41 sats \ 23 replies \ @kr 20 Mar \ on: Million Sat Madness is a Keynesian Beauty Contest meta
interesting ideas, appreciate the write-up.
in theory, i like the idea of splitting up rewards over different timeframes but the complexity of our rewards structure is already a lot to handle.
explaining the details of daily (or monthly) rewards to a total newcomer is still a challenge, and i think adding in 4 tiers for rewards distribution makes it even harder for a newcomer to grasp.
reply
I am still waiting... 😂
That complexity is a feature not a bug. At least that's how I'm thinking about it.
I think what we want is for stackers to just zap the content they find valuable. One way to do that is to make it not worth their time to attempt to game the rewards system and to just have a general understanding that you get rewarded for zapping good content.
In experimental econ, one of the ways experiments can fail to elicit responses aligned with the incentives is what's called a saliency problem. The difference in payouts is not large enough for the participant to bother figuring out the incentive structure and they just do what they feel like.
This is essentially imposing a saliency problem onto Stacker News.
reply
I am still saying what I always said here on SN: people should zap wisely.
After the halving, people will start appreciating more each sat.
And the halving is near.... right after this MSM.
And will be BRUTAL.
reply
i hear you on the saliency problem, but SN funds MSM with 3 million sats so we do want to make stackers aware of the contest (and the sats).
raising rewards amounts has been positively correlated to improved key metrics across the site, so i think downplaying the sats might have the opposite effect.
reply
My issue isn't with the reward being large, it's that people are potentially trying to get the reward by gaming the system instead of trying to get the reward through heightened engagement. The problem is stackers having a sense that there's a big payoff differential between zapping the content you like and zapping the content that will earn a reward.
Raising the reward amount is awesome and I'd expect it to yield positive outcomes.
reply
reply
This is largely anecdotal, but I have noticed much lower zaps on my link posts. People know link posts don't do as well, so there's a disincentive to zap them.
Personally, I intentionally increased my post zapping about a week ago. I had been rationing my sats a little bit, but realized that I was shooting myself in the foot. I'm not zapping anything just because of the author, but I am probably clicking on certain people's posts more often, which leads to them getting larger zaps.
I don't expect to see this really peak until next week anyway, as people sprint to the finish. I did see a comment in the saloon about someone checking out from stacker news until next week so that they can be sure to have resources at their disposal.
I'm not saying we have some huge problem. This is more an expression of concern that over time the behavior on SN will conform to the incentive structure and the incentive structure has a well known flaw that is beginning to show itself.
reply
fair enough, and appreciate you raising these concerns.
the daily rewards structure has its own flaws too, and on balance the key site metrics seem to be better in March (total zaps, items created, etc…).
the only metric that hasn’t really budged is new signups, but we’ve got a separate experiment planned for that shortly.
reply
I'll just say one thing, in the first week of April I'm going to do like Siggy and take a week off :) pura madness
reply
As always, I look forward to the new experiment.
reply
The rewards might not be big enough yet - but in general people game the system to make money. It is common on all the other social medias.
Over MSM, I have been giving zaps more freely, and in general give them to anyone who replies to me thoughtfully. I will 100% reconsider this now that I know I would get more back giving them to highly zapped comments.
Maybe that makes me a jerk, but those are the incentives.
reply
that’s not exactly how the incentives work, but i get your point.
all else equal, zapping the most popular comment after everyone else has already zapped it is not going to move the needle as much as zapping an interesting comment which others haven’t yet discovered.
reply
I see just as much value in those who find good content no one else has zapped and spreading the wealth. In fact, that is what I think must happen for an economy to flourish. See Jack's 'balancing redistribution' from billionaire to underfunded FOSS, and OpenSats. I think the premise that 'zapping popular content ' should be rewarded is flawed at the outset. It is mitigated by making it more specific: 'zapping popular content early' but I also see that it makes the incentives a game of guessing, which cannot be ideal at scale.
reply
the premise that 'zapping popular content ' should be rewarded is flawed at the outset
Right. The issue is that they don't want people just randomly zapping crap for the sake of a reward payout. They want people to zap "good" content. How do you gauge that other than by equating "good" and "popular"?
Btw, the trust score does a lot of work here, as I understand it. However, that also compounds the problem. It doesn't take long to get a sense of who has the high trust scores and then you can attempt to mimic them.
reply
It always has to be 'high trust scores to me'. My zaps go to reward those I trust locally and their rewards can only go to me if they trust me locally.
just randomly zapping crap for the sake of a reward payout
There must be something I don't understand. I don't see how this could be economically viable? How would spending sats randomly earn me sats overall?
Any system must incentivize creators more than zappers. Zappers being incentivized to zap may be necessary, but it reminds me of the fallacy that tax/theft + welfare programs are needed because humans aren't kind enough to take care of each other. Not equivalent, but similar. If it is needed, perhaps it should just be by lottery. I have to think on it more.
reply
I don't see how this could be economically viable? How would spending sats randomly earn me sats overall?
It doesn't make sense as an equilibrium outcome (absent some outside subsidy). However, most stackers are zapping well below the equilibrium level right now, so there are sats to be made just by zapping content. I'm living proof of this and I've been admonishing everyone else for their miserly behavior pretty much the entire time I've been on SN. The hodl mentality is working against them in this case.
I actually like incentivizing zappers more, since there's no other financial reason to do it. If people are rewarded for zapping "well", then by necessity creators are being rewarded. Shifting the burden of producing zaps onto creators seems like pushing on a rope to me.
reply
Just a quick thought, I wonder if changing the color of their username or cowboy hat based on how much they have zapped would cause others to zap them back more.
reply
Many cool cowboy hat ideas have been discussed. I tried to aggregate them here: #383551
I think there's some low-hanging fruit there.
reply