~~shitstorm~~ dispute of Oct 2006' - it would not be hard to be creative, and if in doubt use the date the case was filed.--Doc^g 21:33, 4 January 2007 (UTC)

Looking back over the last few months, there a few with personal names that probably could have been given issue or article names. Like I said, if you are concerned, contact a clerk. Thatcher131 21:41, 4 January 2007 (UTC)

I'm looking forward to the prospect of Wikipedia:Requests for arbitration/Quantifying the extent to which provocation and a good record of contributions excuse incivility, myself. :-). AnonEMouse ^(squeak) 21:35, 4 January 2007 (UTC)

A bit hard on the fingers, but it sure as hell won't be called /Giano 2, if it ever happens. Thatcher131 21:38, 4 January 2007 (UTC)

Or /Giano II, as it were. --Cyde Weys 23:26, 4 January 2007 (UTC)

How about Snide Wars? Giano 18:36, 5 January 2007 (UTC)

Or Snark Trek? Thatcher131 18:42, 5 January 2007 (UTC)

Ex accuso lucellum?--Mackensen (talk) 18:43, 5 January 2007 (UTC)

I tend to agree with this. The only time Clerks should use individual usernames is when we have a situation like a Husnuck or a Deltabeignet, which were about that one user only. Currently, surveying the list of cases, the only other one to include usernames is Piotrus-Ghirla, which is probably a good naming idea also given what the dispute was about. Naming that one Piotrus or Ghirla would have been bad, and exactly the things clerks should try to avoid. Daniel.Bryant ^{[ T · C ]} 03:36, 5 January 2007 (UTC)

I say we should rename the Giano case to Wikipedia:Requests for arbitration/Giano v. Tony Sidaway v. Kelly Martin v. InkSplotch v. George. By the way, some earlier cases did have the "vs." label, such as Wikipedia:Requests for arbitration/FuelWagon v. Ed Poor. Hbdragon88 04:59, 5 January 2007 (UTC)

If every case should be renamed for an article, perhaps the earlier so-called "Giano" case would become Wikipedia:Requests for arbitration/War of all against all. Newyorkbrad 05:06, 5 January 2007 (UTC)

No, that's a terrible idea for a name of the case. We stopped using "v." for a reason. --Cyde Weys 18:07, 5 January 2007 (UTC)

I don't think Hbdragon88 was making a serious suggestion, any more than I was. Newyorkbrad 18:15, 5 January 2007 (UTC)

Think of it as another barrier against frivolous RFARs. "If you can't type a 100-character URL in properly, your dispute isn't serious enough, go to mediation." (Only mostly humorous.) AnonEMouse ^(squeak) 18:56, 5 January 2007 (UTC)

Collaborations involving page-banned users

There are currently two reports at Arbitration enforcement of users who are banned from certain pages seeking to collaborate with other editors to continue to work on the articles.

Iantresman (talk · contribs) is banned from Plasma cosmology and its talk page for three weeks under the terms of his probation. He posted some comments to User talk:Ionized, which Ionized posted verbatim to Talk:Plasma cosmology [4] [5]. Ian and Ionized argue that even if verbatim reposting is a problem, Ian can provide "information" which Ionized can use to edit the article.
KyndFellow (talk · contribs), who is permanently banned from Sex tourism and related articles and their talk pages, has been advised that he can collaborate with another editor on article changes so long as he does not edit directly himself [6], and he is seeking to do so [7].

Of course, this is only a problem in cases where an editor is banned from article talk pages as well as the articles themselves. Still, some response needs to be devised, as article bans (either direct or imposed per probation) are part of many arbitration cases. I see a couple of equally logical responses.

The policy regarding banned users should be extended to users under temporary or permanent page bans. Proxy editing is prohibited. Editors who collaborate with a banned editor, either by verbatim reposting or by using their suggestions, get the same page bans imposed on them.
Page bans are issued for disruptive behavior, not for providing particular content. The banned editor may offer his suggestions to other editors, who, at their own discretion, may choose to make verbatim edits, partial edits, or ignore the advice. If these third party edits are seen by uninvolved admins as carrying on the disruption, the third party editors may be appropriately blocked or banned. If the third party editors behave responsibly, the edits are treated like any other edit the user may make, even though they were suggested by a banned user.

