Good day Wikipedians,
I have of late got into a football management computer game. Don't panic, I will be relating this post to Wikipedia, hang on. I'm really enjoying the game. To such an extent that I've actually started to follow football. I've never particularly liked football. I only started playing the computer game cos there was a free demo. Now I like the computer game so much I'm following football in the real world.
After quite a few hours of playing it struck me that all I was really doing most of the time was evaluating numbers: player abilities rated out of 5, 10 or 20 depending on the stat in question. Numbers of goals. Numbered position in league. Tier of football I'm playing in.
I don't know why this should be so compelling. Watching numbers change. But the game is incredibly successful (some editions have broken records for fastest selling computer game according to our articles).
The numbers are clearly giving us players an emotional response. They engage.
Last year, during the Strategy process and before I started playing this game, I proposed that what Wikipedia needed was "more rewards" for editors. I proposed a few things. In the end we got Wiki-love, which I support and like, but they isn't really like what I proposed at Strategy. To be honest I can barely remember what it was I proposed back then...
I still think we could do with more rewards and maybe this damned game has given me an answer.
More editor stats.
All of us who have been around for some time know that edit counts are not very reliable indicators of effort. Nevertheless we still do keep a public record of editors with high counts. I think there's a reason for that. I think it's because we still, despite protestations, know that an edit count does tell us *something* about a Wikipedian. Even if it's just "(s)he edits a lot".
I believe I'm right in saying that the Foundation is in the process of setting up something like Toolserver. I suggest we plan to put it to work. I suggest we expand greatly the stats we keep on individual editors and form league tables from them. I believe that aiming for a place in a table will motivate people. I realise that a) this is unproven and b) there will be objections, particularly regarding 'gaming the system and 'unintended consequences' but perhaps we can discuss those and mitigate them (more later).
New Stats that could be placed in league tables could include:
* Length of service (difference in days between first edit and last) * Number of consecutive days/months/weeks where 5 or more edits have been made (or 50 edits, or a hundred): in short there could be quite a number of these tables that relate to consistency and number of edits all of which, I feel, might spur people on to keep contributing. * Most characters/bytes added (without being removed) * Most blocks for admins * Most welcomes, barn stars awarded * Most reverts / undos * Average reader-rating of articles user has edited at least ten times
You could also have these as percentage of number of edits and rank for those too, eg welcomes, blocks or reverts as a percentage of total edits, (with a minimum number of edits to qualify for inclusion on the table).
Now, it could be (WILL be!) that someone decides "I'm going for the revert league title" and starts doing things we wouldn't ideally like (to put it mildly). However their presence at the head of the league, I feel, will actually subject their edits to greater scrutiny. People will look at their contributions and it may well result in needed censure, showing their activity to be undesirable and action could be taken accordingly. Also, you may have people in the top table who aren't even *trying* and their presence at or near the top might cause some examination of their contribs.
Perhaps you can think of some league tables that would really push desirable behaviours at minimal risk of negative ones?
If you don't like this idea I'd like to hear the concerns, HOWEVER! I would also like you to just entertain the idea and - even if you're against - think of some individual editor stats that could be tracked you think *may* provide useful feedback, even if you ultimately don't think we *should*.
So: I propose we greatly increase feedback on user performance to drive people on. Support editor stats today.
User:Bodnotbod
Hi Bod,
We've actually got some of these already, many in widespread use.
I'm not the only editor with {{User:EVula/Userboxes/admin since|year=yyyy|month=mm|day=dd}} or {{User Wikipedian For|day=dd|month=mm|year=yyyy}}
Or indeed {{user contrib}} and even {{User:WereSpielChequers/Userbox Editcountitis}}
http://en.wikipedia.org/wiki/Wikipedia:ADMINSTATS gives you blocks per admin and a whole bunch of other stats including several top 25s.
I was involved in getting http://en.wikipedia.org/wiki/Wikipedia:EDITSextended from the 4,000 with the highest edit counts to the top 8,000 with extra separate files for the next two thousand (for space issues we couldn't do 10,000 in one list). Extending the list to 8,000 brought it into range for many newer or less active editors. You now need well over 12,000 edits to get in the top 4,000, in early 2008 it was less than 6,000 - it takes as many edits to get into the top 8,000 today as it did to get into the top 4,000 in early 2008.
I don't know if the other features you wanted exist, but if there is demand they may well do already.
