The most recent Veritasium video. I found the deep look into the history of the Prisoner's Dilemma amazing and wanted to make sure others got to see it.
The most recent Veritasium video.
I found the deep look into the history of the Prisoner's Dilemma amazing and wanted to make sure others got to see it.
I feel like I had learned most of this information already at one point or another, but he did a great job at helping me to actually understand it. The visualizations of the tournaments and the...
I feel like I had learned most of this information already at one point or another, but he did a great job at helping me to actually understand it. The visualizations of the tournaments and the explanation of the different strategies were excellent.
This was a fascinating video, and I feel like I can actually use this info in my real life, long term. I wanted to ask a question of those in this thread that are married/in a relationship: How do...
This was a fascinating video, and I feel like I can actually use this info in my real life, long term. I wanted to ask a question of those in this thread that are married/in a relationship:
How do you see the Tit for Tat strategy working in a relationship/marriage?
I have often heard the advice, "Don't go to bed mad," regarded as pretty good relationship advice, and this makes intuitive sense to me. However, employing a Tit for Tat strategy implies that I'm going to be "retaliating" against my partner some of the time, which seems much more risky than, for instance, in a career. I understand that it's important not to be a pushover, and so you eventually need to show them that you're not gonna keep taking their crap if they keep it up. But also I wonder if this strategy needs to be modified in a marriage? E.g. more forgiveness than elsewhere in your life? If so, I would guess it's as simple as say, 20-30% more forgiveness than normal to get the vast majority of overall possible benefit. But I'm really interested what perspectives others may have! Particularly those of you that are married.
I would also guess this is where attachment styles become important. The attachment style of your partner is somewhat analogous to what strategy they would employ in this game, so your optimal strategy will depend on what theirs is. And then, over time, it's ideal if you both can work together to modify your strategies to be more compatible with one another.
I've read the books and looked at the experiment and the code (though admittedly in undergrad roughly one billion years ago), and I can't stress enough that Axelrod is not giving advice on...
I've read the books and looked at the experiment and the code (though admittedly in undergrad roughly one billion years ago), and I can't stress enough that Axelrod is not giving advice on interpersonal relationships. It is important to remember in the prisoner dilemma is that within the context of the experiment, you are not on a team with anyone. Your goal is not to make things better for everyone, but specifically to make things better for yourself. The long term strategies involved are entirely selfish; if you cooperate, you make things better for yourself, and you don't worry about making things better for everyone, but cooperation is still better than not cooperating.
Marriage is different, or at least it should be. When you enter into a marriage, or any social relationship that is based on trust and love, then you should not be only worried about your own selfish "winning" - you should be considering the needs of others. As such, I think Tit for Tat is awful in terms of marriage; when you are married, you are a team, and you should generally consider your outcomes to be intertwined. You should adopt strategies together for the benefit of both of you (or all if you in a family).
I think it's really important to remember that this experiment is not a guideline for how to live an individual life and have friends; it's for adversarial games, where you are playing with points, and it is generally extrapolated to diplomacy, but it is not a guideline to have effective interpersonal relationships; humans are a lot more complex than the strategies talked about and every strategy was being selfish but that still resulted in cooperation being effective, because cooperation is a winning strategy even when being 100% selfish. With interpersonal relationships, it isn't effective to be 100% selfish, but if you are, cooperation is still a winning strategy.
Sorry if this was preachy, I had flashbacks to game theory class, and the prof expressly stating not to try to use this for interpersonal relationships because we'd all end up hating each other.
Ahh I see, glad I asked this question then! I didn't realize these experiments are very much not to be used in the context of relationships, hence why I found the tit for tat strategy very weird...
