Hacker News new | past | comments | ask | show | jobs | submit login
Interactive introduction to game theory and trust (ncase.me)
405 points by GranularRecipe on Aug 11, 2021 | hide | past | favorite | 62 comments



I've seen this once. But just now I noticed something:

>the fewer "repeat interactions" there are, the more distrust will spread.

Doesn't this explain quite well why big internet communities devolve into a cesspool? I'm not behaving like a jerk (or worse, a moron) if I know people around me will recognise and avoid me, but as the odds of that happening get closer to zero, it becomes more advantageous to fling shit.


Not only that, but it becomes harder to recognise consistent norms. If you post an assertion and I respond with a 350-word cited explanation of how you are well-intentioned-but-factually-incorrect, am I

A. Behaving like a know-it-all jerk?

B. Behaving like a contributing member of a curiosity community?

Depends on the community norms and how well we understand them.


This is exactly why every month or so when I check out Facebook to see what's going on I don't respond to things I see like I would here, no matter how much I might want to for a bit.

Even if a similar topic comes up on FB and I've just posted a large comment here with multiple sources against a point of view, on FB at most I'd probably post "Eh, I'm not sure that's quite right, at least according to this. <link>". Almost nobody there is actually interested in having their point of view challenged or learning more about the thing they're posting, from my experience, so that's about all you can do in many situations without things just devolving or people ignoring you.


Yep. And it's totally fine for different forums to have different jobs!


So it’s sort of like the difference between a small town and a big city. In a city one is semi anonymous and can do lots of things that one otherwise couldn’t do in a small town where someone will know who you are and places social pressures which affects behavior.


I think you’re spot on, but a counterexample seems to be Twitter, where both low- and high-profile accounts engage in shit-flinging because it evokes reactions, and the algorithm doesn’t discriminate between positive and negative engagement.


Because of the character limit, I think Twitter is more conducive to emotional outbursts (both positive and negative, but considering that it takes a lot more emotional energy to be positive than negative we see more negativity) due to the quick nature of it. It only takes a couple seconds to pound out 250 chars.


My emotional outbursts can easily be thousands of words. There's no way to actually express emotions on Twitter. I suspect that's part of the reason for all the images and memes and links to articles.

Edit: and bitter arguments.


The character limit introduces a lot of misunderstandings.


This is my experience with Twitter. There isn't enough space to write out something in a way that minimizes the chances of misunderstandings. However, I think a big part of the problem is that those explanations are needed in the first place. Why do we so easily jump to awful interpretations? That's how you end up with rage that quickly spreads and even worsens as context for what the person said is lost or the quote isn't repeated at all but instead just the bad-faith interpretation is passed around.

The effects of that go beyond just the awful emotional impact. People get fired for that kind of controversy. (I hesitate to call it "cancel culture" because of the implications in that term of who's canceling who).


>Why do we so easily jump to awful interpretations?

I think the root cause is failure to assess certainty levels. For those people, "I know X" and "I don't know, but some weak evidence leads me to believe X" are exactly the same thing.

So they draw a conjecture based on what someone else said, assume that conjecture is correct, and vomit it as incontestable truth. Other assumers join in, take the anecdote as incontestable truth, and you have the rage spreading like a wildfire.

[Incidentally this kind of behaviour was what made me ditch Reddit. Not the power-tripping mods, or the corrupt admins, but the idiotic users. I'm glad I don't use Twitter, it seems like they fall in the same bag.]


Another thing that the game emphasized (based on the 1984 book “the Evolution of Cooperation“) is how misunderstandings increase mistrust. The Web is full of misunderstandings.

The other takeaway is to foster Win-win situations as opposed to zero-sum games.


Yes that and the more immediate absence in the particular interactions: a lot of people wouldn't be as hostile and disrespectful in person for fear of losing teeth.


I think you have a point, but the type of interaction is important too. IRL, you have to trust based on the fear that if you deviate, others might minimax you in the future. On social platforms, no one can really minimax you in a tangible way, hence all the cesspool behavior.


Not just internet communities. This applies to cities, or even countries, as well.


