One of the major recurring themes throughout all of this, has been non-patrollers dictating how patrollers should behave. There are enough exceptions to that over-generalization to belabor the point, but it remains a genuine concern of mine.
I think most sysops are willing to step up on the rare occasion where it seems prudent (and the controversy surrounding it is loud enough.) But there must be a better way. Perhaps if every active sysop were somehow required (suggested?) to, for one month a year, patrol at least 3,000 entries in one month.
One "patrol" would be just that - marking an edit as patrolled. With the WT:WL and maybe 50 manual patrols per day (~1/2 hr) that could be accomplished with - well, with about 1/2 hour a day, 5 days a week, for a month. It would very obviously be that individual sysop's discretion, whether to block or warn, in each specific case. We could go with six sysops per month, "required" to patrol, rotating them out at the end of each month. (So, once every 13 or 14 months, you might be expected to step up for a refresher, if still active.) The progression could be in date-sysopped order, alphabetic order, or random order. It could pick on three newbie sysops paired with three old-timer sysops, etc.
Obviously, there would be no penalty for not meeting the "quota." But it would be easier to give someone who rarely patrols, a badge of merit, i.e. This Sysop Isn't Clueless. It would completely eliminate the basic inequality between those sysops who do, and those who don't.
I feel very strongly that those who never patrol, cannot even see the repetitious patterns of abuse. Watching is no comparison to doing. Subtle patterns that are difficult to quantify become obvious in a month's time; moreso over a greater period of time (I think Ruakh likes to call that "paranoia" on my part.)
Mostly when I talk about your paranoia, I mean when you decide that long-standing contributors are conspiring against you and this project; but in the context of patrolling, I mean when you decide that a common type of error is in fact a common type of vandalism or trolling. "Subtle pattern" does not generally mean "pattern of evil". —RuakhTALK12:59, 12 February 2008 (UTC)[reply]
Please keep your insults to yourself. You seem to have a pathological distrust of anything I say, no matter how many times I've dug through the archives to prove to you I'm not making stuff up (a truly worthless task, as you then say things like "oh, well I guess consensus has changed.") There's a productive thread here - some little good thing out of all of this mess you helped create (instead of dying down quickly) - you don't need to disrupt it with slander. --Connel MacKenzie16:21, 12 February 2008 (UTC)[reply]
WTF? I distrust you because I know from experience that you feel no compunctions about inventing facts (case in point: right here). —RuakhTALK00:03, 13 February 2008 (UTC)[reply]
Such a policy or recommended procedure would very intentionally have no teeth: if you don't patrol, you should expect your comments about patrolling to be dismissed without consideration. And you wouldn't get a cute little star that says you have performed your annual required patrolling (instead of a barnstar, perhaps a bloody dagger icon?)
Having a regularly-scheduled "changing of the guard" would inherently keep "over-patrollers" from patrolling past the burn-out point.
Such a system might help keep new sysop nominations flowing...an old timer sees that his rotation is coming up, so the month before nominates six new sysops that each will want their "I qualify as a real sysop" badge of honor, perhaps bumping him/her from the queue.
Judging from the variety of questions on IRC about patrolling lately, I can't help but observe that every sysop who patrols, implicitly learns a tremendous amount about Wiktionary policy subtleties that they simply never ran across previously. The little bit of distributed experience creates an enormous pool of knowledge about how things have been handled in the past (i.e. what consensus turned out to be.) That would tremendously help expedite WT:ID questions, WT:RFV/WT:RFD challenges and recurring WT:BP topics.
