Some thoughts on SN rewards \ stacker news ~meta

3752 sats \ 39 comments \ @elvismercury 28 Mar meta

Been noodling on this topic for a bit (some recent discussions here and here and this great one), and here are what seem to me to be the most important questions to ask with regard to the SN rewards system:

what do users like and want
what user behavior helps the SN community to flourish
what behavior can support the viability of SN as a business
how can we identify that behavior using available data
could we make new data available that would reveal it better
are there features that would encourage users to exhibit more of this useful behavior, or perhaps, decrease counter-productive behavior
how do we design incentives in accordance with the previous answers

All of these issues are more complicated than they appear at first. For instance, a ton is buried in the first one -- what does it mean to flourish, anyway? You could define that in a bunch of ways:

total volume of zapping
some measure of surface area of numbers of users x total zaps
number of daily posts
diversity of daily posts
summation of the relevance of posts over time, e.g., evergreen-ness
complexity of created content according to semantic analysis or link structure
stacker activity over time

and many of these have their own measurement problems, in turn. A minimum circularity is probably inescapable, e.g., part of the definition of flourishing has to do with supporting SN as a going business concern, but not all of it; and part certainly has to do with users getting what they want, but not all of that, either, and anyway, which users? And what about potential users who are not here, but could be, as a kind of natalist utilitarianist argument?

It will be interesting to consider March Madness in terms of questions like these. I don't have instincts about it, partly bc I'm not following the leaderboard or other stats, but at a minimum it seems the reward distribution should look different? Although perhaps it won't look as different as we think it will? What if the top X zappers earned exactly what they would have earned anyway? That's analysis waiting to be done; similarly with the other metrics.

Any aspiring computer science, economics, psychology, or sociology PhDs in the audience? DM me to riff on dissertation topics. Or, I guess, reply in the comments :)

view all related items

377 sats \ 1 reply \ @DarthCoin 28 Mar

I cannot give you the best answer for all those questions because I do not have the right information data. I think only SN founders and admins can give you the best answer to this post.

What I can give you is MHO about how I see SN until now (being one of its first users).

I see SN like were the faucets in the early days of Bitcoin

People were receiving a lot of sats for almost doing nothing. But that was a method to bring new users easily. In those early days, was needed more testing and people using it, in order to understand it. SN is doing something the same, but in small quantities. As I always said, be wise with zapping. Not everything on SN merit a zap, especially a large amount of sats.

SN is designed to be a better reddit but instead of upvote/downvote will be with sats. Users MUST feel "the pain" when they zap 1 sat or 1000 sats. Upvotes on SN are the same as on reddit, one user = one vote. But SN is adding something new, the incentive of amount of sats, not just the vote. The amount of sats is ONLY the zapper appreciation to the OP, for its effort. But that is not affecting the "ranking" of that post. Is more like a V4V reward.
All those thinking that SN is an "assmilking" cow... they are wrong. SN is a fair reward system, driven only by each stacker "sats power".

In all this MSM experiment I zapped more than usual. Why? Because I wanted to see how the algo is working or not. But usually I am very careful with how much and to who am I zapping.

43 sats \ 0 replies \ @elvismercury OP 28 Mar

I like the analogy to an old-timey faucet -- SN as faucet is a fascinating way to think about it.

847 sats \ 13 replies \ @k00b 28 Mar

You could define that in a bunch of ways:

What's fun is that we don't use any of these metrics directly. The only thing we directly measure are aspects of spending behavior.

Your spending on items that other people who have historically spent well are also spending on.
Your creation of items that people who have historically spent well are spending on.

It is a Keynesian Beauty Contest. But the whole point of the algorithm, at least as it's currently constructed, is to incentivize creating and surfacing a consensus of beauty.

In future algos, I'm interested in exploring more localized consensus, where rewards come from cliques of people you trust/people who trust you. As long as I'm fantasizing, I also want territories to be able to experiment with reward rules of their own making because I think any person is unlikely to have the single best idea for every territory.

139 sats \ 2 replies \ @elvismercury OP 28 Mar

As long as I'm fantasizing, I also want territories to be able to experiment with reward rules of their own making because I think any person is unlikely to have the single best idea for every territory.

You could imagine allowing territories to provide their own distribution function -- you give access to some kind of daily data bundle (computed by SN) and the territory can, in return, provide you with a JSON blob of reward allocation over users, the results (presumably) of their computation over that bundle. (This seems sort of ornate, vs the territory owner providing a function that SN uses to compute, but security risks inherent in that seem prohibitive.)