I think either response can be justified with reference to various policies. What is the arbitration committee's intent here? Thatcher131 04:53, 6 January 2007 (UTC)

A user who does a forbidden act for another user steps into the shoes of the other and is fully responsible. The penalty which would have applied to the other user applies to the user who acts in their stead. Fred Bauder 05:21, 6 January 2007 (UTC)

Should they be warned first? The page-banned user knows they are doing wrong, but the other user might be an innocent party until told about the history of the case. Carcharoth 02:40, 9 January 2007 (UTC)

I'm not an Arbitrator, so this isn't binding by any stretch of the imagination; but WP:AGF would say so. A final warning would be appropriate, with a link to the remedies and this statement by Fred B. Daniel.Bryant ^{[ T · C ]} 05:16, 9 January 2007 (UTC)

Yes, they are responsible for their disruptive acts, but equal amounts of disruption from new users and established users are treated differently with good reason. We should react to the disruption, but not in the same way as with the banned user unless the new user become persistent as well. Dmcdevit·t 07:15, 9 January 2007 (UTC)

I'm not an Arbitrator (and I reported the KyndFellow issue above). Disruptive acts aren't the issue here. If a page-banned user edits their banned page, even in a non-disruptive fashion, they are subject to being blocked. However, a third party may be believe it is okay (helpful even) to edit on behalf of the banned user. There needs to be a routine warning for the third party in this situation, per Carcharoth's suggestion. Only then could the third-party editor reasonably be subject to the banned user's penalties, per Fred Bauder. / edgarde 07:31, 9 January 2007 (UTC)

Fred's statement is either misleading and trivial (of course if someone reproduces the behaviour that lead to the banning of a user, he or she will be in trouble) or, in my opinion, neither fair nor useful. If someone is banned from a store for shoplifting, we don't forbid others to shop for him. Likewise, if someone has been banned from an article, we ban the offending behaviour, not necessarily (all of) the content. Otherwise we would open up a huge can of worms. What is "collaboration"? Is some piece of information tainted forever because a banned user told me about it? Or even just broadcasted it to the world? How do we detect off-wiki collaboration? I would certainly like to hear some other arbitors opinion, and/or a more detailed answer from Fred. --Stephan Schulz 14:18, 9 January 2007 (UTC)

Not at all. In many jurisdictions, an accomplice to a crime is just as culpable as the primary party. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 18:48, 12 January 2007 (UTC)

There are instances where an individuals judgement is suspect, to the extent that their contributions are a net negative, even though some of the contributions would be good. Filtering the contributions through a third party can mean that overall contribution is positive. So I wouldn't like to see the option of using third parties eliminated outright. Regards, Ben Aveling 21:32, 13 January 2007 (UTC)

Certainly some discretion should be applied when judging the edit. Someone should not be condemned simply by association if they have done nothing questionable, but making edits requested by a banned user in that user's stead is something that will be more looked at with more discrimination, at the very least. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 20:33, 14 January 2007 (UTC)

The arbcom has dealt with proxy editing (editing an article in the place of someone who has been banned from it) previously in one of the Lir cases. We prohibited the practice. Raul654 03:49, 16 January 2007 (UTC)

Another so soon?!?

Okay, I once again renew my call for an expeditious review of the acceptance mechanism. Look at this case -- a user has been indefinitely blocked due to sockpuppetry. The admin in question removed the unblock template from his userpage and then protected his userpage (effectively rendering him mute, unable to even request advocacy). He alleges abuse, and it at least seems on the surface that some procedures regarding blocks may not have been followed properly. Please note I'm not addressing the sockpuppetry or administrative abuse issues here... I'm simply saying that this may be another example of detriment caused by the rule change. Four ArbCom members have found cause to hear the case, three have rejected. A net of +1, very unlikely to ever get heard, and yet it apears to four ArbCom members that this merits further investigation (I'm inclined to agree -- protecting user talk pages?)... Sorry to stir the waters more, but this just stuck out at me. /Blaxthos 10:06, 21 January 2007 (UTC)

I haven't looked at the merits of that particular application, but the issue of whether to change the current language on case acceptance really ought to be resolved. The meta-discussion above seems to have petered out, which is quite understandable as it's a dull topic, but still.... One arbitrator having changed his rejection to willingness to accept based on other arbs' interest in the case is a good step toward collegiality in this area, of course.

Having given this more thought, I propose: "A case will be opened expeditiously if at any time it has a total of four net 'accept' votes (i.e., accept minus reject votes) or has accept votes from a majority of the total number of active, non-recused arbitrators. Otherwise, a case will be opened if it has received more accept than reject votes as of 10 days after filing."