WereSpielChequers
On 30 September 2011 17:46, Bod Notbod bodnotbod@gmail.com wrote:
Good day Wikipedians,
I have of late got into a football management computer game. Don't panic, I will be relating this post to Wikipedia, hang on. I'm really enjoying the game. To such an extent that I've actually started to follow football. I've never particularly liked football. I only started playing the computer game cos there was a free demo. Now I like the computer game so much I'm following football in the real world.
After quite a few hours of playing it struck me that all I was really doing most of the time was evaluating numbers: player abilities rated out of 5, 10 or 20 depending on the stat in question. Numbers of goals. Numbered position in league. Tier of football I'm playing in.
I don't know why this should be so compelling. Watching numbers change. But the game is incredibly successful (some editions have broken records for fastest selling computer game according to our articles).
The numbers are clearly giving us players an emotional response. They engage.
Last year, during the Strategy process and before I started playing this game, I proposed that what Wikipedia needed was "more rewards" for editors. I proposed a few things. In the end we got Wiki-love, which I support and like, but they isn't really like what I proposed at Strategy. To be honest I can barely remember what it was I proposed back then...
I still think we could do with more rewards and maybe this damned game has given me an answer.
More editor stats.
All of us who have been around for some time know that edit counts are not very reliable indicators of effort. Nevertheless we still do keep a public record of editors with high counts. I think there's a reason for that. I think it's because we still, despite protestations, know that an edit count does tell us *something* about a Wikipedian. Even if it's just "(s)he edits a lot".
I believe I'm right in saying that the Foundation is in the process of setting up something like Toolserver. I suggest we plan to put it to work. I suggest we expand greatly the stats we keep on individual editors and form league tables from them. I believe that aiming for a place in a table will motivate people. I realise that a) this is unproven and b) there will be objections, particularly regarding 'gaming the system and 'unintended consequences' but perhaps we can discuss those and mitigate them (more later).
New Stats that could be placed in league tables could include:
- Length of service (difference in days between first edit and last)
- Number of consecutive days/months/weeks where 5 or more edits have
been made (or 50 edits, or a hundred): in short there could be quite a number of these tables that relate to consistency and number of edits all of which, I feel, might spur people on to keep contributing.
- Most characters/bytes added (without being removed)
- Most blocks for admins
- Most welcomes, barn stars awarded
- Most reverts / undos
- Average reader-rating of articles user has edited at least ten times
You could also have these as percentage of number of edits and rank for those too, eg welcomes, blocks or reverts as a percentage of total edits, (with a minimum number of edits to qualify for inclusion on the table).
Now, it could be (WILL be!) that someone decides "I'm going for the revert league title" and starts doing things we wouldn't ideally like (to put it mildly). However their presence at the head of the league, I feel, will actually subject their edits to greater scrutiny. People will look at their contributions and it may well result in needed censure, showing their activity to be undesirable and action could be taken accordingly. Also, you may have people in the top table who aren't even *trying* and their presence at or near the top might cause some examination of their contribs.
Perhaps you can think of some league tables that would really push desirable behaviours at minimal risk of negative ones?
If you don't like this idea I'd like to hear the concerns, HOWEVER! I would also like you to just entertain the idea and - even if you're against - think of some individual editor stats that could be tracked you think *may* provide useful feedback, even if you ultimately don't think we *should*.
So: I propose we greatly increase feedback on user performance to drive people on. Support editor stats today.
User:Bodnotbod
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
On Sep 30, 2011 9:47 AM, "Bod Notbod" bodnotbod@gmail.com wrote:
Good day Wikipedians,
I have of late got into a football management computer game. Don't panic, I will be relating this post to Wikipedia, hang on. I'm really enjoying the game. To such an extent that I've actually started to follow football. I've never particularly liked football. I only started playing the computer game cos there was a free demo. Now I like the computer game so much I'm following football in the real world.
After quite a few hours of playing it struck me that all I was really doing most of the time was evaluating numbers: player abilities rated out of 5, 10 or 20 depending on the stat in question. Numbers of goals. Numbered position in league. Tier of football I'm playing in.
I don't know why this should be so compelling. Watching numbers change. But the game is incredibly successful (some editions have broken records for fastest selling computer game according to our articles).
The numbers are clearly giving us players an emotional response. They
engage.
