Game theory illustrated by an animated cartoon game
76 comments
·May 19, 2025hayst4ck
alickz
>If you tolerate bad faith, you ask for more bad faith behavior.
Yet some bad faith behavior comes from misunderstanding, so to be completely intolerant of perceived bad faith behavior would be the same as entering into the endless "tit-for-tat" cheating loop mentioned on the site
"I'm intolerant of your (perceived) bad faith (because you're intolerant of my (perceived) bad faith)"
Works when perceptions aren't muddied, but that is rarely the case
>The lesson is clear, that if you want a world you want to be a part of, then you must become powerful and choose to use that power for good.
I agree broadly, but that logic is also the cause of a lot of cheating
"I cheat because they cheat"
"I need nuclear weapons because they have nuclear weapons"
Tit-for-tat is a great strategy, but it's not a _dominant_ strategy
I think that's the most important point the website makes
Gravityloss
In sports it doesn't work with the individual, either tit-for-tat or just succumbing to bad behavior.
Instead there are central organizations created by the players and teams together, and they agree on rules and establish referees and processes so on.
xp84
I feel like your comment sums up many life lessons. It's also the reason I stopped believing in extreme leftism as I matured. I think as young idealists we think "Surely if we just gave the government a lot more tax money, they'd solve the problems we all say we care about" and "Surely if you just give freely to whomever says they're in need, they won't cheat the system!"
Later I started to see the patterns where government spends most of the money lining the pockets of the well-connected, and then on the micro level how many people take advantage of any method of unjust enrichment, given the chance, and you start to desire much more accountability from all parties. And yes, things like, say, exhaustive income verification to qualify for benefits definitely hurts those who are playing by the rules the whole time. It's the cost of having trashy individuals in society who exploit everyone relentlessly.
gitremote
> I feel like your comment sums up many life lessons. It's also the reason I stopped believing in extreme leftism as I matured. I think as young idealists we think "Surely if we just gave the government a lot more tax money, they'd solve the problems we all say we care about" and "Surely if you just give freely to whomever says they're in need, they won't cheat the system!"
As an older leftist, what you describe looks more like the liberal (centrist) view. Actual leftists think the current system is broken. Cops are not acting in good faith using tax payer money, so we should defund the police departments to some degree and divert those funds to social programs to help the vulnerable. Politicians and Congress people (including Democrats) in office are not acting in good faith and should be barred from trading individual stocks to prevent insider trading.
xp84
> Politicians and Congress people (including Democrats) in office are not acting in good faith and should be barred from trading individual stocks to prevent insider trading.
No objection from me there.
> defund the police departments to some degree and divert those funds to social programs
I will just say that I have zero faith that the "social programs" will do much besides serving as jobs programs for people with degrees that would otherwise not be useful to society. Meanwhile, less cops and less support for the idea of policing leads to more of the exact same opportunistic bad behavior that I was talking about. There's a statewide experiment on this topic that's been going on for years, called California.
ericrosedev
It looks like Copykitten is the sweet spot to me, with a focus on keeping miscommunication to a minimum. I wonder where, between 0% and 1%, there is a noticeable deviation, because I find the idea of the Copykitten more nuanced than the copycat, but the copycat always wins somewhere between 0 and 1.
dinfinity
> Later I started to see the patterns where government spends most of the money lining the pockets of the well-connected
It is definitely not most of the money in developed countries. Being cynical is fine, but don't throw the baby out with the bathwater. Governments do enormously important and valuable things with most of the money. Just take a look at where money is spent in the budgets of developed countries.
irjustin
> The lesson is clear, that if you want a world you want to be a part of, then you must become powerful and choose to use that power for good.
This particular lesson we've seen play out time and time again at a state level. Europe, USA, Singapore and even China.
The best thing that happened was corruption was either kept out or effectively eliminated.
When leaders are bad actors, it's pretty obvious, but an explicit example is the Philippines in the 50's-60's and how its economic powerhouse was squandered away.
Today, Vietnam is in the process of cleaning house and showing dividends while Indonesia really could really do amazing stuff if it go its act together.