About "quotas:" it it were obvious that the sysop were at least "trying" to patrol for several days a week for four weeks, I think it should count, whether it works out to be 3,000 edits (approximately one day, these days) or 90,000 or only 1,000, total for the month. The numeric quota would remain a "suggestion" (adjusted for inflation) verified by a few other patrollers with standing, I guess. --Connel MacKenzie07:21, 12 February 2008 (UTC)[reply]
This is an excellent idea, sign me up. (I patrol occasionally, but never systematically enough to make much of a difference.) Question: is there a patrol log (so that it is possible to see who patrolled an edit and when)? -- Visviva07:39, 12 February 2008 (UTC)[reply]
That's another reason for people (especially me) to be put off by adminship. How quick is it to patrol edits? Like, how many can you patrol in a minute? --Keene09:57, 12 February 2008 (UTC)[reply]
With Lupin's Navigation popups and "advanced patrolling" turned on in WT:PREFS, anywhere from a couple dozen per minute, to one per minute. (With Whitelisting turned on, 400 in a minute or two is not uncommon.) In general, the edits you patrol are the "obviously OK" ones, or ones that you've listed on RFD/RFV already. --Connel MacKenzie10:16, 12 February 2008 (UTC)[reply]
That sounds a lot more serious than what I've put in. I've only ever patrolled edits to see what it was like, or because I was bored. I have a tendency to do a bit of editing when I look at recent changes though, so I don't think I could average more than one a minute. As with any other measure, we are reminded that it's just a number. DAVilla14:55, 12 February 2008 (UTC)[reply]
DAVilla, you won't earn your bloody dagger (OK, maybe we should pick a more reasonable icon/logo) without actually doing a bunch of batches of serious patrolling. Yes, there are tricky ones that require investigation, but if you average only 30 per half hour, you aren't patrolling. That isn't even one drop in the bucket. --16:38, 12 February 2008 (UTC)
Well, the political mechanism of convincing generally non-patrolling sysops to at least try it for a while is certainly something to be worked out. I tossed out the words "somehow required" there for lack of a better way to express it. --Connel MacKenzie16:30, 12 February 2008 (UTC)[reply]
I think that's not such a bad idea, but I would want the periods (maybe weeks instead of months) to be voluntarily selected, and I would want to have more carrot than stick. No one should feel like they're required to participate, although they'd be highly encouraged. Little stars (no bloody daggers, please) could be referred to when conversations such as this arise. Also there could be several quotas, and bronze-silver-gold or similar system. If you're close to your quota, just continue into the next period to cap it off. DAVilla14:42, 12 February 2008 (UTC)[reply]
A week is far too short; one month would be a bare minimum. As I said, peculiar patterns emerge - some of which are weekly cycles. Yes, more carrot than stick - but how? --Connel MacKenzie16:30, 12 February 2008 (UTC)[reply]
Simple. Put the logo on each admin's user page, above the Babel. In my mind it would be a collection of metalic or colored stars rewarded over the past year, with a maximum. For instance, if the maximum were five, I may have one very old gold star, two silver stars, and three bronze stars, but only two bronze stars showing. When the gold star becomes a year old, it is moved out, showing two silver and three bronze. If I then earn another bronze star, it would back up the others, but being over the maximum it wouldn't show. I need to aim for silver or gold to better myslef. Eventually the silver stars will age as well, and after a year of no patrolling, all would be lost.
How about two weeks at a time? with double and possibly tripple shifts allowed, but a mandatory break after that, "mandatory" in the sense that additional patrolled edits would not count towards any reward. Come to think of it, the maxing out of stars already reinforces the idea of taking breaks. DAVilla18:40, 12 February 2008 (UTC)[reply]
A month seems like a really short amount of time for this, particularly when you've got cooperation from five (or more) other people. That's why I left the estimate of 3,000 - an intentionally too low number. I suppose if someone patrolled 30,000 in their first week, I'd be happy to say they were off the hook. (But I'd also have to check what it was they were marking as patrolled - were they doing all the warn/rollback/edit/delete/blocks that go along with it, or just marking stuff incorrectly. No high-five, if they do more harm than good!)
If you saw something - some behavior - twice...you'd think it was a coincidence (if you even noticed it.) If you see it three times, you could still dismiss it out of hand. A month is barely enough time to get a taste of what I'm talking about. In a month's time, you won't gain any continuity, but you will see a few things over and over again - even if your 1/2 hour a day is at all different times of the day. --Connel MacKenzie18:44, 12 February 2008 (UTC)[reply]
There is a compromise between the two, but it's even more complicated than what I proposed, so I won't try to hash out the details unless you want to hear it. The point would be to try to make several of the periods, or subperiods if you will, consecutive, accomplishing the above, while the subperiods themselves could be short. The reason I think that short timeframes are better is that people cannot generally commit to a full month, and certainly not on calendar month boundaries. If I'm extremely busy most of the year, but I can contribute heavily from the middle of December to the middle of January, then there isn't any reason why that should be looked at more negatively than contribution from the start to the end of February.
