tommorris.org

Discussing software, the web, politics, sexuality and the unending supply of human stupidity.


Wikipedia and teaching game rules

As always, on Wikipedia there’s a big dust up going on about new page patrolling or NPP. Basically, the issue is something like this: when you create a page on Wikipedia, it gets listed on a page called Special:NewPages. If you are a logged in user and are confirmed, you can click on pages in that queue and mark them as patrolled. New page patrollers check through that list, usually from either the beginning or end of the queue, and make sure that new pages coming in are of decent enough quality.

Only there’s loads of work and not really enough people to do that work. It’s rather taxing work too: if you mark it as patrolled, you have to potentially live with the possibility that you might be the last person to look at that page. It might have a massive copyright violation, it might contain some subtle vandalism, whatever. You have to intuitively tell in about a minute or less if it needs further attention, and if it does, tag it as such, fix it, or nominate it for deletion (or perhaps delete it outright if you are an admin).

There is constant anguish amongst the new page patrollers, because the queue never seems to really go down, however hard you try. It’s always a few days away from the 30 day limit: once a page is over 30 days, it drops off the end of the new pages queue, and we don’t really want that to happen.

The other complaint is that new page patrollers don’t necessarily do a very good job: they miss obvious issues and mark stuff as patrolled, and they unnecessarily scare newbies by falling on them like a ton of bricks, deleting their pages and leaving them lots of warnings. Some have suggested that we need to certify new page patrollers, basically ensure they are competent and give them a user right to ensure that the NPPers are doing their job right. There are problems with this: some people go into NPPing because it is the first real maintenance task they come across that they can do that doesn’t require special rights. To do recent change patrolling, you tend to need rollback to use things like Huggle. And, of course, there’s adminship… which is hard to get.

Last year, a plan was floated to try making it so that new users wouldn’t be able to create new pages. To try it. Test it for six months and see if it makes the encyclopedia better. The Foundation turned it down and told the community on English Wikipedia that they should stay calm because they would have other stuff coming soon to fix the new pages problem.

Besides the political issues of WMF-community relations, the issue with the autoconfirmed trial (ACTRIAL) is that it puts a restriction in place, but does nothing to reduce the demand. People will turn up at Wikipedia and want to create new articles. We might channel them into the AfC process instead of articles going into the main namespace… but the issues haven’t disappeared.

The motivation behind ACTRIAL was noble: we would rather have users who spent a few days lurking and finding their feet by editing existing articles before jumping in and creating new articles. In the olden days of the Internet, lurking was a lot more widely practiced than today. Now, of course, you should be able to turn up at a new social site like Twitter or Facebook and get going very easily. It’s an age of user experience (much as I might, in my grumpier moments, lament that), and if a website doesn’t hold your hand and guide you through the etiquette of how you behave and what you use the site for, that’s obviously a fault with the website rather than a moral failing of the user, as a previous generation might have said.

The root of a lot of the problems with newbies on Wikipedia is that we need to explain a shit ton of rules to people when they most probably don’t care, have a very short attention span, and want to have fun and edit stuff.

Fortunately, there’s already a field who have handled this. Anyone who listens to user experience people bang on about game design will know the answer already: game designers. Pick up almost any video game and it’ll have strategies for teaching new players how to play the game. Some of these are not necessarily very well done. It can often be cliché: dramatic opening, then some obvious plot device pops up like a character whose only point is to teach you how to perform simple tasks even though you are a super-duper knife-wielding commando badass who presumably knows all this shit back-to-front. Or perhaps they have a dumb foil who can’t quite do it and asks you, the super badass, to show you.

In Final Fantasy VII, Barret, being a gun-toting muscleman, can’t quite get to grips with materia and asks you to explain it to him. Even better is when they break the fourth wall, and your NPC buddy tells you to press the ‘X’ button to fire while jiggling the joypad like this.

There are ways, then, to induct people to the game world. What they are depend on what sort of game it is. Something like Pong is fairly obvious. Most people don’t need much help to play some kind of colour matching game like Tetris or Bejeweled. Something more sophisticated will have more complex tutorial modes. Wikipedia is definitely at the harder end: more EVE Online and less Asteroids.

There’s plenty of advice online and in game design circles on tutorial modes and how to design these introductory procedures. Let’s take as an example a post called Teaching players how to play your game, which I found just by Googling. It seems to contain some sensible advice and, more importantly, some examples of how not to do it. Wikipedians might have some of those eerie deja vu feelings while reading.

Don’t teach too much too soon

If we tell people that before even touching Wikipedia they need to understand N, V, RS, GNG, NPOV, COI, MOS, AfD, MOS, ANI, WTF and all that, they’ll just fuck off back to Facebook. Or ignore it all and learn nothing from any of the advice we’re giving.

Don’t give the answers before the questions

Again, we can spout policies at people, but they don’t understand why we have policies. I’ve had this with users before: I remember calmly explaining the username policy to a user, and then told them that I’d have to softblock their account and they’d have to re-register with a personal rather than corporate account. If I had just blocked and templated, rather than explained the point of the username policy, they probably would have been angry. On Wikipedia, policies are usually created in response to something. We have reliable sources as a policy because people often use shitty sources.

This is obviously a hard one, as we don’t necessarily know what questions the user is going to ask. If the Foundation are working on improving the new editor experience, they probably should be working out roughly the order of the questions they are asking, answering those questions, and then cutting out all the clutter that doesn’t help the user do the right thing.

Don’t be a bastard

Yeah, shouldn’t this be obvious? Shouldn’t it be obvious that this should be obvious? Obviously. But obvious point obviously needs obvious reinforcement. The problem here is that unlike a game, we have a social environment. And that’s a lot harder to design. You might be lucky and get some nice person who’ll be really helpful, or you might get an utter bastard. But, of course, that’s fixable too: if you design the software in the right way, it might be possible to make it a lot easier for people to be nice and a lot harder to be mean.

The learning curve for Wikipedia is there, and it is hard, and it is probably necessary: most people do actually need to learn various skills in order to be able to write a decent encyclopedia article. And, no, the learning curve will still be there even with a visual editor. (The learning curve can be even steeper on other wikis: if writing a good three paragraph encyclopedia article is hard, think about how hard it is to write a decent dictionary definition, or a decent news article, or a decent scientific description of a biological species.)

I don’t have any solutions, sadly, just a suggestion that people who are interested in issues around new page patrolling and newbie treatment think about it through the lens of game design (I mean, people already think Wikipedia is a giant RPG, may as well make it easy for newbies to join in). The issue is how do we balance the openness of welcoming all editors and the issue of all the work their inexperience causes for patrollers. Think about it as a game tutorial design problem, not as a supply and demand problem. I’d write more, but then this might spiral out from being a blog post into being a serious, considered piece of writing.