User talk:Scott MacDonald/Targeted Flagging

Fails the Ted Kennedy test

Ted Kennedy is hardly one of our lesser notable BLPs surely? Knowing the poster child incidents is Flagged Revs 101 I would have thought. MickMacNee (talk) 14:19, 19 December 2009 (UTC)

Whilst the Ted Kennedy false death thing was embarrassing for wikipedia, in actual fact, it was reverted quickly and anyone finding it would only have had to check CNN and notice there were no reports. I'm sure it is not the first time the Kennedy family have had false stories on the internets, and I doubt the episode caused Kennedy much distress. This type of thing is not the BLP problem of vulnerable articles I want to look at here. Plus, you'll never get consensus to use flagging on such articles.--Scott Mac (Doc) 17:52, 19 December 2009 (UTC)

Ted Kennedy is the reason for the Jimbo quote on your petition. MickMacNee (talk) 11:55, 20 December 2009 (UTC)

When somebody says the right thing for the wrong reason, it's still the right thing (besides, the point of the quote in context is less that Jimbo thinks we need flagged revisions, and more that the timelines in getting them are all out of whack). Steve Smith (talk) 12:56, 20 December 2009 (UTC)

What a load of shit. Don't even start interpreting what other people mean by their own quotes to serve your own purposes. MickMacNee (talk) 15:32, 20 December 2009 (UTC)

And just to add, if people really want to start justifying that petitions existence, then maybe they should start actually posting on that petition. Or not. MickMacNee (talk) 16:14, 20 December 2009 (UTC)

Em, you brought the petition up and the quote up here, no one else.--Scott Mac (Doc) 16:17, 20 December 2009 (UTC)

My mistake. Totally separate proposals. Nothing to see here. MickMacNee (talk) 16:34, 20 December 2009 (UTC)

I like it

As someone who has followed flagged revisions for a while and is more midfield on the BLP issues, I like your suggestions. Especially the point that flagging should be deployed on articles with low edit volume first. This may be hard to get through to people since many are used to protecting high-profile articles. We have to push the point that reviewing is a limited resource, and putting frequently edited articles under flagging is expensive.

In the coming trial, flagging will be available to all admins as a protection method. I would suggest that initially, it should not be systematically applied to low-profile BLPs, but rather encourage admins to use it for BLPs that have been vandalized, but not enough for permanent semi-protection. When it is proven to work efficiently, it can be rolled out gradually to a large portion of BLPs, if reviewers can keep up.

As a trade-off, I think we should sacrifice review quality to some degree. I would rather see some vandalism get through the review because of sloppiness, than a lot of it not getting checked at all. Even a person who is not very well versed in BLP issues can catch a lot of libel. Just having a second person check would help a lot. There are a lot of low-profile BLPs and I don't think a small cadre of BLP experts will be able to keep the backlog short.

In general, we shouldn't expect flagging to perfectly solve the BLP problem. Compare it to semi-protecting all BLPs, which is way to radical to get consensus, and still only stops the newest vandals. Flagged protection on all low-profile BLPs would actually be more powerful, and if the backlog can be kept short then I think the community will accept it. --Apoc2400 (talk) 15:15, 19 December 2009 (UTC)

Personally I am opposed to the "coming trial". It seems to me that a tool that ought to be deployed to prevent serious and credible libels on BLPs will be misused to fight run of the mill vandalism which, while embarrasing to wikipedia, is not really the nub of the BLP problem. Let's rather address the real issue.--Scott Mac (Doc) 17:54, 19 December 2009 (UTC)

Well, a lot of the serious BLP cases are rather simple. Even obvious things like "He is gay and is dating John Smith" can be quite damaging. The really sneaky libels that are designed to appear credible and not get reverted is not very common. Sorry the above comment got long, btw. --Apoc2400 (talk) 20:12, 19 December 2009 (UTC)

Most obvious things are not actually problematic - most simple vandalism harms wikipedia but not the subject. See the first part of User:Doc glasgow/The BLP problem for details.--Scott Mac (Doc) 20:25, 19 December 2009 (UTC)

I like this too. I'm ashamed to admit that the inverse correlation between number of articles subject to flagged revisions and competence of reviewers/quality of review never occurred to me, though it seems obvious now that you point it out. I'm not always good at predicting what will go over well and what won't, but it also seems like this might soften the resistance of all but the most hard-core "anybody can edit" zealots. Building consensus isn't my strong point, but I'd love to do what I can to make this go. Steve Smith (talk) 01:59, 20 December 2009 (UTC)

I'm unconvinced. Define notable in this context if you want to convince me. One of the problems with open editing and BLPs is that those clever enough to not simply write X is a giant jackass get away with it, even on "notable" BLPs. They are not reverted because they conceal their true motives by citing so-called reliable sources, ignoring WP:UNDUE and simular policies and guidelines. Per WP:CONSENSUS these edits must then be discussed to death before they are reverted or the person doing the reverting will be fiercly attacked and eventually blocked for 3RR. There's a common misconception on this site that vandals, POV pushers and defamers are morons that can be dealt with through WP:RBI. They are not. Anything less than a full implementation of flagged revisions will be ineffective. Vyvyan Basterd (talk) 08:04, 20 December 2009 (UTC)