27 sats \ 1 reply \ @k00b 28 Mar

We could run the code in an isolated environment. It takes some input of a specified format and returns an output in a specified format.

27 sats \ 0 replies \ @k00b 28 Mar

Maybe I'll finally get to create a visual programming language. Or maybe programming with natural language will be good enough by then.

137 sats \ 3 replies \ @elvismercury OP 28 Mar

I love the KBC metaphor.

Casting the reward function in that light then pushes the design space down a level -- it becomes a different grounding problem -- you define beauty based on the consensus of the users you've attracted. But you're still attracting users. Those are still choices that the definition of beautify is grounded in.

You could find that you've made choices that result in a curious definition of beauty, for good or ill.

48 sats \ 2 replies \ @k00b 28 Mar

You could find that you've made choices that result in a curious definition of beauty, for good or ill.

Ah yes, it's all a bit semi-supervised in that I did a lot of data labelling early on to kick start the whole thing.

The way I've thought about this so far is that so long as the group of "people who have historically spent well" can change, ie the historical input to the algorithm is a recent window of time, the definition of beauty can progress.

27 sats \ 1 reply \ @k00b 28 Mar

Also, if it wasn't clear, by "data labelling" I mean I would zap a lot, ie I am the history before history.

100 sats \ 0 replies \ @elvismercury OP 28 Mar

Yahweh: I am who I am Popeye: I am what I am @k00b: I am the history before history

109 sats \ 4 replies \ @0b1CoinObi 28 Mar

The cliques are already here alive and kicking, you just have to spend any amount of time in certain territories to see it first hand. Certain members merely have to post any old comment to get zapped, they've already got their followers. Which in turn allows them to game the system. The mere mention of cliques makes me feel sick, it's disclusionary, the try hards will try harder to suck up to the clique members and how will people get a trust rating? Territories experimenting with their own reward rules will just make it the wild west. Those with the best reward incentives will get the most traffic forcing other territories to follow suit. It's another race to the bottom.

233 sats \ 1 reply \ @k00b 28 Mar

Certain members merely have to post any old comment to get zapped, they've already got their followers. Which in turn allows them to game the system.

Building relationships and earning people's favor is a pretty virtuous "game" isn't it? What alternative game would you like to see played?

It's another race to the bottom.

I think it's a race to the top, ie "the best reward incentives will get the most traffic." Should the worst reward incentives get more traffic? I know you don't mean that but I'm not sure what you do mean.

15 sats \ 0 replies \ @justanumber 29 Mar

earning people's favor is a pretty virtuous "game" isn't it?

I call it sucking up, brown nosing, kissing arse, virtue signalling 😜😂 But, seriously, that's the groundings for a social credit system. I think what this showing me is that myself and a handful of other people on here are truly ungovernable 😂

15 sats \ 0 replies \ @justanumber 29 Mar

The word clique filled me with dread too

0 sats \ 0 replies \ @elvismercury OP 28 Mar

I get what you're saying, but this seems like one of those things where everybody agrees that 90% of the internet is shit, but they can't agree on which 90%. Which is the difference btwn your description being total pathology (for me, this would be a btc maxi echo chamber screaming the catechisms at each other) and being the sign that things are working out great (a bunch of people who say interesting things being perpetually rewarded for saying interesting things).

Territories experimenting with their own reward rules will just make it the wild west. Those with the best reward incentives will get the most traffic forcing other territories to follow suit. It's another race to the bottom.

Oh wow -- for me, that sounds like a race to the top! An ecosystem of innovation.

0 sats \ 0 replies \ @justanumber 29 Mar

more localized consensus, where rewards come from cliques of people you trust/people who trust you

How would that work?

It sounds like it has the potential really accentuate what already happens naturally, create exclusivity, plus alienate and make it difficult for new people joining.

48 sats \ 5 replies \ @Coinsreporter 28 Mar

IMHO, users like daily rewards more than the contest thing. I am personally averse to it and like the contest more.

42 sats \ 4 replies \ @elvismercury OP 28 Mar

Sometimes what users like, or think they like, is contrary to what it causes them to do. Will be interesting to see what's knowable based on March.

122 sats \ 3 replies \ @k00b 28 Mar

Something I only have a theory for and has a bunch of confounders is that the number of comments per post is at an all time high this month. I think without daily rewards to recoup costs, and post costs being relatively high, people are engaging more with others to recoup costs.

210 sats \ 1 reply \ @elvismercury OP 28 Mar

I keep almost talking about this -- not wrt March, but generally. I'm like, surely posts that generate lots of discussion are good posts, and people who help a lot of discussion occur are great users but then I'm like, this would be literally the easiest metric in the world to game without some extraordinarily sophisticated network analysis, or, if that's not true, I'm too stupid to solve it in a less sophisticated way.

