Something we've been discussing for years is whether the rewards system is properly incentivizing behavior on Stacker News. It's not easy to evaluate, but there's a concept from experimental econ that might help us think about it.
Saliency is a property of incentive systems. A salient system is one that provides sufficient payoffs for people to put in the effort to figure out their optimal strategy.
This is a major point of tension in designing economic experiments, because increasing saliency usually means increasing the expense of the experiment. But, an experiment that doesn't have salient payoffs will fail to provide useful results because the participants don't bother trying to think about their strategies.
Not Salient
In one sense, the SN rewards are obviously not salient. Even those of us who have spent a lot of time thinking about the details of this system don't really try to optimize our rewards.
Every so often, someone successfully games the rewards system by zapping small amounts to what will become top posts very quickly. The rarity of these occurrences is proof that the system is not salient. It takes a lot of effort to net an extra few thousand sats and there are easier ways to get a few thousand sats.
It's also not a bad thing that the rewards aren't salient, because they don't precisely incentivize desired behaviors. As I've previously written about, Stacker News rewards are a Keynesian Beauty Contest.
Vaguely Salient
@k00b has talked about wanting the rewards system to essentially implement the heuristic of "MOAR rewards for doing good stuff" and I think that's largely been a success.
People do produce higher quality content here than on most free-to-post platforms, comments tend to be both more numerous (relative to user count) and more thoughtful, and people zap much more than they do on nostr.
Thousands of sats are sufficient incentive to zap good content and be more thoughtful when making posts and comments.
This is the sweet spot, where the rewards are large enough to incentivize adopting a pro-social heuristic but not so large as to incentivize figuring out the antisocial optimum.