Possibly more important than the length of each period is the ratio of maximum stars to periods per timeframe of expiry. In my example there is a one year expiry so 26 two-week periods in that timeframe, and a ratio of 5:26 or about 20% which seems too low. DAVilla19:12, 12 February 2008 (UTC)[reply]
Excellent idea. Sign me up for the crash course in blunder avoidance. There must be some minimal level of "documentation" that doesn't disclose anything of great value to evil-minded lurkers. DCDuringTALK15:25, 12 February 2008 (UTC)[reply]
The Help:Patrolled edits page could certainly use some expansion. The Rat Patrol announcement here has some other handy comments. The combination of tools makes a handy technique: enabling popups in WT:PREFS and "patrolling enhancements" (+"expert" after a 1/2 hr, or sooner if you get sick of the alerts) speeds up patrolling enormously, by sifting out most of the good edits in advance. As some have pointed out here and there, starting from the bottom and working your way up tends to be more interesting. --Connel MacKenzie16:21, 12 February 2008 (UTC)[reply]
I have added "How to patrol" based on what I think, and a response I gave on someones talk page. There is more that could be said but I am not sure how much is better learned on the job. Please feel free to update it all ye who patrol.Conrad.Irwin19:22, 12 February 2008 (UTC)[reply]
I am strongly against the idea of letting sysops compete with each other (and not only because I don't like loosing :), we should be working together - not racing to patrol as much as possible - that is when mistakes happen. We need to be able to patrol in a relaxed state, with the knowledge that if we miss something someone else will pick it up. Conrad.Irwin19:22, 12 February 2008 (UTC)[reply]
I agree that a relaxed state would be important, but it might still be possible to encourage people to patroll. (But the opinions of those who don't do a certain task shouldn't be dismissed, that would create a caste system.) To make it more relaxed, what if it was, instead of a certain number of patrolled edits, a certain time of spending with patrolling? This might be something that those who are not administrators could do too, I guess they couldn't mark pages as patrolled (?) but they could still watch recent changes and revert vandalism. But hopefully it's not going to become a requirement that everyone has to do same things. Could the stars work so that after the year they would go to the subpage, so they wouldn't disappear? And I'd like to suggest that one could not earn more than maybe two stars during a year so that people wouldn't worn themselves out. The idea would be to get as many as possible to do it, so that no one would have to do it too much. Best regards Rhanyeia♥♫20:33, 12 February 2008 (UTC)[reply]
The way I see it, it wouldn't be a certain number of patrols in a strict time period like "one calendar month", but a much looser one like "three to five weeks", so that we wouldn't feel the rush to complete them. There would also be an informal sign-up sheet to avoid having too many patrollers at certain times and too few at others. The competition is not in who can catch the edits first, but in who can dedicate their time, in chunks, to patrolling. I suppose it's possible that some people could take it too far, and focus more on marking the patrols than on fixing the content and educating the editors, but it's the same problem with edit counts, and I don't think anyone would be so fooled into scrutinizing them that way. DAVilla21:56, 12 February 2008 (UTC)[reply]
haha, little stars for RC patrolling? Amazing! I thought people stopped appreciating little stars when they were about 13. At least, my kids do. Anyway, I wish the star system lots of luck. I'd be happy to dispense stars. --Keene00:21, 13 February 2008 (UTC)[reply]
If we are going to be rewarding sysops with little stars then we are going to have to reward other contributors with little stars, otherwise it will be unfair, where will it end? Wiktionary is small enough that those who work well can be seen to be working well by others. It should not be necessary to bribe people in order to get them to do things that need doing. Those who don't want to do them should not be marked out as less "good" than others - and these stars are for marking out the good people. Please can we ignore this terrible idea and get back to doing something more productive, like actually patrolling ;). Conrad.Irwin16:00, 13 February 2008 (UTC)[reply]