Last year, during the Strategy process and before I started playing this game, I proposed that what Wikipedia needed was "more rewards" for editors. I proposed a few things. In the end we got Wiki-love, which I support and like, but they isn't really like what I proposed at Strategy. To be honest I can barely remember what it was I proposed back then...
I still think we could do with more rewards and maybe this damned game has given me an answer.
More editor stats.
All of us who have been around for some time know that edit counts are not very reliable indicators of effort. Nevertheless we still do keep a public record of editors with high counts. I think there's a reason for that. I think it's because we still, despite protestations, know that an edit count does tell us *something* about a Wikipedian. Even if it's just "(s)he edits a lot".
I believe I'm right in saying that the Foundation is in the process of setting up something like Toolserver. I suggest we plan to put it to work. I suggest we expand greatly the stats we keep on individual editors and form league tables from them. I believe that aiming for a place in a table will motivate people. I realise that a) this is unproven and b) there will be objections, particularly regarding 'gaming the system and 'unintended consequences' but perhaps we can discuss those and mitigate them (more later).
New Stats that could be placed in league tables could include:
- Length of service (difference in days between first edit and last)
- Number of consecutive days/months/weeks where 5 or more edits have
been made (or 50 edits, or a hundred): in short there could be quite a number of these tables that relate to consistency and number of edits all of which, I feel, might spur people on to keep contributing.
- Most characters/bytes added (without being removed)
- Most blocks for admins
- Most welcomes, barn stars awarded
- Most reverts / undos
- Average reader-rating of articles user has edited at least ten times
You could also have these as percentage of number of edits and rank for those too, eg welcomes, blocks or reverts as a percentage of total edits, (with a minimum number of edits to qualify for inclusion on the table).
Now, it could be (WILL be!) that someone decides "I'm going for the revert league title" and starts doing things we wouldn't ideally like (to put it mildly). However their presence at the head of the league, I feel, will actually subject their edits to greater scrutiny. People will look at their contributions and it may well result in needed censure, showing their activity to be undesirable and action could be taken accordingly. Also, you may have people in the top table who aren't even *trying* and their presence at or near the top might cause some examination of their contribs.
Perhaps you can think of some league tables that would really push desirable behaviours at minimal risk of negative ones?
If you don't like this idea I'd like to hear the concerns, HOWEVER! I would also like you to just entertain the idea and - even if you're against - think of some individual editor stats that could be tracked you think *may* provide useful feedback, even if you ultimately don't think we *should*.
So: I propose we greatly increase feedback on user performance to drive people on. Support editor stats today.
User:Bodnotbod
I absolutely agree. We keep running into this problem (edit count as the only universal metric for success) all over the place. You mentioned the Wikimedia Labs project (i.e. the Toolserver equivalent), but I think one thing we could do now would be to go take a look at the mockup currently built for GlobalProfiles and let engineering staff know what stats/info you think could/should be included in order to mitigate the "editcountitis" issue. Note that it's just a proposed design doc, so if you agree that we need better indicators of the work people do for the encyclopedia now's the time to speak up.
https://secure.wikimedia.org/wikipedia/mediawiki/wiki/GlobalProfile/design
Steven
Re the suggestion that edit count is the only universal metric for success. Editcount is a very long way from being the only universal metric for success.
Length of Tenure and cleanliness of block record are just as universal.
Adminship and other userights, are also universal, though I'd concede that adminship on some projects is seen as easier to get than on others, but then automated and semiautomated edits are perceived as less worthy than manual ones.
Featured content is at least Wikipedia wide, though I confess I don't know wiktionary or Wikisource well enough to know if they have an FA style system.
I'm pretty sure that Barnstars are universal, though maybe someone could do a survey of the wikis to check that.
Linguistic skills. Babel boxes are another thing that matter in all our multingual wikis.
Wikiness is a new one on me - http://toolserver.org/~erwin85/xcontribs.php?user=WereSpielChequers I've just come across it in the Stewards elections, so it is both somewhat specialised and at the same time something that editors from many different wikis can appreciate and clearly many are judging each other by. You could argue that it is a function of editcount and linguistic skills, but I think it more than that as I probably outscore some editors who are far from monolingual.
Toolserver access and the ability to code are universally valued. As I believe are various other editing skills.
If you want more I'd suggest creating some modules of Computer based training, with of course appropriate userboxen for those who complete them with a high enough score. It would be great to have one for Newpage patrollers to guide them through the minefield that is speedy deletion tagging, and there are several other areas where a shift from learning on the job to learning via a gamelike training modules would raise quality, reduce angst and I believe greatly reduce levels of newby biting. If we invested in some for new admins we might even address the two most common reasons for block histories amongst our most active editors.