So instead, I hold my tongue.

27 sats \ 0 replies \ @Undisciplined 28 Mar

I totally agree with all of that. Hopefully, someone smarter than us figures it out.

0 sats \ 0 replies \ @Coinsreporter 28 Mar

Yes, that's what I also feel.

452 sats \ 7 replies \ @Voldemort 28 Mar

I am going to spew some thoughts. Hopefully some of them will be meaningful.

What is the goal of zapping, in the first place?

My thoughts that might be way off base, are that zapping was aimed at incentivizing meaningful contributions to the site (rather than trolling), incentivizing people to post here over some other site, and enabling people to reward others for posting things that they like.

Those are rather surface level. In a deeper sense, zapping ought to make the question "what do users like and want" irrelevant. Zapping serves as a quasi-price mechanism that aggregates this information. Where zaps go is what the users want. But this is where it could get muddied, especially once the site gets bigger - rewarding people who give zaps acts as a subsidy that is compounded by giving zaps to people who zap popular posts early. Subsidies blur market signals.

Here is an example: Suppose someone documented their entire process of developing a new recipe in weekly installments. Most people might not like it. I would. But why would my zaps that go towards a niche topic that I value earn me less than giving zaps to one of the many bitcoin posts that get more zaps because they are popular? In theory, the only reward I should get is the continuation of the material that I like. So if it sounds like I am whining about not getting more rewards, I am actually arguing for zero rewards from zapping.

I have more thoughts on this, but will respond with them in a comment below.

Rather than discuss what it means for a community to flourish, I will consider signals that a community is flourishing.

My first thought went to animals - people talk about animals released to the wild in this way. Signals of flourishing might be gaining weight, being accepted into a group, mating.

For communities of people, signals might be demand to be in the group, membership fees increasing (could be actual fees or something like home prices), people not leaving.

It is fascinating that so many people on SN are vehemently opposed to "shitcoins" because if SN had its own currency you could use the price of the currency as a proxy for flourishing. You'd also not get arbitrary increases in prices for things like territories. Tokenomics matter. But I digress.

Typically, membership fees are a function of supply and demand. Prices increasing might indicate there are more new members than leaving members. So Maybe SN could measure something like how many accounts haven't been active in a certain number of days/weeks/months vs. new members.

I don't think new members need to be incentivized, however. If they are signing up they likely already realize the value.

I outlined what I would do for daily rewards in a different post. Basically it is a function of how many sats their post received compared to others in the same topic at similar times. Potentially also a function of the max sats a single person gave it. In my view, the goal of daily rewards should incentivize me to give people more rewards if they have content I like. Right now it is to give rewards to things other people like. Unfortunately I don't like what most people on this site like so that means

Even if someone agrees with me they are not incentivized to give me sats if they think others will disagree. In turn, this disincentivizes me from making more contributions.
If I want to max my rewards I am forced to give sats to things I don't like.

I want to make sure it is clear I am not complaining. I am a fan of the site. I am merely speaking in hypotheticals. IE, I still create posts even if they don't get zapped much.

310 sats \ 6 replies \ @elvismercury OP 28 Mar

Many interesting thoughts in this post, but lunch is not infinitely long.

Even if someone agrees with me they are not incentivized to give me sats if they think others will disagree. In turn, this disincentivizes me from making more contributions.

This is the bizarre beauty and horror of Keynesian beauty contests, as I linked in the writeup: the fallout from being rewarded by what other people are rewarded for. Here's a thought experiment:

Imagine that SN were visited by a user who posted My Little Pony fanfic. The current SN readership suggests these posts would not be richly rewarded, although who can say. And this fictional user would reasonably say: well, they don't appreciate my genius here, fuck this noise.

Is that a bad outcome? Under what circumstances would it be a bad outcome?

Being slightly less ridiculous, I have posted things that were a ton of work that were not well-rewarded. I admit to feeling kind of annoyed, like my internal model that I am a valued member of this place had been violated. So then what?

Do I tune my interactions, in order to win better rewards?
Do I persist, even though people don't value those kinds of things, in the hopes that I will discover an audience for them and they will eventually come to be valued?
Do I say: well, this isn't that place for that.

You can make a case for any. Which one you do depends on what you're after, I suppose -- the old "find the right objective function" thing @Undisciplined just mentioned.

360 sats \ 1 reply \ @Voldemort 28 Mar

Is that a bad outcome? Under what circumstances would it be a bad outcome?