Concur. If I'm going to get stars on my user page for patrolling I will certainly stop immediately. The sysops we have don't need cutesy things to do the work. (What is needed is just getting enough people doing it regularly that it doesn't burn out one person after another!) DAVilla: rather than inventing fancy stars and boxes that we will not use, it would be far more useful to spend your time actually patrolling. Robert Ullmann16:33, 13 February 2008 (UTC)[reply]
I very much doubt if anyone would come along and put stars on your user page without your consent. :-) Beyond that, each user -- each person -- has a different motivational profile; some people like to have visible markers of the work they have done (something which is woefully lacking for tasks such as patrolling, which generate no obvious product). Personally I would probably keep such records in numerical form on a subpage, but I see nothing wrong with having graphics on one's userpage either. It's certainly a step above barnstars and Babel templates (which don't really reflect anything meaningful at all). If visible tokens of work done help motivate some users to contribute more actively, surely that is a good thing for the project. -- Visviva16:41, 13 February 2008 (UTC)[reply]
I'd like to spend more time patrolling, but I find it incredibly boring, and hence I have to motivate myself first. I'd like to keep the cutesy little box even if I'm the only one who uses it. No, no one is going to put the box on your user page. Otherwise I will beat Connel with a carrotstick. ;-) DAVilla19:43, 13 February 2008 (UTC)[reply]
Ok, so if this is something that admins can put on their own pages to keep themselves going I can sort of accept that, but please no-one try and force one on me :). I am still of the opinion that this is a bad thing though, people will come across them and start thinking, hey - I want a box like that... and the use of them may well spread until they loose their meaning - unless of course we don't allow mere users to get them, which reeks of elitism. (Though this is a very weak slippery slope argument). I also feel that if admins grudgingly do their spate at patrolling every week, just so they can keep their shiny stars the quality of patrolling is not going to be high (though don't get me wrong more users doing it reasonably well is better than not doing it at all). Conrad.Irwin20:25, 13 February 2008 (UTC)[reply]
Mere users can't patrol edits. That's already elitist. But we should welcome other types of rewards as well, and we should be open to receiving new admins, which I think we already are.
Only five stars over the last year are displayed in the box, each taking about a month to earn. So you don't have to grudgingly do your spate every week. I imagine the lowest level will in the low hundreds, so it won't be difficult to keep the box full, either. The point is first of all to encourage patrolling by recognizing the effort, and secondly to encourage doing it cyclically. DAVilla21:44, 13 February 2008 (UTC)[reply]
I think this latest variant is a very poor idea. Giving ratings so that non-patroller's comments can be quickly dismissed was the original idea (sortof - but not so harsh.) Giving yourself a ranking is...erm, not gonna work. P.S. The stars make me vomit. The swords don't convey the right idea, either. --Connel MacKenzie20:31, 13 February 2008 (UTC)[reply]
I'm not giving myself a ranking, I'm giving you the opportunity to please rate me, objectively of course. Your bot would be the one to fill in the reward levels, not me, and maintain them on a weekly basis. Granted Conrad and Robert don't want to give you that opportunity, but I already do, and I would imagine DCDuring will as well. I'm thinking that some might not want it on their user page, but wouldn't mind seeing the compiled results on a separate page. The latter is more what you had in mind, no?
You can put up whatever symbol you like on your own page. If you want to use the same dagger but just make it bloodier and bloodier then fine—though it's going to reflect on your personality, so really I should have taken the comment as tongue and cheek, no? I was just showing off the flexibility. I have deleted the swords, and you probably won't find any bloody daggers on Commons anyway. I'm thinking sheilds or nets might be a good idea. DAVilla02:29, 14 February 2008 (UTC)[reply]
I've compiled the results over the past week (Feb. 6 thru Feb. 13) and removed the names for objectivity. These are the counts of patrolled edits per individual. They exclude automatic patrols, which I think are edits from accounts with admin or bot standing. I don't know if these include whitelisted patrols, but those should definitely be excluded as well. The point is to reward volunteered time.