Four accept votes period (the original rule) would also work, although it gives a minority of the committee the theoretical right to put a case on the agenda (cf. the real-world analog noted above), but there are arguments to support that. The current "four net" policy I still think is untenable: cases shouldn't be rejected by a vote of 4-3 to accept, much less 8-5 to accept. Newyorkbrad 12:13, 21 January 2007 (UTC)

Newyorkbrad's suggestion makes sense. Adding a minimum number of accept votes (to the third case) does not seem necessary. What is the likelihood that a case will sit at 3-2 for a week and a half?

What rule applies to the current RfAs? Four net? Jd2718 16:32, 21 January 2007 (UTC)

I think it's written down as four net. -Penwhale | Blast the Penwhale 21:41, 21 January 2007 (UTC)

A 4-3 accept case was just rejected. Given Newyorkbrad's workable, sensible solution, this would be the moment to attract some attention for it. Jd2718 23:37, 21 January 2007 (UTC)

I think Newyorkbrad's wording is a little baroque tho. As written, the Gnetwerker case would *still* have been dismissed since it was not a majority vote of *all* active members. It's also a bit of re-inventing the wheel, since this kind of issue has arisen in other similar bodies. Applying a quorum principle to votes is pretty simple to write down, and has a clear effect.

"Requests will be accepted when at least six committee members have voted and there is a majority of votes. Requests not accepted after ten days will be removed." --Barberio 13:01, 22 January 2007 (UTC)

No, under my proposal the case would have been accepted, because the "majority of ALL" is required only before the 10 days are up (to avoid cutting off discussion if the result might change). Still, if my wording is baroquen, it can certainly be improved. Newyorkbrad 13:55, 22 January 2007 (UTC)

NYB's idea is good, rephrased to: IF [(Accept > arbitrator/2) Or (Accept - Reject >= 4) OR (Accept > Reject AND T >=10] THEN Case = OK; else NO CASE. -Penwhale | Blast the Penwhale 15:11, 22 January 2007 (UTC) (Sorry, had to do something like this so I don't go out of my mind.)

Of course, there's nothing to say the voting would have gone the same way on Gnetwerker's case if these rules had been in place; for example, Fred might or might not have changed his vote, and other members might have raised "no" votes if it looked like it might be accepted otherwise. I'd tweak the accept slightly -- I'd say "if accept > (arbitrator/2 - 1)" just because it feels more like the Supreme Court model -- if one vote shy of a majority wants to review a case, that's good enough. However: (accept > reject and T>=10) is no good; a 3-2 that sits for a week shouldn't be accepted. There should be a minimum total vote -- a quorum, I guess -- for simple majority acceptance. --jpgordon^∇∆∇∆ 17:11, 22 January 2007 (UTC)

Number of votes required for case acceptance

At some point last summer, the arbitration policy was changed so that instead of simply 4 "accept" votes being required for ArbCom to open a case, it now requires 4 net accept votes. Questions were raised about the pros and cons of the change, and I believe it was indicated that the matter could be revisited after the first of the year when the new arbitrators were installed. As it would be better to have this discussion in the abstract rather than at a time it's dispositive of a pending application, I raise the matter here. Newyorkbrad 18:16, 8 January 2007 (UTC)

For what it's worth, I went over all the cases from 2006 and posted an analysis here. There seems to have been only one case acccepted under the 4 vote standard that should not have been accepted in all of 2006. In the 2 months since the policy was changed to 4 net votes there may have been one case rejected that arguable should have been heard. Thatcher131 23:36, 8 January 2007 (UTC)

I'll post thoughts and links on this when I get back from a trip (Thurs.) if the committee is interested in pursuing the matter. In brief, "4 accept votes" is problematic because it means "4 accept, 9 reject" gets accepted. "4 net votes" is also problematic because it means "8 accept, 5 reject" gets rejected. Maybe "at least 4 accept votes provided it's a majority of those voting"? Newyorkbrad 23:41, 8 January 2007 (UTC)

Both seem to have their own advantages; I'm not too sure which one (or a combination of some sort) would be the most effective. What do others think? (Yes, of couse we're interested in your comments on this, Newyorkbrad.) Thanks! Flcelloguy (A note?) 02:24, 9 January 2007 (UTC)

Well, I'm not NYB, but here's my thoughts (and excuse the instruction creep):

After seven days, if a case recieves more "accept" than "reject" votes (eg. 4/1, 6/3 etc.), and "accept" votes tally at least four, and there are two more "accept" votes than "reject" (ie. two net "accept" votes), the case is accepted.
If a case hasn't recieved four votes to either "accept" or "reject" after seven days (eg. 3/2, 3/3, 1/3 etc.), the case is given three more days to try and meet either; cases which don't have the +2 net "accept", but have four "accept" votes, are also afforded this extension. Cases which have four votes in "reject" are dismissed after seven days if they fail the above bullet point.
Anything else is rejected, and if the cases which are valid for the extension above still can't meet the "accept" requirements after the extra three days, they are rejected. Daniel.Bryant ^{[ T · C ]} 05:12, 9 January 2007 (UTC)

