Is zero-sum the default economic position?

[The once and future king]

[ QUOTE ]
Until John Nash came along some very bright people didnt realize that you could prove that non-zero sum but non-cooperative games could reach an equilibrium, and if there were no such equilibrium there would be no win-win solutions.

[/ QUOTE ]

You are aware that Mr Nash has recently refuted a great deal of his own theories including the one above.

[The once and future king]

Economics does not always work in the human scale.

I might lose my job to more competitive worker somewhere else, this ultimately in the long term makes the pie bigger, yet in the short to medium term my slice has decreased dramatically.

Also to get my slice back I may have to relocate or retrain. Both things that piss people off if they have strong attachments to their jobs and their communities.

Basically economics is on a macro scale so is beyond the immediacy of human perception. People will see zero sum if in the short to medium term they must suffer loss in some way for an abstracted unpercievable (but real) gain in the long term.

[hmkpoker]

[ QUOTE ]
Economics does not always work in the human scale.

I might lose my job to more competitive worker somewhere else, this ultimately in the long term makes the pie bigger, yet in the short to medium term my slice has decreased dramatically.

Also to get my slice back I may have to relocate or retrain. Both things that piss people off if they have strong attachments to their jobs and their communities.

Basically economics is on a macro scale so is beyond the immediacy of human perception. People will see zero sum if in the short to medium term they must suffer loss in some way for an abstracted unpercievable (but real) gain in the long term.

[/ QUOTE ]

Well said. The essence of the problem is that humans reason with heuristics, which can very easily allow them to see any gross benefit or detriment that they want to see.

For example, the benefits of Wal-Mart (cheaper products, increases in wages in the third world, more money for new business demand) and the detriments (local retail businesses tanking) are both huge aggregates. It's pretty much impossible to assign a number to either. This does not stop people from trying to qualitatively rationalize them heuristically, though.

PRO WAL-MART: Sure, some local retailers are losing out, but hundreds of millions of consumers are saving lots of money and living conditions in the third world are improving dramatically!

ANTI WAL-MART: Yeah, people are saving a few bucks and a bunch of chinese people are getting a few more pennies a day, but tons of small American businesses are going bankrupt!

ZERO SUM ATTITUDE: One one hand, American consumers are saving money and the standard of living is increasing in the third world, but at the same time, many American businesses are becoming outcompeted, so it all kind of evens out.

This really is how most people reason, and is perhaps best illustrated in the Simpsons episode where Principal Skinner and Edna Krabbappel banter back and forth to a local electorate about whether or not to raise taxes for more school funding.

[bobman0330]

[ QUOTE ]
[ QUOTE ]
There is no way to implicitly collude. Its impossible. No matter what I pick, you are going to make your selection. We cannot collude. It is essentially the same as saying "Wait until he makes his pick, look at it, and then make your pick."

[/ QUOTE ]

If you get on the game show, and surprise! it turns out that your best friend is the other contestant, aren't you and your best friend going to implicitly collude (even without discussing it beforehand)? If the answer to that question is "yes", then it is not true that there is no way to implicitly collude.

[/ QUOTE ]

This isn't collusion, this is cooperation. The reason you don't dick your friend over is that he's your friend and you derive utility from him being happy and because if you defect on him, he won't be your friend any more.

If you're a moral person and you just don't like cheating other people out of money, that's one way out of the PD. If there's an external enforcement mechanism, like contract law or broken friendship, that's another way.

[John Kilduff]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
There is no way to implicitly collude. Its impossible. No matter what I pick, you are going to make your selection. We cannot collude. It is essentially the same as saying "Wait until he makes his pick, look at it, and then make your pick."

[/ QUOTE ]

If you get on the game show, and surprise! it turns out that your best friend is the other contestant, aren't you and your best friend going to implicitly collude (even without discussing it beforehand)? If the answer to that question is "yes", then it is not true that there is no way to implicitly collude.

[/ QUOTE ]

This isn't collusion, this is cooperation. The reason you don't dick your friend over is that he's your friend and you derive utility from him being happy and because if you defect on him, he won't be your friend any more.