WereSpielChequers
I absolutely agree. We keep running into this problem (edit count as the only universal metric for success) all over the place. You mentioned the Wikimedia Labs project (i.e. the Toolserver equivalent), but I think one thing we could do now would be to go take a look at the mockup currently built for GlobalProfiles and let engineering staff know what stats/info you think could/should be included in order to mitigate the "editcountitis" issue. Note that it's just a proposed design doc, so if you agree that we need better indicators of the work people do for the encyclopedia now's the time to speak up.
https://secure.wikimedia.org/wikipedia/mediawiki/wiki/GlobalProfile/design
Steven _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
On Sat, Oct 1, 2011 at 11:15 AM, WereSpielChequers < werespielchequers@gmail.com> wrote:
Re the suggestion that edit count is the only universal metric for success. Editcount is a very long way from being the only universal metric for success.
Length of Tenure and cleanliness of block record are just as universal.
Adminship and other userights, are also universal, though I'd concede that adminship on some projects is seen as easier to get than on others, but then automated and semiautomated edits are perceived as less worthy than manual ones.
Featured content is at least Wikipedia wide, though I confess I don't know wiktionary or Wikisource well enough to know if they have an FA style system.
I'm pretty sure that Barnstars are universal, though maybe someone could do a survey of the wikis to check that.
Linguistic skills. Babel boxes are another thing that matter in all our multingual wikis.
Wikiness is a new one on me - http://toolserver.org/~erwin85/xcontribs.php?user=WereSpielChequers I've just come across it in the Stewards elections, so it is both somewhat specialised and at the same time something that editors from many different wikis can appreciate and clearly many are judging each other by. You could argue that it is a function of editcount and linguistic skills, but I think it more than that as I probably outscore some editors who are far from monolingual.
Toolserver access and the ability to code are universally valued. As I believe are various other editing skills.
If you want more I'd suggest creating some modules of Computer based training, with of course appropriate userboxen for those who complete them with a high enough score. It would be great to have one for Newpage patrollers to guide them through the minefield that is speedy deletion tagging, and there are several other areas where a shift from learning on the job to learning via a gamelike training modules would raise quality, reduce angst and I believe greatly reduce levels of newby biting. If we invested in some for new admins we might even address the two most common reasons for block histories amongst our most active editors.
WereSpielChequers
The point is not that there aren't other things worth valuing or that some set of very, very experienced people in the community value. It's that there are ways we could make things other than edit count a core part of defining identity in the projects. Because they're not right now.
Toolserver access? Clean block log? Barnstars?
How many times have you had to take time and explain what those even are? Most people who've ever edited don't even really understand the concepts much less think of them as metrics for success. Ask 100 random Wikipedians -- especially those with an edit count lower than 1,000 or who aren't in a big project like English or German -- about any of those things, and I bet you they won't know what half of them even are.
It is not universal if it's only visible and understandable to people who are already extremely active in Wikipedia.
I absolutely agree. We keep running into this problem (edit count as the only universal metric for success) all over the place. You mentioned the Wikimedia Labs project (i.e. the Toolserver
equivalent),
but I think one thing we could do now would be to go take a look at the mockup currently built for GlobalProfiles and let engineering staff know what stats/info you think could/should be included in order to mitigate
the
"editcountitis" issue. Note that it's just a proposed design doc, so if
you
agree that we need better indicators of the work people do for the encyclopedia now's the time to speak up.
https://secure.wikimedia.org/wikipedia/mediawiki/wiki/GlobalProfile/design
Steven _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
Editor since 2007 is instantly understandable to everyone, I would say it is at least as easily understood as editcount
I've never had to explain what a clean blocklog was, and I don't recall having to explain barnstars. Toolserver access, yes I have explained to people, but even our most untechnical editors have no problem grasping the concept that there are people who can do various levels of programming.
But your question was about "a core part of defining identity in the projects" if you want "a core part of defining editors wiki skills outside of the projects" then you need a completely different focus. When it comes to newbies and near newbies then I would suggest a more targeted personalised approach is better. Goodfaith editors don't need to be told bout the blocking process until they have an encounter with someone who merits a block, likewise FAC is rather more relevant to some editors than others.