That really is instruction creep. It is very complicated in contrast to 4 net accept votes in a week. Important matters get accepted without any delay or controversy. Fred Bauder 15:02, 9 January 2007 (UTC)

How about a simple 4 supports and >50% support? Ral315 (talk) 17:45, 9 January 2007 (UTC)

Isn't that what Newyorkbrad said? :-) "Maybe "at least 4 accept votes provided it's a majority of those voting"?" - seriously, this does seem the most sensible proposal. 3/2/1-anything fails (even 3-0). 4-0, 4-1, 4-2, 4-3 all pass, but 4-4 fails. More controversial might be that 5-4, 6-5, 7-6, and 8-7 would all pass this way. Incidentially, arbitrators do still vote and take part in cases that they voted to reject, don't they? Or do they concentrate on cases they voted to accept? Or do they tke part in all cases while they are 'active'? Carcharoth 18:04, 9 January 2007 (UTC)

Presently there are between 9-14 arbitrators participating on each case, with most cases closing after 6-7 arbitrators have voted. Thatcher131 19:11, 9 January 2007 (UTC)

7-6 should not pass; that is the flaw with the proposed system. Daniel.Bryant ^{[ T · C ]} 01:18, 10 January 2007 (UTC)

How about a little common-sense application of WP:IAR? If a case is on the border between being accepted and not and there are issues that need to be examined, or a majority of arbitrators do want to see it, or what not, why not just use discretion and use the case anyway? I don't think it necessarily makes sense in this instance to try to figure out the best hard-and-fast rule. --Cyde Weys 18:12, 9 January 2007 (UTC)

Well, from the clerks' point of view I'm not going to IAR and open a case that doesn't meet the criteria set forth by ~~my Insect overlords~~ the arbitration committee, although the arbitrators are free to do so themselves. In my opinion, the old system only failed to reject one case that should have been rejected, out of more than 100 in 2006. The new system rejected the case against Cbuhl79 at 4-2, but since Cbuhl79's last edit was 11 days before the case was canned, that was also argulably the right decision. The proposed hybrid solution (4 votes and over 50%) is functionally the same as 4 total votes since there was no case in 2006 where the hybrid solution would have resulted in a different decision from 4 total votes. Thatcher131 19:11, 9 January 2007 (UTC)

Thatcher131, let's not Monday-morning quarterback the Cbuhl79 case -- at the time the issue was decided (by the new rules) there was no way to determine if Cbuhl79 would resume editing or not, and thus if it was the right decision or not. The precedent set by the circumstances at the time of the decision arguably shows the fatal flaw of the rule change. That the troll simply went away is secondary, and doesn't negate the problem that was revealed. :-) /Blaxthos 23:48, 9 January 2007 (UTC)

I have been reading along on this issue for the past few rounds (approximately 6-8 weeks I guess) and have refrained from commenting as I thought it was primarily an issue for the arbitrators to discuss, but I see others are jumping in, so I will too. Although my Wikipedia experience is not as deep as some, I do have some experience with decision-making in small groups (of the in-person variety.) I would like to suggest a simplified version of Daniel.Bryant's proposal, specifically that a case be accepted only if it has at least four "accepts" AND the "accepts" outnumber the "rejects" by at least two votes, within seven days (if that is the correct number of days.) (Or in the parlance used on here, 2 net accepts and 4 gross (if you will) accepts.) This is not "instruction creep" and it is not complicated. (All I have really done to Daniel.Bryant's proposal is to remove the automatic extension of time at the end of the seven days; if the arbitrators want that, they should have it, but I got the sense that it was this aspect that drew the "instruction creep" comment. The numbers of votes are exactly as Daniel.Bryant proposed, he just stated it in a more complicated way by including a "majority" vote, which does not need to be stated because of the "two net accepts" requirement, which is actually a sort of supermajority.) To make this even simpler to implement, someone could make up a little table showing how many votes are required per number of voters. (I figured out the numbers, but I do not know how to make tables.) This method would ensure that at least four arbitrators have participated in the decision, and that there is more than a bare majority in favor of accepting it, while avoiding the rejection of cases in situations where, for example, 6 accept and 3 reject. I think the 6-3 vote (leading to rejection) is a good example of something you want to avoid. A member of the community asks the ArbComm to hear a case that he/she thinks has considerable importance; a majority of the committee (9 of 15) give their opinion about whether to accept; two-thirds of those voting (6 of 9) vote to accept; but the case is rejected because it has 3 net accepts instead of 4. I think a member of the community would be justifiably irritated by such a result. So, there's my suggestion. 6SJ7 21:58, 9 January 2007 (UTC)