I’ve played this and it’s fantastic. Do play it to the end.

I found it extremely insightful and their observations connected long standing disparate dots for me. It’s as if a big jumbled up puzzle suddenly clicked into place.


Played this sometime in the past and had a lot of fun. I did not want to, given the 30 minute play time estimate, but I ended up playing the anxiety demo [0] on the same site, and spending about an hour total.... it was incredibly informative and eye-opening.

[0]: https://ncase.me/anxiety-demo/


I just finished playing this and I wasn’t particularly impressed. It seemed to be saying that your anxiety is controlling you when you choose not to use, Facebook, Instagram, Twitter, and Tinder. I don’t believe that using social media apps is a necessary or meaningful thing to do with your life. Nor do I think it’s fair to imply that people who choose not to engage on those platforms are suffering from anxiety disorders. In fact it often seems more the opposite way around.

However, I’m not entirely sure that this was the intended insinuation. Yet it’s hard to read anything else into it since its about a girl trying to engage with her friends on her phone across these various platforms and being stopped by the ‘anxiety wolf’. She even gets her health dinged for sharing a news story, which to some extent could be viewed as engaging with reality.

Engaging with reality can most assuredly cause anxiety, however I don’t really view this as improper. I don’t believe that mental health should be thought of as a framing or mindset one has to the extent that it frames reality away. I.E. I may feel better if I pretend covid is not real though I am not making a healthy choice if I do.


> It seemed to be saying that your anxiety is controlling you when you choose not to use, Facebook, Instagram, Twitter, and Tinder.

This is a really odd reading to me. Especially if you play through the scenario multiple times, you see that the Anxiety Wolf controls Human regardless of what you do: even if you turn down the party invitations or if you never interact with the news story on Twitter at all.

It's not that engaging with social networks is bad, it's that Anxiety Wolf views every interaction and outcome as dangerous. If you eat bread, you're eating junk food. If you don't eat bread, you might have an eating disorder. If you don't share a story, you're disengaged. If you do share a story, you didn't fact check it enough. If you eat lunch alone, you're going to die alone because you can't make friends. If you try to get a date, your date's a serial killer. You can't win; there isn't a move that Anxiety Wolf will be happy with other than being in a constant state of panic all the time or curling up in a ball and crying, because Anxiety Wolf is scared of everything.

The point of the first section of the game is that Anxiety Wolf has an unhealthy relationship with risk analysis, which ultimately leads into the point of the second act of the game -- that the absence of Anxiety Wolf also causes Human to have an unhealthy relationship with risk analysis in the opposite direction, to the point where Human starts ignoring imminently dangerous situations.

> I.E. I may feel better if I pretend covid is not real though I am not making a healthy choice if I do.

Covid is actually a pretty good example here. Anxiety Wolf wants you to get vaccinated (which is a very good idea, because you should be scared of catching Delta). But Anxiety Wolf was also the voice early on in the pandemic telling people to dip their fruit in bleach in case somebody in the grocery store had touched it. It's important to figure out how to distinguish between those two suggestions.


My play through definitely implied that social media was good and that not using it was bad. Seems to me that anyone could play this and decide their own interpretation of "anxious wolf" which is exactly my criticism of it. Things that contain more depth are less subjective in their interpretations. I would even take aim at the idea that "fear of being a bad person" is something we over-estimate, I often wish more people had this fear (like vaccines because it's good).

It reminds me of the whole Colbert paradox, where he polls equally well with all sides of the political spectrum despite engaging with politics daily. No matter what you believed you felt he believed the same things as you.

This is not that I disagree with your overall view here - I don't, people should absolutely apply logic in reasoning to their rick/reward calculations rather then going on pure gut feeling - however I just do not believe this game really goes into any of this in a manner fitting to the complexity of the subject. But really, how could 6 min flash game do that?


> Seems to me that anyone could play this and decide their own interpretation of "anxious wolf" which is exactly my criticism of it.

Right, I think this is very intentional. The author even interrupts the game at one point to tell the player that they should choose the choices that personally give them the most anxiety. My playthrough ended up focusing in on a lot of the "productivity" lines; are you wasting time doing X, is it responsible to engage with people if you don't have a clear set of goals about where it will end, that kind of thing.

