There’s a game on Twitter right now called Spymaster — I’m not exactly sure what it is because I haven’t joined. I just know about it because a lot of people I follow on Twitter are playing it, and one gets the impression that any action they take in the game (“assassinating” each other or whatnot), gets tweeted back into their followers. Apparently, there are settings in the game to limit the tweetspam, but nobody seems to be doing that.
Maybe they should rename this game to SpamMaster.
It reminds me, and some other people, of the early days of Facebook platform, around two years ago. Then, application developers gained users — and pissed them off — by sending out automated messages from the application to friends’ news feeds. Facebook responded, eventually, by altering its algorithm to mostly stop these app notifications from appearing.
In its latest redesign, though, Facebook has dropped its news feed algorithm in favor of letting users manually decide what they want to filter, whether application notifications or anything else. This is more like how lifestreaming service FriendFeed has worked for awhile — you decide who to follow on the service, and you can decide what sort of things those people share that you want to see in your feed.
Twitter needs some sort of solution along these lines — after all, it wants to be a platform company for social games like Spymaster, right? But doing that is going to be tricky. How does Twitter decide what to filter automatically? Anything? Maybe create a filter for third-party applications, that appears on the right-hand menu column on the Twitter homepage interface (that’s what FriendFeed’s interface already looks like, for what it’s worth). Maybe Twitter could use Facebook Connect and other social data-sharing services to filter friends by service? So you could see a Twitter for just your Facebook friends, for example. Maybe filtering using hashtags (Twitter’s user-created filtering mechanism) about particular topics, as some have suggested?
In the meantime, lots of Twitter users are being spammed, and it’s annoying.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.