I guess I like 6SJ7's suggestion as well. "2 net, 4 gross" really isn't instruction-creepish at all, and can be stated very succintly and understood just as easily. --Cyde Weys 22:05, 9 January 2007 (UTC)

I do have a nack of overcomplicating things :) The main concern I had originally was cases at 7-6, 6-5 etc. should not pass. Even 7-5 is debatable, as is 6-4, but still there is a concensus among ArbCom to accept, as opposed to one Arbitraror casting a "swing vote". Daniel.Bryant ^{[ T · C ]} 01:21, 10 January 2007 (UTC)

I apologize for the fact that this discussion has become a bit more complicated than I expected. Subject to correction by someone who was here at the time (such as User:Uninvited Company), I think the original rationale for 4 votes was to get cases open quickly when it was clear arbitrators wanted to hear them, without waiting further, and that is still a laudible objective if things start off 4-0; not sure if the arbs would want a case opened at 4-3 or would prefer to wait for more votes at that time. To Thatcher131's point, I think the distinction is more likely to matter with a greater number of arbitrators. Hopefully the rule agreed upon can be encapsulated in a way that tables should not be necessary! Newyorkbrad 22:12, 9 January 2007 (UTC)

I would suppose it comes down to a balancing act. The criteria for accepting a case should be sufficiently strict that the ArbCom isn't bogged down by frivolous cases. At the same time, it should not be so difficult to grant cert that worthwhile cases are being cast by the wayside. From the analysis above, it appears that a change in the rules would have made a difference in very few cases last year (though I'm not sure about the effect of the new, enlarged ArbCom.) I am inclined to argue that the ArbCom ought to hew towards the less-restrictive standard, simply because it avoids an appearance of unfairness (a 6-3 support-oppose tally resulting in rejection?) while having a very minimal effect on the actual workload.

For reference, the United States Supreme Court (USSC) employs the so-called rule of four, which stipulates that the Supreme Court will hear a case given the support of four of the nine justices. In essence, if at least a substantial minority thinks a case has some meat to it, the Court is compelled to hear the issue. TenOfAllTrades(talk) 22:21, 9 January 2007 (UTC)

Purpose?

Amid the discussion of all the possible permutations of options for ArbCom acceptance mechanisms, it occured to me that one particular facet hasn't been discussed much... at the risk of sounding dense, what was the intended purpose of the change? Was it to alleviate type I and/or type II errors? Why did it occur, and did it have the intended result? Might be useful in trying to find a better solution. /Blaxthos 23:54, 9 January 2007 (UTC)

Before the reject votes really did not matter expect they were not an accept vote. The same as not voting at all. This made them matter more. FloNight 00:26, 10 January 2007 (UTC)

I'm a little leery of the 'matter more' argument. I mean, I assume that Arbs aren't sitting around alone at night, drinking rye straight from the bottle, quietly weeping to themselves because their votes don't matter enough. (At least, I hope that they aren't....) I am concerned that the large 'net' margin makes reject votes matter more—more in fact than an accept vote, in that substantially more accept votes are required than reject votes to determine whether or not a case is accepted. If we're up to a fifteen-member ArbCom, a case can be rejected even if as many as nine of the fifteen Arbs think it has merit. TenOfAllTrades(talk) 03:02, 10 January 2007 (UTC)

My preferred beverage is Merlot. And since I've been an arbitrator for less that a month, I still manage to use a wine glass most of the time. ;-) Your question has some merit but it is important to remember that this matters in very few cases. Many cases are rejected as being premature with the hope that some type of resolution will occur without ArbCom involvement. If the situation is not resolved then the situation can be revisited. FloNight 19:47, 17 January 2007 (UTC)

Exactly my point. In essence, one reject vote vetos four members who believe the case warrants acceptance -- the balance of power, so to speak, has been completely tilted (not to mention if one factors in the fact that the full ArbCom rarely participates). Far too much power in one vote, if you ask me... I like the model based on the SCOTUS procedure, personally. However, this still begets the question... what was the purpose of the change? /Blaxthos 03:08, 10 January 2007 (UTC)

And this becomes relevant...