If you're a moral person and you just don't like cheating other people out of money, that's one way out of the PD. If there's an external enforcement mechanism, like contract law or broken friendship, that's another way.

[/ QUOTE ]

I think it is also implicit collusion, regardless of the reason for your motivation or the fact that he is your friend.

The highest EV in the situation is for you both to forgoe the chance at the extra hundred dollars in order to ensure that you both win a million dollars apiece. That requires implicit collusion or cooperation.

Say the other player on the game show was a stranger to you until you met and shook hands before the game started. Would you dick him out of a million dollars just to have an extra chance (not even a certainty) of winning a hundred dollars? This is a case where an extra hundred is relatively meaningless in comparison to the goal of winning the million. If two friends can silently choose the copperative and highly profitable strategy for both of them, so can two strangers who think it through.

Even if you are right on this matter, I still think there must be something fundamentally flawed in the so-called expert strategy if it returns a far lower EV that would the strategy employed by two morons playing randomly. It does return a MUCH lower overall EV so how do you resolve this paradox? Game theory against ANY opponent is supposed to protect your EV not destroy it. Yet in this case it fails miserably against an equal game theorist.

Thanks to you and vhawk for taking the time to discuss this with me.

[vhawk01]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
There is no way to implicitly collude. Its impossible. No matter what I pick, you are going to make your selection. We cannot collude. It is essentially the same as saying "Wait until he makes his pick, look at it, and then make your pick."

[/ QUOTE ]

If you get on the game show, and surprise! it turns out that your best friend is the other contestant, aren't you and your best friend going to implicitly collude (even without discussing it beforehand)? If the answer to that question is "yes", then it is not true that there is no way to implicitly collude.

[/ QUOTE ]

This isn't collusion, this is cooperation. The reason you don't dick your friend over is that he's your friend and you derive utility from him being happy and because if you defect on him, he won't be your friend any more.

If you're a moral person and you just don't like cheating other people out of money, that's one way out of the PD. If there's an external enforcement mechanism, like contract law or broken friendship, that's another way.

[/ QUOTE ]

I think it is also implicit collusion, regardless of the reason for your motivation or the fact that he is your friend.

The highest EV in the situation is for you both to forgoe the chance at the extra hundred dollars in order to ensure that you both win a million dollars apiece. That requires implicit collusion or cooperation.

Say the other player on the game show was a stranger to you until you met and shook hands before the game started. Would you dick him out of a million dollars just to have an extra chance (not even a certainty) of winning a hundred dollars? This is a case where an extra hundred is relatively meaningless in comparison to the goal of winning the million. If two friends can silently choose the copperative and highly profitable strategy for both of them, so can two strangers who think it through.

Even if you are right on this matter, I still think there must be something fundamentally flawed in the so-called expert strategy if it returns a far lower EV that would the strategy employed by two morons playing randomly. It does return a MUCH lower overall EV so how do you resolve this paradox? Game theory against ANY opponent is supposed to protect your EV not destroy it. Yet in this case it fails miserably against an equal game theorist.

Thanks to you and vhawk for taking the time to discuss this with me.

[/ QUOTE ]

No problem, and I am most certainly no game theory expert. I'm trying my best to argue my understanding of the concepts, and specifically what is at play in your game show and other PD-like situations, but I make no grand claims that I am absolutely correct. A lot of your points are well-taken.

BTW, I still think you should read that thread I linked earlier. [img]/images/graemlins/grin.gif[/img]

[bobman0330]

[ QUOTE ]
Say the other player on the game show was a stranger to you until you met and shook hands before the game started. Would you dick him out of a million dollars just to have an extra chance (not even a certainty) of winning a hundred dollars? This is a case where an extra hundred is relatively meaningless in comparison to the goal of winning the million. If two friends can silently choose the copperative and highly profitable strategy for both of them, so can two strangers who think it through.

[/ QUOTE ]

This is mostly a payoff question. For me, getting $1 million and my friend or even a stranger getting $1 million is better for me than getting $1,000,100. So that might not even be a PD situation.