2883
1863
1294
1124
1064
760
727
439
271
171
157
139
108
76
44
44
25
19
15
9
8
5
3
3
1
1
Clearly there are a number of individuals, including myself, who mark edits as patrolled when we see them, but do not actively hunt them down. The point of this is to encourage us to participate more actively, but not to set an impossible goal, so I'm thinking that over the course of four weeks 300 and 1000 patrols would be good for bronze and silver.
On the upper end, the point isn't to make current patrollers work harder, rather to recognize them for the effort that's already put in. I'm thinking that 3000 would be a reasonable bar for gold. I don't see any reason to push the upper end to exhaustion. DAVilla02:54, 14 February 2008 (UTC)[reply]
You have to be very careful with lists like that, a lot of the time I have rat patrol open in the background which automatically hits patrol for a lot of whitelisted non-sysop users. Also, when a sysop edits a page the software, I am not sure if it is MW or rat patrol, marks all previous edits of a page - yes these are probably roughly fair but care should be taken to ensure that people aren't unduly mislead. Also there was mention of a bot to do this, is there one that will already - or for the first while will you just go by whatever process you used to generate the above list? Conrad.Irwin15:33, 14 February 2008 (UTC)[reply]
Yeah, I mentioned that above, but it really deserves more attention. Rat patrol and the whitelist are effective and should be praised, but patrol counts from that are not fair if some users run rat patrol and others don't, or if access times are biased due to location or provider. When no one is running it we'd end up checking edits from whitelisted users, which is a good idea from time to time anyways. That's only temporary because rat patrol will pick up old edits as soon as it's online again.
It's not like the software has eyes as good as ours; it just assumes all of those edits are good. If this is intended to reward time dedicated, then including those patrols would be the wrong direction to go. I have no idea how it works, so I wouldn't know how to remove them from the count if subtracting "automatic" patrols hasn't done that already. I would guess rat patrol has to keep its own tally. DAVilla21:01, 14 February 2008 (UTC)[reply]
Note: The primary reason Rat Patrol was written, we to mark edits from anons as "patrolled" after SemperBlotto cleans their entries up. All the rest is fluff. --Connel MacKenzie01:55, 15 February 2008 (UTC)[reply]
(Sorry for moving your comment up.) So in most cases credit should be given to SemperBlotto, correct? Is there any way to count or approximate that, or at least to avoid giving Conrad Irwin credit for something that he never looked at? DAVilla02:41, 15 February 2008 (UTC)[reply]
Rat patrol does two things, marks whitelisted non-sysop users, and pre-sysop non-whitelisted edits, so while Semper probably gets more credits than everyone else I don't know whether he is in the majority or everyone else combined beats him. The user who visits the patrol page gets credit for the patrol, so unless everyone was running Rat Patrol and it was trained to behave correctly, I can't see a nice way to do this. Unless of course we get some javascript to load up the page history after a sysop edit and get it to patrol everything there, but that again relies on every sysop using it - as rat patrol would have to be stopped or modified not to do this. Conrad.Irwin19:29, 15 February 2008 (UTC)[reply]
If we're starting to worry more about who gets "credit" for patrols than about whether every contribution is being adequately reviewed, we're going in the wrong direction. -- Visviva12:00, 16 February 2008 (UTC)[reply]
In the course of discussing who gets credit for patrols, we've uncovered the problem of marking edits as patrolled that should not all be, just because a sysop has modified the page. In many cases they should be, e.g. for new and very short pages, but in some cases they should not. DAVilla19:17, 17 February 2008 (UTC)[reply]
This is such a long discussion that I don't know where to put my comments, so I'm just returning the carriage and feeding the paper a bit. First of all, I don't like the idea of star-based rating, only because I see what it's turned into on enWP. Moreover, there are a number of tasks to do besides patrolling edits, so that even if we do barnstar patrolling, we should bear in mind that the barnstar means only "this admin is an active patroller" and not "this admin is an active admin". As such, such barnstars should only have influence toward how we rate that admin's opinions regarding patrolling-related issues, nothing else.—msh210℠21:58, 13 February 2008 (UTC)[reply]