The University of Phoenix case is at 3/3/0/0. It is unlikely to achieve four net votes, but possible to gain the +2 net proposed. Thoughts? Daniel.Bryant ^{[ T · C ]} 04:07, 14 January 2007 (UTC)

The purpose of the change (which happened in October, incidentally) is kinda foggy to me too; it came along with some other streamlining moves. Not sure it was discussed enough. --jpgordon^∇∆∇∆ 08:55, 14 January 2007 (UTC)

I believe it was definitely not discussed enough, and as time progresses we see more cases that (in my humble opinion) are being rejected while a majority of ArbCom members feel that it should be accepted. I personally prefer the acceptance mechanism that is modelled after the Supreme Court. In any case, I believe it's time to do something -- widen the scope of discusssion? RfC? Who does these things? Who gets to decide these issues? If ArbCom are the ones to decide the rules by which they accept cases, it would be wholly ironic if we send it to WP:RfArb. /Blaxthos 20:49, 17 January 2007 (UTC)

Well, we haven't actually seen any case in which a majority voted for acceptance and yet the case was rejected, have we? The case that was a 3-3 case turned into a 3-6 case when I asked for more votes on it (that wasn't what I'd hoped for, but that's the problems with individual opinions -- they're often different from mine.) If it ever does seem to make a difference on a case, we can review the rules for acceptance there and then; since we (as in ArbCom) made the procedures, we get to change 'em too, and on the fly if we want to. (For example, if we find ourselves with a case that's 6-5, my guess is that we'd be able to work out between us that a case with almost half the board wanting to hear it is worth hearing.) --jpgordon^∇∆∇∆ 21:33, 17 January 2007 (UTC)

Not to be a thorn, but I do think Wikipedia has a duty to inform its userbase of what the operating procedures will be. One of the foundations of our democratic society (both by Constitutional law and Supreme Court review) states that there will be no de facto laws... I think that's a pretty good premise for us to emulate here as well. I'm not trying to ignore the practical wisdom of your proposal, however I think that it is only forestalling more confusion and increasing the likelihood of additional type I/II errors, as well as supressing users' liklihood to attempt at even filing ArbCom cases. Just pointing out the yang. /Blaxthos

Be careful what you wish for. It looks like there will be a clear example in the Gnetwerker case where a significant majority of Arbcom members voted for hearing a case, but the 'four net votes in ten days' rule leads to it being dismissed.

This really isn't that new a problem, and other bodies that may sit with variable numbers of their members set a Quorum. I think the best, and simplest to word solution would be to say "Requests will be accepted where at least six current arbitration committees members have voted and a simple majority voted to accept. Requests that have not been accepted within ten days will be removed." Assuming a quorum of six would be acceptable to the arbcom for voting to take cases. --Barberio 09:17, 21 January 2007 (UTC)

Works for me, but I'd love to hear from the ArbCom on why the standard was changed in the first place—what problem was recognized and solved by this solution? TenOfAllTrades(talk) 15:02, 23 January 2007 (UTC)

AMA advocates' status in cases

Surely some here know me for being an AMA advocate that from time to time appears in the halls of ArbCom defending people. This time, I have a doubt. What status have formal or informal advocate during a case? Are we "parties" or just "others"? If we are "parties", then, can we make motions, endorse them, object, request in the workshop or just comment as an uninvolved user? My opinion is that advocates should be considered a party, as we're involved (indirectly, yes) in the case... but, in the other hand, no arbitrator has ever thought on ruling on an advocate... It's quite confusing to me and that's why I request this clarification. Thanks in advance! --Neigel von Teighen 19:12, 18 January 2007 (UTC)

- Any user may be an informal advocate, but an AMA advocate speaks for the user they represent. They are not a party but may speak for the party they represent. In the past no advocate has effectively represented a user, but the role is open. Great care should be taken to make only motions which make sense to the arbitrators. Focus on adequately presenting relevant evidence in a useable form and on framing proposals in terms of core Wikipedia policies. Fred Bauder 02:01, 23 January 2007 (UTC)
Comment: A rule that a party's advocate in a mediation automatically becomes a party to an ensuing arbitration case might inadvertently discourage editors from taking on the role of advocate. Hopefully, it is a rare situation in which an advocate's own conduct becomes the focus of inquiry by ArbCom, so I don't think formal "party" status is necessary. A sensible rule would be that advocates have the same standing as any other editor to present evidence, make workshop proposals, etc., but that of course when an advocate is commenting in the capacity of advocate, it's good practice to note that fact. When an ArbCom case is filed, providing courtesy notification to anyone who was acting as an advocate is also an appropriate thing to do. Newyorkbrad 19:20, 18 January 2007 (UTC)
Comment • As an advocate myself, I would say that we are just another editor, and should be treated as such. There should be no preferential or special treatment given, and their status as a party should be judged on the merits, or lack thereof, of their actions, and the length of their involvement. If they are not directly involved in the dispute, other than by acting as an advocate, than I would be compelled to think that they would not be a party. After all, we do not bring the previous mediator on a case into a case simply because they were the mediator in the prior attempt at dispute resolution. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 19:27, 18 January 2007 (UTC)