[ QUOTE ]
I still think there must be something fundamentally flawed in the so-called expert strategy if it returns a far lower EV that would the strategy employed by two morons playing randomly. It does return a MUCH lower overall EV so how do you resolve this paradox? Game theory against ANY opponent is supposed to protect your EV not destroy it. Yet in this case it fails miserably against an equal game theorist.

[/ QUOTE ]

Well, this is why the PD is considered to be so important, and why a lot of the insights of game theory take people by surprise. The natural assumption is that behaving rationally should result in an optimal outcome. But sometimes, rationality can work against you.

But there's not a paradox. The expert strategy increases your EV by $100. Playing against an expert reduces your EV by $1 million. If you can "trade" your option of increasing your EV by defecting in exchange for your opponent's option, then it's a bargain. Signing an enforceable contract would do this. But if you can't, you're in trouble. Unless your opponent is altruistic and doesn't want to screw you over, he will if he's smart. Because his choice will not affect what you do.

[John Kilduff]

[ QUOTE ]
[ QUOTE ]
Say the other player on the game show was a stranger to you until you met and shook hands before the game started. Would you dick him out of a million dollars just to have an extra chance (not even a certainty) of winning a hundred dollars? This is a case where an extra hundred is relatively meaningless in comparison to the goal of winning the million. If two friends can silently choose the copperative and highly profitable strategy for both of them, so can two strangers who think it through.

[/ QUOTE ]

This is mostly a payoff question. For me, getting $1 million and my friend or even a stranger getting $1 million is better for me than getting $1,000,100. So that might not even be a PD situation.

[ QUOTE ]
I still think there must be something fundamentally flawed in the so-called expert strategy if it returns a far lower EV that would the strategy employed by two morons playing randomly. It does return a MUCH lower overall EV so how do you resolve this paradox? Game theory against ANY opponent is supposed to protect your EV not destroy it. Yet in this case it fails miserably against an equal game theorist.

[/ QUOTE ]

Well, this is why the PD is considered to be so important, and why a lot of the insights of game theory take people by surprise. The natural assumption is that behaving rationally should result in an optimal outcome. But sometimes, rationality can work against you.

But there's not a paradox. The expert strategy increases your EV by $100. Playing against an expert reduces your EV by $1 million. If you can "trade" your option of increasing your EV by defecting in exchange for your opponent's option, then it's a bargain. Signing an enforceable contract would do this. But if you can't, you're in trouble. Unless your opponent is altruistic and doesn't want to screw you over, he will if he's smart. Because his choice will not affect what you do.

[/ QUOTE ]

But thinking that way screws you also, and for enormous money, whenever you are against anyone thinking similarly. Game theory is supposed to protect your EV no matter who you are up against and regardless of the strategy they employ. I still don't see how two morons could (and would) outperform two game theory experts if the expert strategy is sound in this example, which leads me to think that the expert strategy is not considering everything it ought to be considering. An expert strategy employed by both players should not be dominated EV-wise by moronic or random strategy employed by both players, but it is. Furthermore why should a purely altruistic strategy outperform all other strategies when both parties employ it? I believe there is more than meets the eye to this example.

[vhawk01]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Say the other player on the game show was a stranger to you until you met and shook hands before the game started. Would you dick him out of a million dollars just to have an extra chance (not even a certainty) of winning a hundred dollars? This is a case where an extra hundred is relatively meaningless in comparison to the goal of winning the million. If two friends can silently choose the copperative and highly profitable strategy for both of them, so can two strangers who think it through.

[/ QUOTE ]

This is mostly a payoff question. For me, getting $1 million and my friend or even a stranger getting $1 million is better for me than getting $1,000,100. So that might not even be a PD situation.

[ QUOTE ]
I still think there must be something fundamentally flawed in the so-called expert strategy if it returns a far lower EV that would the strategy employed by two morons playing randomly. It does return a MUCH lower overall EV so how do you resolve this paradox? Game theory against ANY opponent is supposed to protect your EV not destroy it. Yet in this case it fails miserably against an equal game theorist.

[/ QUOTE ]

Well, this is why the PD is considered to be so important, and why a lot of the insights of game theory take people by surprise. The natural assumption is that behaving rationally should result in an optimal outcome. But sometimes, rationality can work against you.