Ahh I see, glad I asked this question then! I didn't realize these experiments are very much not to be used in the context of relationships, hence why I found the tit for tat strategy very weird when thinking about a relationship situation.
fwiw, there's a whole nother type of game that can be modeled in game theory called a cooperative game, in which both parties get the same reward based on how successful the strategies they used...
fwiw, there's a whole nother type of game that can be modeled in game theory called a cooperative game, in which both parties get the same reward based on how successful the strategies they used each were together. That type of game can be used to model certain aspects of language/communication, for instance (which is why I know about it -- my background is in linguistics, not math). Imo cooperative game theory is probably more directly applicable to marriage than adversarial game theory.
The given game-theoretic scenario considers self-interested actors who cooperate only opportunistically. I should hope that this is not how your marriage works!
The given game-theoretic scenario considers self-interested actors who cooperate only opportunistically. I should hope that this is not how your marriage works!
Well, you don't, other than maybe a highly modified version of it. Tit-for-tat is a strategy for use in an untrusted domain. Not only that but it's an incomplete model, or at least a very...
Well, you don't, other than maybe a highly modified version of it. Tit-for-tat is a strategy for use in an untrusted domain. Not only that but it's an incomplete model, or at least a very sanitized model, of interfacing with the real world and all its complexities. Once you encounter more social complexities you need to refine it. As Robert Sapolsky pointed out in a video I recently watched, when working in more noisy or complex domains, what's more effective than just pure tit-for-tat is tit-for-tat with forgiveness, where signal-to-noise ratio is higher and you sometimes have errors in communication which can cause conflict.
So with more rules comes a more refined version of it which in your case would be incomplete though I would like to see what a more team-based model would be like such as with contractual partners. I think soon enough when extrapolated to more real-world scenarios, these interactions become complex enough that it's harder to make more formal rules regarding interacting in such domains as with marriage. Although as humans we tend to do a good enough job at that level since it's the domain we've evolved to deal with for millennia.
But I do think it's interesting how the more formalized tit-for-tat with forgiveness has been a less formalized basis for fairly stable civilizations for millennia and that possibly a good amount of human conflict is just due to communication errors involving non-optimal protocols of human interaction. Of course, due to imperfect protocols and methods of communication, this only covers a part of the problem with human conflict and there's a lot more to be said of humanity's desire to compete, even when communication is optimal. But as said by others, TFT is just too simple a method to use for marriage and doing so will have a non-optimal result for you so you'd be better off abstracting to more complex disciplines in softer sciences such as sociology, psychology and such.
I love Veritasium. It was really interesting to see the evolution of strategies through the generations. Very reminiscent of ecological patterns, once the prey is hunted to extinction, the...
I love Veritasium. It was really interesting to see the evolution of strategies through the generations. Very reminiscent of ecological patterns, once the prey is hunted to extinction, the predator population wanes. Thanks for sharing OP
I first learned about The Prisoner's Dilemma from William Poundstone's book. It's a good overview and also a biography of John Von Neumann. https://www.goodreads.com/book/show/29506.Prisoner_s_Dilemma
I first learned about The Prisoner's Dilemma from William Poundstone's book. It's a good overview and also a biography of John Von Neumann.
The same month that I finish the last few books of The Expanse series (where the Prisoner's Dilema becomes a central plot point), Veritasium makes a video about it. Talk about coincidences,...
The same month that I finish the last few books of The Expanse series (where the Prisoner's Dilema becomes a central plot point), Veritasium makes a video about it. Talk about coincidences, specially when I've never even heard of this particular dilemma before.
I don't have much knowledge about the history of the prisoner's dilemma, and I found this video useful for that purpose. It was an interesting watch. However, I think the presenter fails to...
I don't have much knowledge about the history of the prisoner's dilemma, and I found this video useful for that purpose. It was an interesting watch. However, I think the presenter fails to acknowledge its limitations, and overgeneralizes. This is especially apparent in the title.
Here are some factors that limit the scope of the strategy presented in this video.
First, there are -often very strong- power imbalances in the real world. They exist in day-to-day interactions, such as in the work place or schools, and other places like state departments or companies. These imbalances also are part of the political landscape, whether on personal or a bigger level. Some examples being, a minority against a majority person in a social situation, or a voter against a president or a party. Tit-for-tat isn't applicable in these cases, and there are many, many situations like this.