I'd suggest doing something like putting the comment in the party section and then signing it, say, "NvT as advocate for RealParty". Unless acting directly as advocate -- i.e., speaking for them -- then you're just another editor with a hopefully useful comment. I think ArbCom can figure out the difference between the real parties and the advocates and is unlikely to include the advocates in any remedies... --jpgordon^∇∆∇∆ 19:33, 18 January 2007 (UTC)

- Well, that's what I myself do: add "User: Imaglang (aka Neigel von Teighen) AMA advocate for User" in the party list and then adding a diff to anything that certifies me as advocate. What I expect from ArbCom is a little guideline on what to do, not a policy. Something we can rely on when an advocate (formal or informal) has doubts on what to do. That's it what we need. --Neigel von Teighen 09:04, 19 January 2007 (UTC)

Comment • I would assume an Advocate is not a party, but an advocate for a party. In a given case were an Advocate represents a party, and performs actions as any other editor, it may raise COI issues. nobs 22:17, 19 January 2007 (UTC)

Comment Neigel asks "can we make motions, endorse them, object, request in the workshop[?]". It seems to me the answer is "yes, of course; anyone can do all that stuff, party to the case or not". As far as I can see, absolutely nothing hinges on whether advocates are considered parties. What am I missing here? PurplePlatypus 05:32, 20 January 2007 (UTC)

- Nobs has hit the point that led me to make this request. There can be COI problems like this: User X makes a motion and Advocate endorses it, counting as two "moves" for the same party in a same "turn"... (proposing-endorsing) I don't know if I'm clear enough... It turns me to be rather unfair in some way... although anyone could go and request an advocate too. Simply put, what I request is a little official guideline written by ArbCom so no doubt nor conflict arrive... Maybe am I being too silly? If so, tell me and withdraw this. --Neigel von Teighen 17:41, 20 January 2007 (UTC)

Advocates have no formal status during arbitration (or, stated another way - they are the same as everybody else). In the past, they have shown themselves clearly and conclusively to be impediments to the arbitration process. In cannot think of a single case they have helped in any way. In short, the AMA is useless. Raul654 18:04, 21 January 2007 (UTC)

Well, I really know we're an impediment, but we try to do the best we can, including myself. And have an idea: please send me a feedback on my work on the ongoing Starwood case after its closure and tell me how I did it and what shall I improve or if I was really useless? Honestly, it can be a good start! --Neigel von Teighen 19:01, 21 January 2007 (UTC)
- I can't speak to the Starwood case (which I haven't yet looked at), but in all past cases, the AMA advocates' arguments have amounted to nothing but pettifoggery. If you wish for things to go different, then - and I say this admittedly without looking at what you have been doing there - I strongly suggest you advocate for the person are representing, and avoid resorting to the AMA's standard toolbox of dilatory tactics. Raul654 21:38, 23 January 2007 (UTC)
Personally, I feel that Advocates could be of use, but currently and in the past they have not been. The problem is that when someone makes an argument on one person's behalf and it is struck down, they tend to take it as a slight against them. I feel that it is important that AMA advocates hold themselves to a certain decorum when working in a case. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 18:48, 22 January 2007 (UTC)
- This is why we thought in our last AMA meeting to do gather arbitrators with our Coordinator and Deputies to talk about these things... Well, in summary, the answer to my request is: "Advocates are the same as anyother editor in the case". Have I undertood it well? If so, then, we can say this request is closed, wouldn't be? --Neigel von Teighen 11:17, 23 January 2007 (UTC)

I honestly must say that I find myself taken aback by comments such as "In short, the AMA is useless," but I cannot deny that historically such observations have been true. In the past, members of the AMA were causing havoc by bringing cases that were far too young in the WP:DR process to MedCom and ArbCom. This, in turn, was mostly due to two things: 1) Advocates who did not have enough direction or practical experience and 2) the fact that the AMA was practically inactive and running "on its own" without any sort of supervision or direction. People were "signing up" with no communication between members and no idea of what to do, the request system was horrific, and the previous Coordinator had resigned months earlier with no acknowledgment from the Association. (This is the state I found it in when I joined).

