Misplaced Pages

This is an old revision of this page, as edited by William Pietri (talk | contribs) at 20:43, 1 February 2010 (→Upraw: Whoa. Did I do that? Too much HTML coding lately.). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Archives

2004 2005 2006 2007 2008

Tigers loose in the museum

Just want to leave a quick note to say how much I appreciated reading the "tiger quote" of yours that's copied here. It's wonderful stuff. --rannṗáirtí anaiṫnid (coṁrá) 02:30, 24 October 2009 (UTC)

Flagged protection

Hi, William. I have little contact with BLPs but I am involved with a number of contentious areas relating to nationalist politics. Flagged protections could change the patterns of edit warring by taking the immediate pressure off to get the "right version" out. Some patterns that can be observed now are:

• IP sock-puppetry, leading to semi-protection
• Edit-warring between established editors (including sock puppetry), leading to full protection

I would be interested to see if flag protection would a) lead to a reduction in IP sock puppetry and b) what effect it would have on the approach of editors to edit warring.

On a), I don't know how it could be measured. If you were to try to measure it, you might measure the number of reverts by registered users against IPs that were not clear cases of vandalism (with the hypothesis being that they would drop). b) is more complex. Right now, locking a page to prevent an edit war simply stops it from being edited. Often this means that even discussion comes to an end because editors just walk away and continue the "fight" elsewhere. Some things you might consider measuring are:

• Do warring editors continue to edit a page after it has been flagged protected (given that their "right version" won't be seen by the majority of readers) or simply walk away and war elsewhere?
• Does flagged protection lead to a healthy editing pattern after protection or do editors simply continue warring? How long do they continue editing for after protections?
• Does flagged protection lead to more or less discussion on pages after protection compared to current protection?
• What frequency of editors (successfully) request that flagged protection be removed compared to normal protection (as this is an indicator that they warring has ended)? How long after the placing of protection compared to the current method? How long does it take to reach a consensus version compared to current system of protection?
• How many edits unrelated to the cause of the page becoming protected take place (i.e. what impact does the current system have for uninvolved editors)?

Some other questions may require focus groups after the period of the trial: Does flagged protection lead to a change in editing pattern altogether - More discussion? Less reverting? Do editor "game the system" either to avoid flagged protection or to cause it?

I hope these suggestions have been useful to you. Best of luck in your effort, --rannṗáirtí anaiṫnid (coṁrá) 23:47, 15 December 2009 (UTC)

Quantifying the effectiveness of flagged revisions

If the following is incoherent, I'm up to late. Ask me to explain. Apologies for the delayed response.

I view this trial as a precursor to emulating dewiki and putting flagged edits on everything. As such, during the trial I believe that articles will be flag protected and stay that way - it won't be (generally) temporary, as it is with page protection. I believe that during the trial only a limited number of pages will be flag protected. These are my underlying assumptions, some of which seem at odds with rannṗáirtí anaiṫnid's.

Regarding rannṗáirtí anaiṫnid's comments:

I do my best to steer clear of contenous stuff these days. Early on, though, I did intentionally stake out a few places where edit warring and other unsociable things were common, trying to "defend" Misplaced Pages from POV pushing.

In sufficiently contentious articles/topics (global warming seems to have had gangs formed with us-vs-them mentalities), you're going to see multiple editors on each side of whatever fence it is that divides them. If one of the warring sides has someone empowered to "mark as patrolled" that will skew the visible version of the article. There ought to be a way of measuring this, although it may require contextual analysis. Perhaps all it requires is a measure of what editors get their edits marked as patrolled by whom within an article (or maybe just across the whole 'pedia, indicating a bias assuming that there's a alignment of interests). Every contributor that got patrolled in one hit would have to be included.

On flagged protection leading to a healthy editing pattern after protection or do editors simply continue warring, I have a guess: because the editors can see the up-to-date version, the edit wars will continue. IPs won't continue, because they can't see their changes, but registered users will.

Many of the things mentioned by rannṗáirtí anaiṫnid for focus groups/editor interviews can actually be measured quantatively. The number of edits against the talk page of the flag-protected articles pre- and post- protection is an indicator of more discussion going on. Edit reverting can also be measured programmatically.

WP:WikiProject Vandalism studies ought to have interesting things to say, but you may need to contact the individual contributors as it seems dead there. I'd be interested in the ratio of IP-good faith edits to IP-vandalism, and the ratio of IP-good faith edits to registered-good faith edits. I'm concerned the IP-good faith edit count will go down. I have a sneaking suspicion that registered-vandalism will go up.

Given so many of these things can me analysed and measured after the fact, it would be helpful if enough data was gathered to make statisically meaningful statements. Vandalism study 1 was interesting, but they discovered after the fact that a sample of 100 articles just wasn't enough to have anything other than a fairly wide range for the vandalism rate they came up with. Assuming a sub-set of pages is marked for flagged protection during the trial, I'd encourage a substantial portion of the articles that would theoretically be begging for it - our most vandalized articles, for example - not receive this protection during the study as a control group. Same applies to BLP articles: explicitly leave some out of the trial. At the same time, include a wide, random selection of articles as flagged protected to measure what the turn-around on edits-to-patrolling is for unattractive ("boring") articles. Get a statistian/trained researcher to advise on this. I'm guessing 2000 general articles need flagging to act as controls, maybe more.

I suspect that vandals that start on a flag protected page will stop around-about there because of the lack of feedback; vandals starting on other pages will continue on to other articles.

Of interest: what happens to the use of Special:Watchlist when this protection is slapped on? My guess is that it may drop a fair bit (this may only happen when every article gets it). Is there an equivalent for unpatrolled changes: articles I care about that haven't been patrolled? Or a bit on the output of watchlist (like minor) showing unpatrolled/patrolled? A drop in watchlist usage could indicate a degree of trust in one's fellow editors.

