As always when it comes to data-driven decision, the math seems scary, but is actually straight forward: the conclusion that the score should go through a logistic formula is actually a direct consequence of the innocuous decision to take a specific likelihood form.<p>Similarly, the key aspect is over-looked: ‘honeypots’? What do you mean by that? Are SuperUsers aware of what constitutes a non-acceptable points? Are edge-cases justified and explained? Most places I see exist but their structure is ambivalent: train stations that have two terminals, several restaurants, a corresponding metro station… Should each platform have its point?<p>The gym I go to recently opened; they have a large swimming section, with two pools (actually five, but the extra three are smaller), changing rooms each with a sauna, etc.: Where can I find a clear explanation of how to describe the place, add venues? Consistent purpose? So, one unit for the whole? ‘Any place big enough so that if you are looking for a friend, you'll be able to notice him there?’ In that case, the bar is separated by a glass wall from the lower pool, so those should be together, far from the lower-deck weightroom. Same for the nearby re-opened hotel who added a spa, a restaurant, a nightclub, a bar… What I see when checking in are seemingly inconsistent structures, and no explanation how to edit them properly.<p>Far more than the fairly straightforward ‘good behaviour’ possible rules (which have little actual impact), clarifying the goal of the tool would help other understand how to implement ‘gamification’. See Wikipedia for a stricking case of that.
Maybe I'm missing something, but it seems like they are just applying a simple Bayesian model to classify change requests? That doesn't seem to have much to do with gamification. Can anyone explain what I'm missing?
Clearly that venue data is valuable to them. Guaranteeing 99% accuracy can be a crucial selling point for 3rd parties. With a subscription-based offering, that data can provide a significant revenue stream.