I don't see any reason why this approach couldn't be extended to any sort of drudge-like work. As you say, patrolling is not the only non-photogenic but critically important task here. Some sort of minimal recognition for a thousand entries created, or a hundred debates closed, or what have you, would certainly be nice. And I think the idea of taking responsibility for a quantifiable block of edits makes more sense than the open-ended (but inherently unenforceable) commitment implicit in Wiktionary:Administrators/Dishwashing. -- Visviva02:43, 14 February 2008 (UTC)[reply]
I would very much like to know about the rating system on Wikipedia to make sure that this system isn't designed in such a way to encourage negative behavior. DAVilla02:59, 14 February 2008 (UTC)[reply]
Any way of competing is bad in my opinion - even their freely given barnstar system, or the edit counters etc.etc. Though I am happy if you want to experiment with this on a very small scale I feel that there should be a community decision (on the Beer Parlour) if we are to expand this. Conrad.Irwin15:33, 14 February 2008 (UTC)[reply]
I have noticed that many individual edits are marked us unpatrolled, even though there have been subsequent corrective edits by admins. This clutters up the list or provides a way to game the system for patrol points, barnstars, and 5% pay raises. It would help to have these items cleared by admins as they do the work.
Having worked through the list for the patrol-education value, I now find mostly translations or entries in languages that I can't help with, at least at the bottom of the list. It would be handy to not see entries for which one had no competence to add value, if that could be readily done without complications or heavy resource requirements. OTOH, the diminishing backlog is itself pretty good already. DCDuringTALK16:25, 14 February 2008 (UTC)[reply]
On your first point, yes, this is annoying; on the other hand it would not be wise for the system to assume that a subsequent edit by a trusted user meant that any issues arising from a bad edit had been corrected. Sometimes people edit one part of an entry without noticing that another section has been vandalized. I believe Robert was going to sic an autopatrol bot on some of the more obvious cases, but I'm not sure what happened to that plan.
As I understand it, patrolling is not intended to check the correctness of the edit, but to check for vandalism, bad formatting, and anything else that seems particularly undesirable or suspicious (a "sniff test"). So you shouldn't necessarily shy away from checking edits to FL entries and translations, although obviously caution is is in order. See Wiktionary:Grease_pit_archive/2006/November#Patrolling_edits and Help:Patrolled_edits. -- Visviva16:44, 14 February 2008 (UTC)[reply]
That GP archive page has Dvorty, inter alia, saying we shuold patrol anything not "obvious junk" and others saying we should patrol anything not "vandalism". Assuming "vandalism" is approximately equal to "purposeful junk", there's a vast difference between "obvious junk" and "vandalism". For example, any edit that I can't vouch for the accuracy of might be vandalism (but isn't usually obvious junk). I feel very uncomfortable patrolling such edits. Is the consensus that we do so?—msh210℠20:42, 14 February 2008 (UTC)[reply]
AFAIK, yes. Those simply get marked as patrolled. If it looks like a bad username, or otherwise suspicious, RFV/RFD one of them (then mark it as patrolled ~= taken care of.) --Connel MacKenzie07:30, 16 February 2008 (UTC)[reply]
My thoughts are only that the patrols for such entries are good ways to simultaneously check for some level of correctness while checking for vandalism. If I can't even read the script (and, except for Greek, it's all greek to me), all I can check for is formatting, if I even understand the peculiarities of the format for the language. It seems I may as well spend my time on something where I can make multiple kinds of contributions (reaching out to new contributors, formatting, correction, rfv, rfd) as well as vandalism patrol. The unpatrolled entries would be great place for folks with those language skills to work, even better than ttbcs and checktrans. I would think it would be very valuable for them to see unpatrolled entries with new translations in languages they knew. DCDuringTALK17:15, 14 February 2008 (UTC)[reply]