It is not possible to define "notable" in any context, so I won't try. I'm suggesting the objective criterion of "number of watchlists" since it both gives a guide to the popularity of the article - but more importantly gives a guide to whether it is likely that problems will be quickly spotted under the current system. I take your point about the variety of problem editors. However, my point is that a full implementation of flagged revisions, ideal as it might be, will a) never get consensus b) be ineffective - since the sheer number of edits to be reviewed will ensure that we have to let just about anyone who wants to be a reviewer be one - and the reviewers will be under pressure to review and approve at a glance. Thus full flagged will be totally ineffective other than for the most obvious vandalism. And obvious vandalism isn't what damages BLP victims - since if the vandalism is obvious to the reviewer, it will also be obvious to the reader.--Scott Mac (Doc) 13:05, 20 December 2009 (UTC)

Point taken about the required number of reviewers. I stand by my support for a full implementation as I consider the risks involved in continuing to allow free editing of BLPs to be greater than the damage that a shortage of reviewers would cause but I understand that this stance makes any kind of implementation unrealistic in the forseeable future. Fair enough. As far as defining what articles to include I remain unconvinced that we should decide this based on the number of people who watchlist them. At the very least it should be possible for otherwise trusted users to request that a problematic BLP is included if that editor feels that the BLP violations are not being handled. As I mentioned above these situations can be very problematic, even for experienced users. Some kind of "Requests for flagged revisions protection" would be useful along with a liberal policy for when to apply flagged revisions. I suppose I was getting my hopes up a bit too high when I saw all those signatures on the petition but it's time that we do something that isn't just another symbolic effort to resolve the problem. Vyvyan Basterd (talk) 14:51, 20 December 2009 (UTC)

The problem with the petition is that many signing it are supporters of something called "flagged protection" which I've always viewed as useless, as it would only mean that non-autoconfirmed edits are flagged. Any libeller with a three day account is then exempt. I'd be happy to go along with a "requests for flagging" page, but I'd want to avoid using flagging on popular articles - as we'd end up flagging thousands of edits to articles which are already scrutinised and where crap is currently reverted quickly - this would dilute the effectiveness for the targeted articles. The criteria should be that it can be applied to any BLP article which has experienced serious BLP violations which have not otherwise been quickly reverted.--Scott Mac (Doc) 15:36, 20 December 2009 (UTC)

That is is only the weakest kind of flagged protection. There is also a stronger kind where edits by autoconfirmed users are also held for review. I am quite sure both will be in the "real soon now" trial. --Apoc2400 (talk) 02:09, 21 December 2009 (UTC)

Then I would suggest the stronger kind be targeted at low-interest BLPs as I indicate, and not used on higher profile articles.--Scott Mac (Doc) 02:14, 21 December 2009 (UTC)

Neutral

Interesting idea, I'm not sure I buy the idea that such BLPs don't attract new editors. But I would like to add one group of BLPs - I would like us as a courtesy to apply flagged revisions to the BLP of anyone who requests it. ϢereSpielChequers 09:45, 20 December 2009 (UTC)

I'd have no problem with that - we'd only be talking of a few dozen articles at most. Sure, new editors do edit such BLPs - my point is that the vast majority of edits are statistically not to these.--Scott Mac (Doc) 12:59, 20 December 2009 (UTC)

Yes, I also thought that point is wrong. Many new users edit articles like their favorite hockey players, singers and such. It's not a vital point though. --Apoc2400 (talk) 13:31, 20 December 2009 (UTC)

It will still be proportionately lower than people wanting to edit celebrity articles. Plus, even if they do, limited flagging should mean their edits are approved almost instantly.--Scott Mac (Doc) 13:51, 20 December 2009 (UTC)

I think this would be a good way of trialling the system, at least. Depending on the numbers, I'd also consider adding BLPs which are nominated for AfD and pass (especially if no consensus). Rd232 ^talk 21:02, 21 December 2009 (UTC)

I don't hate it

Scott MacDonald, you know by now what my opinion of Flagged Revisions is in general. My opinion is unlikely to change. Having said that, this flagging proposal is far superior to any of the previous proposals - let alone the unthinking support for FR by people who don't seem to want to know any details about what it is they're supporting. In particular, you're right about which articles are the actual serious problem, and restricting flagging to only that narrow class of articles might well make if feasible for reviews to be both timely and complete, which could reduce the main negative consequence of reduced editability. There are still some WP:BEANSY issues still remaining with this proposal, but they are just as present in the other proposals as well. If we do enable Flagged Revisions - I still oppose it - then it ought to be done this way. As to convincing anyone else ofit - well, you have my sympathies. — Gavia immer (talk) 21:45, 21 December 2009 (UTC)