Recently, with many Advocate efforts, there has been a resurgence in membership, a reorganization of our structure and and influx of zeal to help and because of that the AMA is back on its feet. We've kept the same goal that we initially held (helping disputes on Wikipedia) yet have a very different way of going about things. As a result we have already relieved ArbCom of dozens of cases and saved many hours of precious time by reducing the escalation of conflicts as they arise and are referred to us.

Things are working well, but they are far from perfect yet, and I feel that the next logical step is for the AMA to foster a closer, functional, and working relationship with ArbCom in order for our processes to be more efficient, and in the end, put less strain on WP:DR. If we end up doing "our job" properly, even fewer cases will rise to the level of Arbitration, and those that do should be properly researched, formed and submitted. What my wishes are in discussing this would be to see that there is some cooperation between us to further these goals and make Wikipedia a better place.

-- (AMA Coordinator) אמר Steve Caruso _(desk/AMA) 03:39, 25 January 2007 (UTC)

I had suggested that we open a formal hearing on this but there wasn't much interest from the other members of the committee. I'll throw out a few comments informally here since I've seen the AMA in action before and have a few specific concerns and believe I can see both sides:

The AMA was organized by individuals who were not especially supportive of the arbitration process. A clean break or a repudiation of this viewpoint might be appropriate.
Arbitration Committee members love to see clearly and concisely presented cases. If that's what you do, great, you'll find that you have our full support in about a picosecond. On the other hand, if you expand cases unnecessarily with trivial counterclaims, you'll be walking in the footsteps of your predecessors.
If you're going to do this, part of your job is to control your clients. They shouldn't be participating themselves in cases except to offer testimony.
If a request for arbitration includes a statement from someone stating that they wish to be represented by a member advocate and that they are going to refrain from direct participation in the case themselves, I would respect that and would expect that most other committee members would as well, as a practical matter.
I would be open to a more formal or official role for advocates once advocates have established a track record as an asset to arbitration itself in addition to helping out by steering cases to the most appropriate forum.

The Uninvited Co., Inc. 10:52, 28 January 2007 (UTC)

I am very encouraged by Steve Caruso's statement above. The function of AMA, done well, can provide a very valuable service to the community and the ArbCom. I also support UninvitedCompany's reccomendations above. Paul August ☎ 19:30, 28 January 2007 (UTC)

Number of votes required for case acceptance

At some point last summer, the arbitration policy was changed so that instead of simply 4 "accept" votes being required for ArbCom to open a case, it now requires 4 net accept votes. Questions were raised about the pros and cons of the change, and I believe it was indicated that the matter could be revisited after the first of the year when the new arbitrators were installed. As it would be better to have this discussion in the abstract rather than at a time it's dispositive of a pending application, I raise the matter here. Newyorkbrad 18:16, 8 January 2007 (UTC)

For what it's worth, I went over all the cases from 2006 and posted an analysis here. There seems to have been only one case acccepted under the 4 vote standard that should not have been accepted in all of 2006. In the 2 months since the policy was changed to 4 net votes there may have been one case rejected that arguable should have been heard. Thatcher131 23:36, 8 January 2007 (UTC)

I'll post thoughts and links on this when I get back from a trip (Thurs.) if the committee is interested in pursuing the matter. In brief, "4 accept votes" is problematic because it means "4 accept, 9 reject" gets accepted. "4 net votes" is also problematic because it means "8 accept, 5 reject" gets rejected. Maybe "at least 4 accept votes provided it's a majority of those voting"? Newyorkbrad 23:41, 8 January 2007 (UTC)

Both seem to have their own advantages; I'm not too sure which one (or a combination of some sort) would be the most effective. What do others think? (Yes, of couse we're interested in your comments on this, Newyorkbrad.) Thanks! Flcelloguy (A note?) 02:24, 9 January 2007 (UTC)

Well, I'm not NYB, but here's my thoughts (and excuse the instruction creep):

After seven days, if a case recieves more "accept" than "reject" votes (eg. 4/1, 6/3 etc.), and "accept" votes tally at least four, and there are two more "accept" votes than "reject" (ie. two net "accept" votes), the case is accepted.
If a case hasn't recieved four votes to either "accept" or "reject" after seven days (eg. 3/2, 3/3, 1/3 etc.), the case is given three more days to try and meet either; cases which don't have the +2 net "accept", but have four "accept" votes, are also afforded this extension. Cases which have four votes in "reject" are dismissed after seven days if they fail the above bullet point.
Anything else is rejected, and if the cases which are valid for the extension above still can't meet the "accept" requirements after the extra three days, they are rejected. Daniel.Bryant ^{[ T · C ]} 05:12, 9 January 2007 (UTC)