For non-wikimedians then I think you need to consider the relevant skillsets that the wikimedian needs to display. I know one editor who answered the PC literacy question in a job interview by saying he was an admin on EN wiki. We have lots of editors each year who apply for university, I know at least one who has done so with a letter of commendation from an academic they have collaborated with on Wiki; This is probably an area we could greatly expand - though you'd need to do some more identifications to the office. In the future I suspect Academia will embrace the digital revolution and get used to the idea of marking candidates for the quality of their update to an earlier work rather than for rewriting stuff from scratch. Once that happens it will be much easier to explain wiki experience in ways that professors understand.
Where the two could overlap is in my old hobby horse of Computer Based Training. I'd love to see the scout movement awarding vandalfighter and Wikipedia editor badges to scouts who've done the training and then demonstrated their new skill.
WereSpielChequers
On 1 October 2011 20:03, Steven Walling steven.walling@gmail.com wrote:
On Sat, Oct 1, 2011 at 11:15 AM, WereSpielChequers < werespielchequers@gmail.com> wrote:
Re the suggestion that edit count is the only universal metric for
success.
Editcount is a very long way from being the only universal metric for success.
Length of Tenure and cleanliness of block record are just as universal.
Adminship and other userights, are also universal, though I'd concede
that
adminship on some projects is seen as easier to get than on others, but then automated and semiautomated edits are perceived as less worthy than
manual
ones.
Featured content is at least Wikipedia wide, though I confess I don't
know
wiktionary or Wikisource well enough to know if they have an FA style system.
I'm pretty sure that Barnstars are universal, though maybe someone could
do
a survey of the wikis to check that.
Linguistic skills. Babel boxes are another thing that matter in all our multingual wikis.
Wikiness is a new one on me - http://toolserver.org/~erwin85/xcontribs.php?user=WereSpielChequers I've just come across it in the Stewards elections, so it is both
somewhat
specialised and at the same time something that editors from many
different
wikis can appreciate and clearly many are judging each other by. You
could
argue that it is a function of editcount and linguistic skills, but I
think
it more than that as I probably outscore some editors who are far from monolingual.
Toolserver access and the ability to code are universally valued. As I believe are various other editing skills.
If you want more I'd suggest creating some modules of Computer based training, with of course appropriate userboxen for those who complete
them
with a high enough score. It would be great to have one for Newpage patrollers to guide them through the minefield that is speedy deletion tagging, and there are several other areas where a shift from learning on the job to learning via a gamelike training modules would raise quality, reduce angst and I believe greatly reduce levels of newby biting. If we invested in some for new admins we might even address the two most common reasons for block histories amongst our most active editors.
WereSpielChequers
The point is not that there aren't other things worth valuing or that some set of very, very experienced people in the community value. It's that there are ways we could make things other than edit count a core part of defining identity in the projects. Because they're not right now.
Toolserver access? Clean block log? Barnstars?
How many times have you had to take time and explain what those even are? Most people who've ever edited don't even really understand the concepts much less think of them as metrics for success. Ask 100 random Wikipedians -- especially those with an edit count lower than 1,000 or who aren't in a big project like English or German -- about any of those things, and I bet you they won't know what half of them even are.
It is not universal if it's only visible and understandable to people who are already extremely active in Wikipedia.
I absolutely agree. We keep running into this problem (edit count as
the
only universal metric for success) all over the place. You mentioned the Wikimedia Labs project (i.e. the Toolserver
equivalent),
but I think one thing we could do now would be to go take a look at the mockup currently built for GlobalProfiles and let engineering staff
know
what stats/info you think could/should be included in order to mitigate
the
"editcountitis" issue. Note that it's just a proposed design doc, so if
you
agree that we need better indicators of the work people do for the encyclopedia now's the time to speak up.
https://secure.wikimedia.org/wikipedia/mediawiki/wiki/GlobalProfile/design
Steven _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
On Fri, Sep 30, 2011 at 5:46 PM, Bod Notbod bodnotbod@gmail.com wrote:
<snip>
I still think we could do with more rewards and maybe this damned game has given me an answer.
More editor stats.
I think what is needed is some way to measure quality objectively. It may be that only quantity can be measured objectively, and that quality can only be measured subjectively. But I'd support something that moved the focus away from quantity towards quality.
Carcharoth
Ideally if we are going to push gammification it should be centered on quality content primarily.