Did dewiki gather stats, or did any researchers generate stats, on the outcomes resulting from flagged protection? Failing that, was the a broad consensus that flagged protection made the world a better place?

Must sleep now.Josh Parris 13:21, 16 December 2009 (UTC)

Very important point re: if one "side" in contentious issue get "reviewer" status. There would need to be ground rules in who and when can mark a page as "reviewed" in the case of content disputes.

WRT to measuring "many of the things mentioned by rannṗáirtí anaiṫnid for focus groups/editor interviews can actually be measured quantitatively", that is true - but how would you interpret them? The aspects that I mean should be measured qualitatively are issues around the way flagged revisions may be a "game changing" intervention. If quantitative measures of edits go up or of discussion goes down, we can only interpret that this means in terms of the way things are now. RFC/focus groups/qualitative respondents of some kind will be needed to measure the to what degree flagged revisions change the game and and so inform us of what changes in behavior may mean. In good research, qualitative an quantitative go hand-in-hand. --rannṗáirtí anaiṫnid (coṁrá) 19:01, 16 December 2009 (UTC)

Are you Wikimedia staff?

I don't see you listed here: wmf:Staff, which is causing confusion about whether you're actually a member of the Wikimedia Foundation staff (or perhaps just a contractor?). Erik is hell-bent on not listing contractors (why? I have NFI), though apparently now Hampton's listed on the "Staff" page to add prominence to his work. Can you clarify, please? (And if you're feeling feisty, jump into this staff index mess and butt some heads about the current idiocy going on over there?) Thanks! --MZMcBride (talk) 06:52, 22 December 2009 (UTC)

Hi! I'm a contractor. I was also puzzled by the caste divisions between staff and not-staff, but as a long-time consultant, I have a non-interference policy when it comes to a lot of company culture issues like that. :-) But yes, I agree that it's confusing for the general public, and I think your desire to change that is entirely reasonable. Thanks, William Pietri (talk) 22:00, 22 December 2009 (UTC)

Upraw

William

While I applaude the work you do on wikipedia, sometimes it's best to get facts straight. Yes I am high annoyed at your 'Blatent copyright infrigement comment on the newly created 'Upraw' Page created but the hobbit_of_doom.

1) I have not copied and pasted a damn thing from the myspace page.. If and I mean IF there is anything resembling it, it is pure co-inscience. 2) Everything I have written towards the page is my own words. 3) I am the drummer of the band that is putting the page up 4) therefore how can you copyright anything that is yourself?? —Preceding unsigned comment added by Hobbit Of Doom (talk • contribs) 09:03, 23 January 2010 (UTC)

Hi! Welcome to Misplaced Pages. Sorry the interface is so confusing; it has clearly caused some misunderstandings here. In March 2008, somebody created a page for Upraw, but they copied the text from the band's MySpace page. For legal reasons, we can't have any text copied from elsewhere. When editor Lankiveil noticed that, he marked the page for deletion; I checked his work, saw that it was a copyright violation, and deleted it.

Now jump forward to this year. It looks like you created a page on your band. An entirely different administrator, PMDrive1061, deleted the page. This time, though, it's because it didn't meet our criteria for including bands. For reasons I don't entirely agree with, we have very stiff requirements for getting your band into Misplaced Pages. Worse, because a zillion people have already tried to put bands they just started 5 minutes ago into Misplaced Pages, editors are very touchy on this subject.

If you'd like to have an article for your band here, you should make sure you meet those stiff requirements. Further, say that straight away in the article you post. And also include links to independent proof of your claims. E.g., if you've won an award, link to the award site. Or link to press articles that confirm what you say.

I hope that helps! Feel free to drop me a line if you have more questions. Thanks, William Pietri (talk) 17:44, 23 January 2010 (UTC)

Biographies of Living Persons Request for Comments

Hi William, a suggestion involving Flagged Revisions was made at the currently ongoing BLP RfC, section Flagged revisions, comment by Nathan. Could you comment if what is outlined there would be feasible, and if so, what timeframe would be required? Thanks for your time. --JN466 21:38, 27 January 2010 (UTC)

Hi! The current proposal, which we are making good progress on, allows administrators to decide which pages get FlaggedRevs applied; it's basically treated as another kind of protection. Which pages get protected that way during the 60-day trial is up to the community. I'm sure it will get tried out on some BLPs. If you'd like to request features beyond what you'll find in the proposal, I'd suggest putting together a separate page with a clear explanation of what you'd want to add. Estimates take time away from developing, so unless there's a strong community consensus to delay development work in favor of estimating your proposal, we'd probably have to wait until after the trial starts to give you useful numbers. Thanks, William Pietri (talk) 20:26, 28 January 2010 (UTC)

glad to here of the good progress, william :-) (and thanks for your efforts in this regard) - I don't know if it's possible, but could you put just a wee bit more meat on the bone in terms of your progress - are you maybe about half way through some stuff, so we should expect to wait perhaps 3 months or so, or is that way too long, or too short? A gentle update would be hugely appreciated :-) best, Privatemusings (talk) 10:19, 1 February 2010 (UTC)

Hi, PM! Most of the work we know we want has been coded, but we can't see it yet. You can see that on our Tracker page. There are operations issues with getting it deployed, and I currently don't know how long that will take to fix. Once we get it up, people will discover other things they want before public release, and I also can't say how long that will take. So I honestly don't know yet, and don't have enough data that I can make a reasonable prediction. Keep an eye on Tracker, though, and you'll know what I know. Thanks, William Pietri (talk) 20:42, 1 February 2010 (UTC)