AdieuToLogic
> This is one of my favorite things on the internet, but it focuses on the positive side of the story which is that groups of people cooperating can defeat a bunch of people who cheat each other. That's a pleasant message.
This is also a reasonable definition of a functional democratic society.
> If you believe in personal agency and personal responsibility and don't believe in magical thinking ...
No man is an island, entire of itself; every man is a piece
of the continent.[0]
> The lesson is clear, that if you want a world you want to be a part of, then you must become powerful and choose to use that power for good.Or perhaps a different lesson is that if you want a better world to exist, choose to live it. Sometimes there will be those which do not share this vision and actively work against your choice.
The question then becomes:
Do you allow bad actors to determine who you are
or is who you are orthogonal to what bad actors do?
Note that "power" has no relevance to the answer of this question.vacuity
> Or perhaps a different lesson is that if you want a better world to exist, choose to live it. Sometimes there will be those which do not share this vision and actively work against your choice.
So you need the power to abide by your choices and not others'. I think the colloquial understanding of "power" is somewhat different than I want to convey. Power is potential, the ability to affect or effect, control[0]. Violence can be powerful, because a dead foe can't hurt you. Nonviolence can be powerful, because a dissuaded foe can't hurt you. Who you are determines what you do. What you do reflects who you are. The path you are walking on, whether to your design or not, requires power to prevent straying.
AdieuToLogic
> Power is potential, the ability to affect or effect, control[0]. Violence can be powerful, because a dead foe can't hurt you. Nonviolence can be powerful, because a dissuaded foe can't hurt you.
I was going more for the "be the change you want" perspective than what perceived power can accomplish. Ultimately, power is an illusion and consumes rather than enables.
> Who you are determines what you do.
Technically this is correct. However, it ignores choice and the value of personal growth found in honest introspection.
> What you do reflects who you are.
Very true. And if one chooses to do something different than what would have been done as an earlier version of oneself, does that not reflect growth?
> The path you are walking on, whether to your design or not, requires power to prevent straying.
Choice is not power, even though it is powerful, unless the assumed one of the multitude of definitions cited is the first:
The ability to do or undergo something.
Which is incongruent with the implication of what I originally quoted: The lesson is clear, that if you want a world you want to
be a part of, then you must become powerful and choose to
use that power for good.
So which is it?marci
Becoming powerful might be the easy part
https://www.youtube.com/watch?v=rStL7niR7gs (The Rules for Rulers / CGP Grey)
hayst4ck
That's one of my other favorite things on the internet. The discussion of it is very high quality, too: https://www.youtube.com/watch?v=ILvD7zVN2jo
Getting into a position of power requiring submission to the power hierarchy that grants power is a very keen insight. It means you are subject to all the corruptive forces everyone in those positions of power are also subject to. People will let you into positions of power if you wear their chains.
"Then you must become powerful and choose to use that power for good" is almost directly from CGP Grey's own analysis/those two videos.
Pournelle's iron law of bureaucracy is another way of looking at it: https://www.jerrypournelle.com/reports/jerryp/iron.html
The man in the Arena speech tickles these topics and is unfortunately now more relevant than ever: https://en.wikipedia.org/wiki/Citizenship_in_a_Republic
Terr_
> getting cheated or cooperating with people who act in bad faith is what creates the cheating
That reminds me of some game-theory stuff, a relatively simple "tit-for-tat with forgiveness" approach does pretty well. It matches a lot of our intuition too: Be nice, punish betrayal, but not too disproportionately.
Not the thing I was thinking of, but found this fun little interactive presentation that goes into some other factors/approaches: https://ncase.me/trust/
martin82
You have perfectly described the downfall of western democracies, who are oh so tolerable.