Gammification is hard to pull off in a way that ensures maintaining quality in output - because by it's very nature such a system is gameable. And you will tend to find, anyway, that the most important contributors have little interest in such things.
If we go this route, it needs to be set up to encourage people to work to produce good content - rather than rewarding "not very much".
This is the problem with Wiki-Love - it makes Barnstars ever more frivolous, and essentially useless in encouraging good work.
Tom
We have FA for those who want to focus on one article, we have http://en.wikipedia.org/wiki/Wikipedia:CUP as a somewhat game like process for content contributors. What aspect of content contribution do we not have a game like feature for?
WereSpielChequers
On 1 October 2011 17:01, Thomas Morton morton.thomas@googlemail.com wrote:
Ideally if we are going to push gammification it should be centered on quality content primarily.
Gammification is hard to pull off in a way that ensures maintaining quality in output - because by it's very nature such a system is gameable. And you will tend to find, anyway, that the most important contributors have little interest in such things.
If we go this route, it needs to be set up to encourage people to work to produce good content - rather than rewarding "not very much".
This is the problem with Wiki-Love - it makes Barnstars ever more frivolous, and essentially useless in encouraging good work.
Tom _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
OTOH quantity has a quality all of its own. So far as I know, there's no good measure of how much text people have contributed that still remain in the article.
It would be a really good idea to measure how many unique strings of characters each editor has added to each article and in general. Presumably anything that still remains is of sufficient quality for whatever level the article is.
On 1 October 2011 16:52, Carcharoth carcharothwp@googlemail.com wrote:
On Fri, Sep 30, 2011 at 5:46 PM, Bod Notbod bodnotbod@gmail.com wrote:
<snip>
I still think we could do with more rewards and maybe this damned game has given me an answer.
More editor stats.
I think what is needed is some way to measure quality objectively. It may be that only quantity can be measured objectively, and that quality can only be measured subjectively. But I'd support something that moved the focus away from quantity towards quality.
Carcharoth
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
Actually, you find that as certain types of articles rise to the level of being featured (or just improve in general), they usually end up having being rewritten almost in their entirety. Some of the information added earlier may still be there, but the wording will invariably have been changed. If there was a way to measure the 'content' of an article, not just the character strings, that would be a better measure.
Also, what you will find when looking for editors who have text strings that survive longest is that you will find those who are most assiduous about keeping track of the articles on their watchlist. That can be either ownership (bad) or stewardship of a well-done article (good).
The assumption "Presumably anything that still remains is of sufficient quality for whatever level the article is" has so much wrong with it that I don't know where to start. It is quite common for the final push for an article to be featured to involve different editors to those that brought it to the current state. And that often involves stepping back, taking a long hard look at the article and the sources, and then ripping up large quantities of the article and rewriting and rebalancing things.
Whether that is building on what went before, or not, I'm still not entirely sure.
Carcharoth
On Sat, Oct 1, 2011 at 5:32 PM, Ian Woollard ian.woollard@gmail.com wrote:
OTOH quantity has a quality all of its own. So far as I know, there's no good measure of how much text people have contributed that still remain in the article.
It would be a really good idea to measure how many unique strings of characters each editor has added to each article and in general. Presumably anything that still remains is of sufficient quality for whatever level the article is.
On 1 October 2011 16:52, Carcharoth carcharothwp@googlemail.com wrote:
On Fri, Sep 30, 2011 at 5:46 PM, Bod Notbod bodnotbod@gmail.com wrote:
<snip>
I still think we could do with more rewards and maybe this damned game has given me an answer.
More editor stats.
I think what is needed is some way to measure quality objectively. It may be that only quantity can be measured objectively, and that quality can only be measured subjectively. But I'd support something that moved the focus away from quantity towards quality.
Carcharoth
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
-- -Ian Woollard _______________________________________________ WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
On 1 October 2011 18:15, Carcharoth carcharothwp@googlemail.com wrote:
The assumption "Presumably anything that still remains is of sufficient quality for whatever level the article is" has so much wrong with it that I don't know where to start.
No, if material lasts for a long period in an article it's highly likely to be fairly good even if it gets rewritten later; and the more material and the longer it lasts, the better.
It's the area under the curve that matters, not whether it *eventually* gets rewritten.
So time_in_article * number_of_unique_characters is probably a fairly good metric.
And you could multiply by the article hit rate to get an even better metric I expect.