> Things that contain more depth are less subjective in their interpretations.

I'm not sure about this. A vision of anxiety that focuses on "this specific action causes stress" would be less subjective, but I think it would also be less accurate. Anxiety (particularly clinical anxiety) is often all-encompassing, when anxiety gets really "fun" is in the moments where you're simultaneously scared of every possible choice at the same time, including inaction. I've been fortunate enough to avoid the worst of these experiences in my life, but from what experiences I do have, the feeling the game invokes of being scared almost just for the sake of being scared, as if fear itself is some kind of defense against the world falling apart -- that does resonate with me, and I think the game does a good job of capturing that.

The other thing that I suspect that some people online have latched onto with this game is just the process of sitting down and visualizing yourself and asking "what am I scared of right now." It's not universal, but for a lot of people I think disassociation can often be a really powerful cognitive technique both for self-analysis and for staging self-interventions during some types of panic attacks, and disassociation is not always something that comes intuitively to everyone.

It's interesting to see the different reactions though. :) It's funny because if anything my criticism of the game is that near the end it gets a little too specific and too prescriptive for my tastes about how to address fears. I like the idea of having a healthy relationship with anxiety, and I do think that idea captures something meaningful; but I don't think that's something that can be set up in only one conversation, and I don't think the motivations/solutions the author lays out for fear are broad enough to capture what that process will or should look like for everyone, the game feels overly optimistic about how hard that process is.

But like you said, it's a short game. I think it suffers a little bit from not having the necessary time available to avoid an abrupt ending that sort of implies treating anxiety just requires having an emotional breakthrough that solves everything.


I’m going to go out on a limb and say that I do not think with disagree very much on this topic, other than our subjective impressions of the game - liking and/or disliking. I do think that we may be talking on slightly different levels of abstraction. Which is tough for me to get right with short online comments. Thanks for engaging on this, though. I enjoyed reading your thoughts very much.


How's a news story reality, and social media not? News stories (especially the sorts you'd share on social media) are usually virtual, detached from reality.


I was not expressing a critical view of news media simply because it is not contextual in this situation. The “game” is actually very non-specific as to what the “news” is - or anything else for that matter. So, I was simply applying something like covid or global warming to it in order to show that the generalized premise of the game does not hold as a very insightful view of human being’s relationships with their anxiety.


The outcome of the various games depend on the way rewards are distributed. If the reward for bad behaviour is high, obviously the bad guys are going to win. This is evident in the 5 step "The Evolution of Distrust". Try reducing the rewards for bad behaviour and increasing for good behaviour :)

So, moral of the story is: Reward good behaviour.


Other morals:

Being "nice" (copycat) is not the same as being a "sucker" (cooperator).

Suckers don't make things worse just for themselves, but for everyone around them, by encouraging bad behaviour.

Even then, people make mistakes, so don't hold grudges for too long. They won't lead you very far.


Just based on the game, suckers are not really the problem. If you turn down interactions for example even the copycat (or weak copycat) looses against the all cheat. So it's not that the always coorporate creates the bad situation.


That's because the copycat, in the first interaction, works exactly the same as the cooperator. And the outcome of that first interaction becomes more important if there are less of them.

The problem I see only pops up when there are more interactions - when the player knows the other will cheat, but still decides to cooperate.


> So, moral of the story is: Reward good behaviour.

I wish. This is an oversimplification.

In many real-world scenarios bad behavior wins unless the rules of the game are changed to encourage cooperation.

E.g. by social norms or legislation or contract


I think the real situation is that every time the system changes, something will try to 'game' it via another strategy.

The morality depends on where/who you were when the system/rules changed.


But you don't know ahead of time who will cooperate, and who will cheat. The rule was "if you spend 1 coin, the other player gets 3 coins". If you change it to "if you spend 1 coin, the other player gets only 2 coins if they cheated", the game isn't fair anymore. Or am I misunderstanding something?