But there's not a paradox. The expert strategy increases your EV by $100. Playing against an expert reduces your EV by $1 million. If you can "trade" your option of increasing your EV by defecting in exchange for your opponent's option, then it's a bargain. Signing an enforceable contract would do this. But if you can't, you're in trouble. Unless your opponent is altruistic and doesn't want to screw you over, he will if he's smart. Because his choice will not affect what you do.

[/ QUOTE ]

But thinking that way screws you also, and for enormous money, whenever you are against anyone thinking similarly. Game theory is supposed to protect your EV no matter who you are up against and regardless of the strategy they employ. I still don't see how two morons could (and would) outperform two game theory experts if the expert strategy is sound in this example, which leads me to think that the expert strategy is not considering everything it ought to be considering. An expert strategy employed by both players should not be dominated EV-wise by moronic or random strategy employed by both players, but it is. Furthermore why should a purely altruistic strategy outperform all other strategies when both parties employ it? I believe there is more than meets the eye to this example.

[/ QUOTE ]

But if you are up against a sample of 50% experts and 50% idiots, you are going to be getting 1 million half the time and 0 half the time. I'm going to be getting 1.0001 half the time and zero the other half. Why isn't my strategy better?

[AlexM]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Say the other player on the game show was a stranger to you until you met and shook hands before the game started. Would you dick him out of a million dollars just to have an extra chance (not even a certainty) of winning a hundred dollars? This is a case where an extra hundred is relatively meaningless in comparison to the goal of winning the million. If two friends can silently choose the copperative and highly profitable strategy for both of them, so can two strangers who think it through.

[/ QUOTE ]

This is mostly a payoff question. For me, getting $1 million and my friend or even a stranger getting $1 million is better for me than getting $1,000,100. So that might not even be a PD situation.

[ QUOTE ]
I still think there must be something fundamentally flawed in the so-called expert strategy if it returns a far lower EV that would the strategy employed by two morons playing randomly. It does return a MUCH lower overall EV so how do you resolve this paradox? Game theory against ANY opponent is supposed to protect your EV not destroy it. Yet in this case it fails miserably against an equal game theorist.

[/ QUOTE ]

Well, this is why the PD is considered to be so important, and why a lot of the insights of game theory take people by surprise. The natural assumption is that behaving rationally should result in an optimal outcome. But sometimes, rationality can work against you.

But there's not a paradox. The expert strategy increases your EV by $100. Playing against an expert reduces your EV by $1 million. If you can "trade" your option of increasing your EV by defecting in exchange for your opponent's option, then it's a bargain. Signing an enforceable contract would do this. But if you can't, you're in trouble. Unless your opponent is altruistic and doesn't want to screw you over, he will if he's smart. Because his choice will not affect what you do.

[/ QUOTE ]

But thinking that way screws you also, and for enormous money, whenever you are against anyone thinking similarly. Game theory is supposed to protect your EV no matter who you are up against and regardless of the strategy they employ. I still don't see how two morons could (and would) outperform two game theory experts if the expert strategy is sound in this example, which leads me to think that the expert strategy is not considering everything it ought to be considering. An expert strategy employed by both players should not be dominated EV-wise by moronic or random strategy employed by both players, but it is. Furthermore why should a purely altruistic strategy outperform all other strategies when both parties employ it? I believe there is more than meets the eye to this example.

[/ QUOTE ]

But if you are up against a sample of 50% experts and 50% idiots, you are going to be getting 1 million half the time and 0 half the time. I'm going to be getting 1.0001 half the time and zero the other half. Why isn't my strategy better?

[/ QUOTE ]

I've only read segments of this conversation, but it's recognized that the expert strategy is to not defect, right? Given that, in your scenario against 50% experts and 50% idiots you will make $1 mil 75% of the time, not 50%. 50% from the experts and 25% more from the randomly choosing idiots. On the other hand, if all the "experts" are choosing to take your option and defect, you only make $1 mil 25% of the time vs. half the idiots. Given these two options for the results of the expert strategy, it should be pretty apparant which is the expert strategy.