Another limitation is that there are many actors in the sociopolitical landscape that has different goals. Political actors, meaning humans, don't just aim to maximize a limited definition of self-interest. They have a vision for society and the future. For example, a religious fundamentalist and an anti-theist have heavily conflicting goals. Therefore, their interactions can't solely be described by means of maximizing self-interest. For an optimal strategy, you have to acknowledge the fundamental differences that lead to conflict, and act accordingly. For example, a political opponent may fake cooperation while undermining you in the long run. You have to be aware of this. Tit-for-tat can be a very dangerous weakness, and in hard conflicts, it can even be fatal.
Another limitation happens when there are multiple actors. You may try to follow the tit-for-tat strategy as an individual actor, but it can still fail. Spectacularly. This happens extremely often in interactions between workers and employers. A worker can refuse the employer's ridiculous demands. But if there aren't laws that are in place and enforced, the employer can just fire the worker and replace them with someone who will compromise.
These are the limitations off the top of my head. As a result of these and most probably more, tit-for-tat can't be generalized. It seems to me that it would be a much healthier to say that tit-for-that has a very specific, limited scope, but it lays the foundations for an approach that should be investigated. Therefore, it's more theoretical than general, and it shouldn't be presented as something generalizable. You may say that the video doesn't explicitly say this, however, my impression is that it's heavily implied in the video that this strategy is generalizable. Especially since the presenter doesn't mention these limitations, there's a highly likely chance it will be interpreted that way, and that the creator of the video intended it that way.
This brings me to a related issue. There are many populizers in the public light, who seem to present easy and profound solutions to complex issues. I think, intentionally or not, they are misleading, and they are often even harmful for understanding the topic at hand. For example, this video doesn't explain the nuances and limitations of this field, and as a result it provides an oversimplified worldview. Considering that I'm just somebody who thought of these limitations in half an hour, without researching this specific field, I'm surprised that these limitations aren't mentioned at all. Taking into account the financial incentive in marketing easy-to-digest solutions to people, I cannot help but have ethical concerns about it.
I agree with a lot of what you say here, especially about videos in this space. When teaching game theory it's really important to present its limitations when applied to real life and how...
I agree with a lot of what you say here, especially about videos in this space. When teaching game theory it's really important to present its limitations when applied to real life and how important every single factor in a game is to changing the result. It's irresponsible to present an optimal strategy in one game or type of game as though it's universally optimal, especially if you're trying to apply it to Real Life.
If you're interested in more straightforward "teaching you stats and game theory" videos, I recommend Primer. They present the basics well for a beginner but even as someone who's studied these things before in uni, they're great for brushing up and understanding why these things work this way on a more fundamental level. The cute animated blobs are also a plus.
The most recent Veritasium video.
I found the deep look into the history of the Prisoner's Dilemma amazing and wanted to make sure others got to see it.
I stumbled across an interactive version of what this video covers recently, The Evolution of Trust. Highly recommend it if you enjoyed the video.
I feel like I had learned most of this information already at one point or another, but he did a great job at helping me to actually understand it. The visualizations of the tournaments and the explanation of the different strategies were excellent.
This was a fascinating video, and I feel like I can actually use this info in my real life, long term. I wanted to ask a question of those in this thread that are married/in a relationship:
How do you see the Tit for Tat strategy working in a relationship/marriage?
I have often heard the advice, "Don't go to bed mad," regarded as pretty good relationship advice, and this makes intuitive sense to me. However, employing a Tit for Tat strategy implies that I'm going to be "retaliating" against my partner some of the time, which seems much more risky than, for instance, in a career. I understand that it's important not to be a pushover, and so you eventually need to show them that you're not gonna keep taking their crap if they keep it up. But also I wonder if this strategy needs to be modified in a marriage? E.g. more forgiveness than elsewhere in your life? If so, I would guess it's as simple as say, 20-30% more forgiveness than normal to get the vast majority of overall possible benefit. But I'm really interested what perspectives others may have! Particularly those of you that are married.