Depends on the game. Lot of games in the real world are traps. Maybe both sides dont/cant see all the info, dont have enough knowledge/skill, have weaknesses/biases, not enough endurance etc etc etc

In those games, best move is to exit the room. Cause no reward/punishment system work.


And punish bad behaviour, reduce miscommunication / improve coordination.


The link really highlights the importance of having win-win situations if both people cooperate (i.e. positive-sum game).


Game theory, just like essentially everything in math, physics and probability, and cs is about adjoints, norms, and fixed points https://github.com/adamnemecek/adjoint/

Nash equilibrium is a fixed point.


That pages needs a lot of work before you try presenting to people who aren't into category theory. I know it's not but to someone who doesn't know the jargon it'll just looks like crankery. Someone who doesn't already know what you're thinking isn't going to be able to get anything out of your comment and that braindump in relation to the OP.


That’s not the experience I’ve had. I have shown it to people who were not into category theory who got it fine. Also it goes beyond CT, CT unlike say linear logic isn’t closed under norm.


"Peace broke out" is one of my favorite openings to the Monty Python Funniest Joke in the world sketch. This is well done!


This is awesome. Did it around a year ago. Eye opening. I irge everyone to check it out.


What happens if you add another level of meta to it, and the object of the game is to avoid complete randomness/undecidability but also for the game to go on for as many iterations, ideally with the lowest amount of mistakes(/entropy?) and defections/cheats.

i.e stuff is happening, there's a lot of predictability but no easily determinable stabilisation endpoint?


In the sandbox, if you change the payoff of both parties cheating to -5, the angelic cooperator seems to win in most simulations.

In a way, this makes sense: if the penalty when both parties cheat is harsher than the upside of swindling your partner, "trust at all costs" wins out!

But on a societal level, I suppose this means that if you want to optimize for do-gooders, you should punish failure harshly, no matter which party is responsible...


Interesting: In the "change the payoffs" version, if you set the cooperate/cooperate to +2 (or +1) and change the cheat/cheat punishment to -3 or -4 or higher, you get a situation that alternates between a greater or lesser number of cheaters and cooperators. The copycats get eaten instantly.


I cannot recommend "The evolution of cooperation" (Robert Axelrod), the source for the game, enough. It is written by a journalist, so it reads well, about a Math doctorate applying the theory in other areas. Richard Dawkings vets it for biology and writes the foreword in the book.


This has been posted a couple of times, also very recently. But never got a lot of discussion.

https://hn.algolia.com/?q=https%3A%2F%2Fncase.me%2Ftrust%2F


I'm certain it was on one of such previous posts that I found it (and played it)

And, as chance would have it, this is where I first heard Komiku's "It's time for Adventure!" (the song that plays in the intro). It has since been what I listen to when programming when silence is not an option


I can't seem to find the exact song, can you link it?


If I recall correctly that's number 13 "Bleu":

https://freemusicarchive.org/music/Komiku/Its_time_for_adven...

Check out the other Volumes too, they're all great


Same for his other game "We Become What We Behold"

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...


This person is fabulous... Followed and loved their games for nearly half a decade now.


Absolutely agreed - their game about spacial repetition is my go-to conversion tool for friends and family: https://ncase.me/remember/


Awesome! I've never read it and I just needed it, thanks for your recommendation.


What a great game.

It would be interesting to see what happens when the types „slide“ towards the next personality (copycat into copykittens into always cooperative) as they get possitive feedback (and the opposite for always cheater).


So beautiful. Love to see ancient wisdom proven mathematically. Bravo!


It is a variable that would make game theory much more applicable.


Gorgeous design and a brilliant way to show case game theory


Put a wrench in the machine to see what happens


what about sequence learning minmax cat?

mmm. or just play nice.


When I started programming, the first mistake I encountered was ignoring the spacing rules in Python, thus creating errors even though the code was correct.


there should be another character similar to the "copy kitten"

co-operate first, then if the other person cheats once then cheat back always


Isn't that the Grudger strategy that's introduced early on?


yes it is, thanks - I must have misread the description

I know a lot of Grudgers


Fool me once...or whatever Bush said.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: