I’ve been fielding a reoccurring comment while presenting our most recent wireframes of Wayfor over the last few days. The comment arises as we demonstrate the way our recommendation engine works to pair a user with an existing issue group. The interaction goes something like this:
A user finds from a list (or through a search function) a broad issue they are interested in, say health care reform (why not). That issue is then loaded into something we call the ‘examiner’ that breaks the issue down into its policy components. In the case of health care, this would be a list of familiar terms like ‘public option’, ‘vouchers’ and ‘interstate exchange’. The user can then investigate any of these policy initiatives further by reading a short factual article about them. Now, everyone’s good and clear up till this point (some people complain about having to read too much). What happens next in the interactions is what has been causing the comment: The user is given the opportunity to designate a ranking for any policy initiative. They can ‘Love It’, ‘Like It’ or ‘Live With It’. Once they have made one selection for each of these designations (one ‘‘Love It’, one “Like It’ and one “Live With It’) the recommendation engine goes to work to find the group that best expresses the user’s position on the issue. At this point, people ask me some variation of this: Well, what if I oppose an initiative? Why is there only the possibility to express one of three degrees of support for an initiative? If I oppose an initiative, shouldn’t I be able to express that? And the answer is yes, of course. But in our system, you oppose something by promoting an initiative that you support not by opposing one that you don’t. That is, our interaction is not designed to support direct opposition within groups. You oppose something by finding the oppositional policy initiative and working within the group that supports it. There is no standing in other people’s (or group’s) way. There’s only forward movement.
For me, this brings to mind the amazing design of our blood veins - those vessels that return de-oxygenated blood back to the heart. Veins are equipped with small, one-way valves that make blood flow only possible in one direction. These one-way valves ensure that the oxygenated blood is always able to be delivered where it is needed and is never competing with de-oxygenated blood at the point of delivery because that blue blood is pushing forward and forward only.
So then, a new design question arises - How do we design a system to deal with the impulse some users might have to stymie a groups progress out of opposition to that groups position on an issue. More simply put, what’s to stop someone from joining a group for the sole purpose of standing in the way of that group’s progress? The answer lies somewhere in the use of user Karma. However, as I’ve expressed in an earlier post, outward facing Karma won’t do the trick for the reasons that Randy Farmer and Bryce Glass lay out so well in their blog about building web reputations. So, instead of publicly displayed Karma that everybody can see, what if users kept their own private Karma about other users? That is, you could rate other users the way you rate songs in iTunes. But rather than those ratings being shared with everyone in the group they would just be for you to see and refer to. That way, you could flag someone who appears to be in the group just to halt progress despite having declared a similar enough position on your group’s issue to get recommended into the group in the first place. If there were then an algorithm that took into account the number of these ‘bad faith flags’ this troublemaker has from other members, perhaps their vote on a group proposal gets less weight, or is not counted. These jerks might not even know that anyone has flagged them and may just continue trying the create havoc unaware that they’re aren’t even being heard. The users who flagged the offender would just learn to ignore them and like your pest of a little brother, they might simply go away.
Having proposed that system in class at then end of last week generated a good bit of excitement and uncovered the notion that the sort of forward only model should be at the forefront of the branding and naming of the project. We need to make it clear that what’s different about us is that instead of standing in the way of other people’s progress, we give you the tools to make better progress if you want your position heard and possibly adopted by your lawmaker. In that way, it then becomes a foot race for groups within our system to get the most members, draft the most convincing platform declarations and get to their legislator first. It’s like free market democracy. It’s as American as apple pie…