Even Plato described this already in The Republic.
frollogaston
The board game "Diplomacy" teaches lessons like this, also how fairness works even between unequal powers.
tiffanyh
Vertasium has a great video talking how Tit-for-Tat (Copycat) wins as a strategy (and how there was a math competition that proved it as well)
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
gowld
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
The ncase.me game goes into the forgiveness part
elliotto
The success of tit-for-tat is a common misunderstanding; it is only the most successful within the makeup of Axelrod's tournament. The OP explains that the strategy is entirely dependent on the environment, and tit-for-tat may not always be the optimal strategy.
gcanyon
No conversation on game theory is complete until someone brings up Golden Balls, and in particular this amazing moment (warning: terrible audio quality). https://www.youtube.com/watch?v=S0qjK3TWZE8
bjconlan
I must say this is amazing. The psychology and manipulation makes me realize how poor I am regarding trust even when the other side is pushing for some unconfirmed equilibrium.
In the game I acknowledge that I was aligned to the "Simpleton" strategy (before it was outlined). Looks like simpleton might actually be applicable in a more general sense too which is a little disheartening.
ChicagoBoy11
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
Another great book on the subject is The Joy of Game Theory by Presh Talwalkar.
netbioserror
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.
gowld
Once weakness in those tournaments is that every interaction has the same stakes, so it doesn't model the possibility of "con artists" who behave generally at low stakes, but cheat at high stakes.
xpe
Varying the stakes would be interesting.
My understanding is this: in these computerized multi-player repeated games, each agent has no knowledge of previous games nor identities of other players. So every game a particular agent plays must follow the same algorithm. This means tournament results can easily be tabulated by tallying performance of each strategy type.
dfltr
I remember that as well, but only because it's mentioned in book 8 of The Expanse.
NullHypothesist
Found this ~10 years ago. Still one of the best things i've ever come across on the internet.
gota
Similar - this is where I found out about Komiku's music; I must've heard to thousands of hours of this while writing code
yubblegum
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
Are you talking about "one-off games" vs "repeated games", or something else?
Repeated games are part of "game theory"
yubblegum
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.
dfltr
I don't know if this is the official term for it, but that just sounds like metagaming[1], i.e. incorporating knowledge of the opposing player (or of trends among a group of opposing players) into how you play the game.
[1]: https://en.wikipedia.org/wiki/Metagame#Competitive_gaming
michalsustr
This also called opponent modeling. Used in e.g. poker, so you can maximise your earnings.
https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=oppo...
gametheory87
[dead]
krisoft
I disagree with the phrasing right the first question it asks.
It says "Let's say the other player cheats, and doesn't put in a coin. What should you do?" and the two buttons are "cheat" and "cooperate". But if the other player doesn't put a coin in then not putting in a coin is not "cheating". It is simply not playing the game with that person.
Cheating would be where you say you will put in the coin (or have already put the coin in) but not doing so.
Tsarp
This is really the best ways to explain basic game theory to anyone. It wasn't until some excellent Profs in Grad school that I finally understand some of the most basic concepts, but would always find it hard to explain to friends and family.
I also think game theory is one of the most important philosophies/life-lessons to understand as you go through life and this is an excellent resource to get people started on the basics.
0cf8612b2e1e
Radiolab had a story about this idea that I enjoyed.
https://radiolab.org/podcast/104010-one-good-deed-deserves-a...
xpe
Radiolab's overdramatic style has the ability to ruin even my favorite topics.
fracus
This web site was so enlightening. I imagine it could often be difficult to put theory to practice as in real life there are so many variables to consider and many of them we can't quantify. Some of the demos illustrated complete opposite results due to a 1% change in the miscommunication variable. But I will say the 3 main takeaways they give you were more general and certainly applicable to real life.
This is one of my favorite things on the internet, but it focuses on the positive side of the story which is that groups of people cooperating can defeat a bunch of people who cheat each other. That's a pleasant message.
Unfortunately, I think the corollary is much more important. What this clearly shows is that on an extremely fundamental level, getting cheated or cooperating with people who act in bad faith is what creates the cheating. If you tolerate bad faith, you ask for more bad faith behavior.
If you believe in personal agency and personal responsibility and don't believe in magical thinking, then it shows on a very mathematical level that your own weakness, the ability for someone to take advantage of you without consequences, is what creates defection rather than cooperation.
The lesson is clear, that if you want a world you want to be a part of, then you must become powerful and choose to use that power for good.