It is a suboptimal outcome if the incentive model disincentivizes a subset of users who deeply yearn for my little pony content to not zap the author.

If there are no users who have such yearnings, it is not suboptimal.

Anyone interested in a betting pool for when we see a my little pony territory?

Being slightly less ridiculous,

ridiculous things lead to interesting thoughts. In all seriousness, I wonder what it means for a community to be able to sustain a sub-community of bronies. Most people view that behavior as completely untenable, so its likely not possible in smaller communities. Or are bronies so unique that they can't be defined as a sub-community, but rather a blight on existing communities that are simply tolerated so as to not break the law?

0 sats \ 0 replies \ @elvismercury OP 29 Mar

Anyone interested in a betting pool for when we see a my little pony territory?

I'm guessing it won't be till the oft-hypothesized re-jiggering of how much territories cost, unless some rich Brony truly has a passion project. Which I am here for, if any such person is lurking.

In all seriousness, I wonder what it means for a community to be able to sustain a sub-community of bronies.

Good question -- there must be some sufficient size where it can be viable on a general-interest site, or a site with another interest that is not Bronies. I've been pleasantly surprised by the activity my territory ~mostly_harmless gets, which is not as fringe as MLP, but which is also not even about any particular topic, more a way of being.

And yet people show up! So maybe whatever the carrying-size required for Bronies is, we're at it?

27 sats \ 3 replies \ @Undisciplined 28 Mar

I have posted things that were a ton of work that were not well-rewarded.

Do you have an example, because that probably means I missed it?

20 sats \ 0 replies \ @elvismercury OP 29 Mar

Oh you sweetheart. I guess it pays to whine :)

20 sats \ 1 reply \ @elvismercury OP 29 Mar

Thinking principally of this one in the Capital Formation series.

To be clear, you should be playing me the world's smallest violin, because I earned almost 4k sats on that. But it's also true that I worked on it every day for a week, let's say ... 3 hours of work to compile, edit, etc. [1]

I did / do those things (and everything here) mostly for myself, but I can't lie and say it didn't take the wind out of my sails a little bit.

[1] If it doesn't seem like it contains 3 hours of work, that's testament to me being ... obsessive about some stuff.

21 sats \ 0 replies \ @Undisciplined 29 Mar

Looks like I had already given it my standard great post zap and it's up to 13k now.

I think there's a small cadre of people who put posts over 10k and it's somewhat random if they'll see your post quickly.

Darth mentioned to me that he circles back through and gives big zaps to posts that generated a lot of good comments.

48 sats \ 5 replies \ @0xbitcoiner 28 Mar

how do we design incentives in accordance with the previous answers

@Undisciplined suggested an interesting and brilliant idea, I can't find the original post, but it was something like this:

reward distribution

Daily
Weekly
Monthly
Yearly

The @sn team has already expressed concerns about the difficulty of implementation, but they have not ruled it out entirely.

460 sats \ 4 replies \ @Undisciplined 28 Mar

Thanks. I don't recall if I ever made the point about how that fosters something closer to evergreenness. If you make a post that people keep coming back to throughout the year, that isn't really rewarded under the daily system, but you might end up with a big payout at the end of the year.

On the evergreenness topic specifically, I wonder if people could get credit for zapping content late, as well as early. Whatever decay function they use, could be modified with another function that grows very slowly over time.

160 sats \ 3 replies \ @elvismercury OP 28 Mar

Earliest and latest zaps to quality content is an interesting idea, a bimodal KBC -- dissertation topic right there.

48 sats \ 2 replies \ @Undisciplined 28 Mar

It would make for an interesting mechanism design project. Of course, you pointed out that the elephant in the room here is properly specifying the objective.

There's no way to evaluate a mechanism, unless you know exactly what SN is trying to accomplish.

42 sats \ 0 replies \ @elvismercury OP 28 Mar

Annoying how it always comes down to that; and surprising how often people don't seem to realize that it comes down to that.

21 sats \ 0 replies \ @justanumber 28 Mar

unless you know exactly what SN is trying to accomplish.

Exactly. What is SN is trying to accomplish?

128 sats \ 2 replies \ @grayruby 28 Mar

The rewards system will never be perfect. I don't mind it generally, I am just not a fan of the way the MSM is structured.

127 sats \ 1 reply \ @k00b 28 Mar

Our plan for April is something way less top heavy which was influenced by your feedback.

166 sats \ 0 replies \ @grayruby 28 Mar

My feedback? You guys must be really desperate. Haha.

Looking forward to seeing what's next.