Tit-for-Tat with occasional Forgiveness and occasional Defection (abuse of trust) often performs better than pure Tit-For-Tat, especially in the presence of random error.
Tit-For-Tat falls into "permanent mutual Defection" tar-pit when playing against a Tit-for-Tat-like opponent that Defects once (perhaps in error) and is non-Forgiving.
Humans are pretty good at repeated-game theory, intuitively.
frollogaston · 27m ago
The ncase.me game goes into the forgiveness part
ChicagoBoy11 · 2h ago
The Evolution of Cooperation is one of the best non-fiction book I've ever read. Through basic algebra it lets you in on appreciating such a deep and profound idea.
wrboyce · 2h ago
Another great book on the subject is The Joy of Game Theory by Presh Talwalkar.
yubblegum · 2h ago
So this little game actually amplifies the distinction between "game theory" and (let's call it) 'relationship theory'. In the former you rely on strategy. In the latter, you rely on established trust.
You run the game once and at the end you are given 'character' headsup on the participants. Next time around playing the same game, you know who is who.
p.s. In effect the distinction can be generalized as 'depth of priors' for the 'bayesian game'.
gowld · 2h ago
Are you talking about "one-off games" vs "repeated games", or something else?
Repeated games are part of "game theory"
yubblegum · 1h ago
Repeated games. Think relationships. Once you have an accurate grasp of the 'character' of your playmate you can approach optimal results.
For example, the character with the flower hat is a 'detective'. We can assume the initial encounter is a coin-toss choice for her and the rest of her choices determined by 'character'. Of course even her first choice is 'in character' (she is 'testing') but if you know her, even if she starts off with a 'cheat' on her first choice, you start off with a 'cooperate'. After that, there is little mystery as to her choices. Or consider the 'grudger'. If for whatever reason you end up choosing 'cheat' once, you know they will never 'forgive' you. etc.
No comments yet
NullHypothesist · 4h ago
Found this ~10 years ago. Still one of the best things i've ever come across on the internet.
xpe · 2h ago
If you want the take-aways, click the next-to-last navigation circle on the bottom of the screen. I won't paste the spoilers here, because I think it would detract the experience.
adi_lancey · 4h ago
this is pretty cool, very nice explanation of concepts I hadn't touched since undergrad econ
daveguy · 3h ago
The simulation didn't cover the problem that the US is having right now -- intentional miscommunication. Unfortunately, there's a reason some countries employ warehouses full of trolls and propaganda spreaders. If you are losing the "game" spreading chaos will level the playing field. It will take the people of those countries to stop their leadership from acting in bad faith before things improve.
This is also the reason some social media outlets have become dystopian hellscapes.
frollogaston · 28m ago
https://ncase.me/crowds/ is kinda about this. Not intentional misinformation exactly, just about how ideas like anger are amplified when you have an overly connected social network. There's supposedly a sweet spot.
I remember that in older formal game theory tournaments, a punish-once single-retaliation strategy won out. It was unconditional on copying, simply that if the opponent cheated, you cheat back once and then forgive until another cheat. Another form of Golden Rule approach. But I think those tournaments were under simpler conditions than the one here.
I like the incorporation of miscommunication, and being able to change the parameters.
searine · 1h ago
This makes me think about The Dark Forest and the chain of distrust that results from lightspeed communications at stellar scale. The universe is harsh...
badmonster · 53m ago
very cute!
gowld · 4h ago
playing time: 30 min • by nicky case, july 2017
clueless · 4h ago
spoiler alert: is copycat strategy the reason united states feels like it's becoming more like the rest of the world, more authoritarian?
yubblegum · 2h ago
The implicit in your q is a negation of the notion of 'social classes'. If you accept the notion of social class (in the political/economical sense) the possibility remains that it is 'theatre' to hoist "authoritarianism" globally over the under classes.
https://radiolab.org/podcast/104010-one-good-deed-deserves-a...
https://www.youtube.com/watch?v=mScpHTIi-kM
---
This seems like a nice rebuild of the math competition performed years ago (as talked about in the video link above).
Direct link to that part of the video: https://youtu.be/mScpHTIi-kM?si=yzZxyeYw4cJA-i37&t=583
Humans are pretty good at repeated-game theory, intuitively.
You run the game once and at the end you are given 'character' headsup on the participants. Next time around playing the same game, you know who is who.
p.s. In effect the distinction can be generalized as 'depth of priors' for the 'bayesian game'.
Repeated games are part of "game theory"
For example, the character with the flower hat is a 'detective'. We can assume the initial encounter is a coin-toss choice for her and the rest of her choices determined by 'character'. Of course even her first choice is 'in character' (she is 'testing') but if you know her, even if she starts off with a 'cheat' on her first choice, you start off with a 'cooperate'. After that, there is little mystery as to her choices. Or consider the 'grudger'. If for whatever reason you end up choosing 'cheat' once, you know they will never 'forgive' you. etc.
No comments yet
This is also the reason some social media outlets have become dystopian hellscapes.
Also, that's one of my favorite websites ever.
https://ncase.itch.io/wbwwb
"WE BECOME WHAT WE BEHOLD a game about news cycles, vicious cycles, infinite cycles"
I like the incorporation of miscommunication, and being able to change the parameters.