Here's my design how the (default) patrol box could look like. Based on the list above, maybe gold was set too high, I'd think from bronze 300, 700 and 1500. Or even 250, 500 and 1000, so that the idea would not sound that only quantity matters, and that things would be done relaxed. With the current rate, only five editors had done over 1000. There was a message on Wikipedia that some Wikipedia allows those who are not administrators to mark edits as patrolled. To allow patrolling for editors who have made 500 contributions might be a good idea and help to get more patrollers. Even then I think this must not lead into elitism, and those who patrol a lot will hopefully provide their insight and experience for the discussions about it, but that everyone can take part of the discussions. :) Does the box necessarily need something for the empty places, or could they be just without anything? Best regards Rhanyeia♥♫10:06, 16 February 2008 (UTC)[reply]
In this case it would be {{patrol box}} which calls it. Rhanyeia's design if you like it can be implemented with parameters |img=star|background=#e6f9f9 and, because we don't all like stars, there are other styles to choose from as well. DAVilla19:25, 17 February 2008 (UTC)[reply]
I've seen it and I think it's impressive work. :) Which talk page could the discussion about this go on, there or somewhere else? Best regards Rhanyeia♥♫11:01, 16 February 2008 (UTC)[reply]
I think we have should have Wiktionary:Patrolling, and Wiktionary talk:Patrolling. It would need to collect these discussions, as well as Wiktionary talk:Whitelist (and subpages) stuff. The Grease pit archives regarding Rat Patrol are also very relevant and should probably be linked (or moved/inserted) in date-sequential order. I've had talk page stuff (about all three) on my usertalk page, as has Robert Ullmann. Opinion: the stars suck. --Connel MacKenzie19:08, 17 February 2008 (UTC)[reply]
Following "Beer parlour" / "Tea Room", we could have "bug juice", "swampwater", "light beer", "fancy teacup" and "1982 Château Lafite Rothschild". --Connel MacKenzie19:40, 17 February 2008 (UTC)[reply]
I've put these numbers on the new page, but they seem a little low. You're aware that the figures listed are for a weeks' worth of patrols, but the numbers apply to a three-to-five week period? DAVilla03:04, 21 February 2008 (UTC)[reply]
I didn't notice that the numbers in the list and in the messages were not comparable. I'll place your numbers there for the time being. :) Best regards Rhanyeia♥♫08:13, 21 February 2008 (UTC)[reply]
I watched the recent changes list some time to see how easy it would be to patrol edits, and it's not easy or fast. One may have to run google searches to know if an edit is good, and someone else may have patrolled it during that, and at least at this time the recent changes list doesn't have very much edits. I think the numbers weren't too low. :) Best regards Rhanyeia♥♫13:56, 21 February 2008 (UTC)[reply]
Not all or even most changes are so complex that you would have to look in a dictionary or run a google search, and anyways you wouldn't have to do that before you mark it. You can mark it first and then decide whether it needs to be flagged. It also seems unlikely that the same two people would bring up the most recent edit. The most recent submission is trumped by the very next, so there's only a window of a few seconds in which pulling up the list would put it at the top. It's possible that someone would run down the list without refreshing, but that just buys you more time. On the other hand, if you actually find that someone else is marking them first I might suggest not trying to patrol the most recent changes at all, instead scrolling halfway down the page. For me there is always a long list of unpatrolled changes, and with the few of us it's not likely that someone else would pick the same one out of the list at random. I don't know how they'd do it with more participants, e.g. on Wikipedia, but I'm guessing that the volume of edits is proportional, and so this isn't as much of a problem as it could seem to be. DAVilla15:59, 21 February 2008 (UTC)[reply]
I tried again and I think it's very difficult, unless there's clear vandalism or nonsense as I just encountered, but some time ago I watched for an hour and there weren't anything so clear. And there aren't so many edits from IP addresses or very new users, and many of them are with different languages. Sometimes down the page there might be the most difficult ones left? I wouldn't take the numbers any higher. The Help:Patrolled edits advises to fix the edit if it isn't good, but if tagging it with a template is another option, do you think those template categories could be linked from Help:Patrolled edits? Best regards Rhanyeia♥♫19:51, 21 February 2008 (UTC)[reply]