Whereas you can get very high edit counts by many well-known ways, even breaking an edit down into many sub-edits can multiply up edit counts, or just doing lots of vandalism reverts.
It is quite common for the final push for an article to be featured to
involve different editors to those that brought it to the current state. And that often involves stepping back, taking a long hard look at the article and the sources, and then ripping up large quantities of the article and rewriting and rebalancing things.
So you're saying that sometimes people rewrite articles, and this contradicts my point how?
Carcharoth
--
-Ian Woollard
On Sat, Oct 1, 2011 at 9:58 PM, Ian Woollard ian.woollard@gmail.com wrote:
On 1 October 2011 18:15, Carcharoth carcharothwp@googlemail.com wrote:
The assumption "Presumably anything that still remains is of sufficient quality for whatever level the article is" has so much wrong with it that I don't know where to start.
No, if material lasts for a long period in an article it's highly likely to be fairly good even if it gets rewritten later; and the more material and the longer it lasts, the better.
Material lasts a long time for two reasons:
(a) It is good and lots of people have checked it and left it alone; (b) It is bad/wrong and no-one has spotted it yet and replaced it or rewritten it.
I don't see how you can devise a metric to distinguish these two case, as you would have to detect the number of people silently checking and approving something (not just reading it). Lots of quality control is *silent* and not detectable in the current metrics. It would be different if there were a way for people to mark text and say "I have this book and have checked this citation, or followed the URL and agree with what is written here". Essentially a way to detect the silent verification that often takes place.
It's the area under the curve that matters, not whether it *eventually* gets rewritten.
So time_in_article * number_of_unique_characters is probably a fairly good metric.
Not in the case of obscure articles written by one person, not linked much from anywhere (but not triggering orphan article bots), and only small changes made over the years. View stats might help here, but probably not much as there are a vast, vast number of articles not visited very much at all. Those would account for most of the "unchanged text" you would be picking up.
And you could multiply by the article hit rate to get an even better metric I expect.
Whereas you can get very high edit counts by many well-known ways, even breaking an edit down into many sub-edits can multiply up edit counts, or just doing lots of vandalism reverts.
Yes, I never said edit count was reliable for anything or useful in any way. I'm only saying that unique text is likely not very helpful either. But the best way to find out is to actually try this and see if it shows anything useful. If it does, great. If not, then try again.
<snip>
Carcharoth
On 2 October 2011 12:28, Carcharoth carcharothwp@googlemail.com wrote:
On Sat, Oct 1, 2011 at 9:58 PM, Ian Woollard ian.woollard@gmail.com wrote:
On 1 October 2011 18:15, Carcharoth carcharothwp@googlemail.com wrote:
The assumption "Presumably anything that still remains is of sufficient quality for whatever level the article is" has so much wrong with it that I don't know where to start.
No, if material lasts for a long period in an article it's highly likely to be fairly good even if it gets rewritten later; and the more material and the longer it lasts, the better.
Material lasts a long time for two reasons:
(a) It is good and lots of people have checked it and left it alone; (b) It is bad/wrong and no-one has spotted it yet and replaced it or rewritten it.
I don't see how you can devise a metric to distinguish these two case, as you would have to detect the number of people silently checking and approving something (not just reading it). Lots of quality control is *silent* and not detectable in the current metrics. It would be different if there were a way for people to mark text and say "I have this book and have checked this citation, or followed the URL and agree with what is written here". Essentially a way to detect the silent verification that often takes place.
You mean, like, umm, Flagged Revisions?
J.
According to our article [[Facepalm]], this is a startrek internet meme indicating "an expression of embarrassment, frustration, disbelief, disgust, shame or general woe. It often expresses mockery or disbelief of perceived idiocy."
Well, that must be right.
Given that, I am wondering why we tolerate a template {{facepalm}} http://en.wikipedia.org/wiki/Template:Facepalm
This does nothing to foster civil discourse among Wikipedians. I've just looked through how it is being used, and whilst I do see the occasional use in self-deprecation, generally it is used as a shorthand put-down: implicitly calling your correspondent an idiot, and his latest contribution self-evidently moronic.
Granted, removing uncivil templates won't magically increase patient and constructive discussion, but I do suspect we'd still nevertheless delete {{jackass}} or {{moron}}. If people are going to mock others, we shouldn't be giving them shortcuts to do so. The existence of the template serves to legitimise such dismissive discourse.
Thoughts?
Scott