I would also guess this is where attachment styles become important. The attachment style of your partner is somewhat analogous to what strategy they would employ in this game, so your optimal strategy will depend on what theirs is. And then, over time, it's ideal if you both can work together to modify your strategies to be more compatible with one another.
I've read the books and looked at the experiment and the code (though admittedly in undergrad roughly one billion years ago), and I can't stress enough that Axelrod is not giving advice on interpersonal relationships. It is important to remember in the prisoner dilemma is that within the context of the experiment, you are not on a team with anyone. Your goal is not to make things better for everyone, but specifically to make things better for yourself. The long term strategies involved are entirely selfish; if you cooperate, you make things better for yourself, and you don't worry about making things better for everyone, but cooperation is still better than not cooperating.
Marriage is different, or at least it should be. When you enter into a marriage, or any social relationship that is based on trust and love, then you should not be only worried about your own selfish "winning" - you should be considering the needs of others. As such, I think Tit for Tat is awful in terms of marriage; when you are married, you are a team, and you should generally consider your outcomes to be intertwined. You should adopt strategies together for the benefit of both of you (or all if you in a family).
I think it's really important to remember that this experiment is not a guideline for how to live an individual life and have friends; it's for adversarial games, where you are playing with points, and it is generally extrapolated to diplomacy, but it is not a guideline to have effective interpersonal relationships; humans are a lot more complex than the strategies talked about and every strategy was being selfish but that still resulted in cooperation being effective, because cooperation is a winning strategy even when being 100% selfish. With interpersonal relationships, it isn't effective to be 100% selfish, but if you are, cooperation is still a winning strategy.
Sorry if this was preachy, I had flashbacks to game theory class, and the prof expressly stating not to try to use this for interpersonal relationships because we'd all end up hating each other.
Ahh I see, glad I asked this question then! I didn't realize these experiments are very much not to be used in the context of relationships, hence why I found the tit for tat strategy very weird when thinking about a relationship situation.
fwiw, there's a whole nother type of game that can be modeled in game theory called a cooperative game, in which both parties get the same reward based on how successful the strategies they used each were together. That type of game can be used to model certain aspects of language/communication, for instance (which is why I know about it -- my background is in linguistics, not math). Imo cooperative game theory is probably more directly applicable to marriage than adversarial game theory.
The given game-theoretic scenario considers self-interested actors who cooperate only opportunistically. I should hope that this is not how your marriage works!
Well, you don't, other than maybe a highly modified version of it. Tit-for-tat is a strategy for use in an untrusted domain. Not only that but it's an incomplete model, or at least a very sanitized model, of interfacing with the real world and all its complexities. Once you encounter more social complexities you need to refine it. As Robert Sapolsky pointed out in a video I recently watched, when working in more noisy or complex domains, what's more effective than just pure tit-for-tat is tit-for-tat with forgiveness, where signal-to-noise ratio is higher and you sometimes have errors in communication which can cause conflict.
So with more rules comes a more refined version of it which in your case would be incomplete though I would like to see what a more team-based model would be like such as with contractual partners. I think soon enough when extrapolated to more real-world scenarios, these interactions become complex enough that it's harder to make more formal rules regarding interacting in such domains as with marriage. Although as humans we tend to do a good enough job at that level since it's the domain we've evolved to deal with for millennia.
But I do think it's interesting how the more formalized tit-for-tat with forgiveness has been a less formalized basis for fairly stable civilizations for millennia and that possibly a good amount of human conflict is just due to communication errors involving non-optimal protocols of human interaction. Of course, due to imperfect protocols and methods of communication, this only covers a part of the problem with human conflict and there's a lot more to be said of humanity's desire to compete, even when communication is optimal. But as said by others, TFT is just too simple a method to use for marriage and doing so will have a non-optimal result for you so you'd be better off abstracting to more complex disciplines in softer sciences such as sociology, psychology and such.
I love Veritasium. It was really interesting to see the evolution of strategies through the generations. Very reminiscent of ecological patterns, once the prey is hunted to extinction, the predator population wanes. Thanks for sharing OP
I first learned about The Prisoner's Dilemma from William Poundstone's book. It's a good overview and also a biography of John Von Neumann.
https://www.goodreads.com/book/show/29506.Prisoner_s_Dilemma
The same month that I finish the last few books of The Expanse series (where the Prisoner's Dilema becomes a central plot point), Veritasium makes a video about it. Talk about coincidences, specially when I've never even heard of this particular dilemma before.
I don't have much knowledge about the history of the prisoner's dilemma, and I found this video useful for that purpose. It was an interesting watch. However, I think the presenter fails to acknowledge its limitations, and overgeneralizes. This is especially apparent in the title.
Here are some factors that limit the scope of the strategy presented in this video.
First, there are -often very strong- power imbalances in the real world. They exist in day-to-day interactions, such as in the work place or schools, and other places like state departments or companies. These imbalances also are part of the political landscape, whether on personal or a bigger level. Some examples being, a minority against a majority person in a social situation, or a voter against a president or a party. Tit-for-tat isn't applicable in these cases, and there are many, many situations like this.
Another limitation is that there are many actors in the sociopolitical landscape that has different goals. Political actors, meaning humans, don't just aim to maximize a limited definition of self-interest. They have a vision for society and the future. For example, a religious fundamentalist and an anti-theist have heavily conflicting goals. Therefore, their interactions can't solely be described by means of maximizing self-interest. For an optimal strategy, you have to acknowledge the fundamental differences that lead to conflict, and act accordingly. For example, a political opponent may fake cooperation while undermining you in the long run. You have to be aware of this. Tit-for-tat can be a very dangerous weakness, and in hard conflicts, it can even be fatal.
Another limitation happens when there are multiple actors. You may try to follow the tit-for-tat strategy as an individual actor, but it can still fail. Spectacularly. This happens extremely often in interactions between workers and employers. A worker can refuse the employer's ridiculous demands. But if there aren't laws that are in place and enforced, the employer can just fire the worker and replace them with someone who will compromise.
These are the limitations off the top of my head. As a result of these and most probably more, tit-for-tat can't be generalized. It seems to me that it would be a much healthier to say that tit-for-that has a very specific, limited scope, but it lays the foundations for an approach that should be investigated. Therefore, it's more theoretical than general, and it shouldn't be presented as something generalizable. You may say that the video doesn't explicitly say this, however, my impression is that it's heavily implied in the video that this strategy is generalizable. Especially since the presenter doesn't mention these limitations, there's a highly likely chance it will be interpreted that way, and that the creator of the video intended it that way.
This brings me to a related issue. There are many populizers in the public light, who seem to present easy and profound solutions to complex issues. I think, intentionally or not, they are misleading, and they are often even harmful for understanding the topic at hand. For example, this video doesn't explain the nuances and limitations of this field, and as a result it provides an oversimplified worldview. Considering that I'm just somebody who thought of these limitations in half an hour, without researching this specific field, I'm surprised that these limitations aren't mentioned at all. Taking into account the financial incentive in marketing easy-to-digest solutions to people, I cannot help but have ethical concerns about it.
I agree with a lot of what you say here, especially about videos in this space. When teaching game theory it's really important to present its limitations when applied to real life and how important every single factor in a game is to changing the result. It's irresponsible to present an optimal strategy in one game or type of game as though it's universally optimal, especially if you're trying to apply it to Real Life.
If you're interested in more straightforward "teaching you stats and game theory" videos, I recommend Primer. They present the basics well for a beginner but even as someone who's studied these things before in uni, they're great for brushing up and understanding why these things work this way on a more fundamental level. The cute animated blobs are also a plus.
I loved this video - and the lessons to be predictable and retalitatory seem well aimed for international relations.