Philosophy professor at Princeton brags about idiotic in-class "game theory" experiment, gets owned in Twitter comments

[Dan Druff]

Sometimes mediocre people think they're exceptional, and then display their mediocrity for all to see.

That occurred today to Lara Buchak, who is a philosophy professor at Princeton.

Look at this bizarre and pointless "game theory experiment" she brags about conducting each year:

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692182967211147432

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692189303449772498

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692363836790104278

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692360415487922534




People reading this were perplexed. What is the point of this? What does it teach? What is the downside to the present class voting for 1 point for themselves, as there's no penalty for taking the selfish/greedy route? Does this teach people to be good for the sake of being good? Maybe you can call this a psychology experiment, but it's definitely NOT game theory! There's no game at all, in fact! Nothing the current students do will alter the outcome of whether or not they're getting 3 points possibly voted from the prior year!

Strangely enough, a small tweak could have actually made this interesting. She could have made it to where you vote to get the bonus point for yourself or vote for 3 points for next year's class, BUT you'll only receive the 3 from the prior year if your class also voted to give 3 to next year. But here it's incredibly simple and stupid. You're either voting to give 1 point to this year's class (including yourself), or 3 points to next year's class -- presumably strangers you don't know.

As you see above, someone on Twitter suggested this, but she confirmed that was not the way it worked. And then another guy suggested a different tweak:

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692354091169538118

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692366079098601931

Embedded Tweet

https://twitter.com/LaraBuchak/status/1692368590001852463


She also doesn't seem to understand the concept of "Nash Equilibrium", which is explained pretty well on this page.

This is because she classifies her little experiment as a "two player game", but it actually isn't at all. The current year's class has zero effect on the previous year's class which is affecting them! There is also no downside to the selfish/greedy choice.


Not only is it weird that a Princeton philosophy professor would come up with such a pointless "game", but it's hilarious that she's bragging about it on the internet. I can only imagine the looks on the faces of her students, who think they must be missing something, when this is proposed.

[lol wow]

princeton record exchange is pretty sick

[Entropy]

Speaking about the decline in academia. I watched the first half and this stuck with me for a few days..
https://www.youtube.com/live/ExneZYvehYM?feature=share
Unz showed his findings to experts like Steve Pinker and even they were shocked.


Traffic is rising fast, bookmark it.
https://www.unz.com/

[BetCheckBet]

Seems pretty obvious this is not prisoners dilemma but more so self interest vs community growth for lack of a better term. Thinking it this way. Someone tells you that you can have five free dollars or get 30 free dollars to give to a stranger. There is no way to rational decide it comes down to personal values. On top of it there is an element of gilt because it’s possible you’ve already benefited but are not aware.

Honestly though no idea why this post got so much attention. It’s a meh issue

[Dan Druff]

Seems pretty obvious this is not prisoners dilemma but more so self interest vs community growth for lack of a better term. Thinking it this way. Someone tells you that you can have five free dollars or get 30 free dollars to give to a stranger. There is no way to rational decide it comes down to personal values. On top of it there is an element of gilt because it’s possible you’ve already benefited but are not aware.

Honestly though no idea why this post got so much attention. It’s a meh issue

These are all elements of psychology, though, not game theory.

As a psychological experiment to demonstrate human greed versus selflessness, it could be interesting. But there's no strategic element to it.

In your example, what would be most interesting would be multiple versions of the same question.

I would take $5 over having $6 awarded to some random, but I would forego the $5 if $10,000 were awarded to a random, even if I wouldn't find out who they are and I couldn't take credit for it.

Where is the point where I'd forego the money to give more to a random? I don't know, and for everyone it's different.

I might run such an exercise in a psych class I'm teaching, but I would never call it game theory.

The tweet got attention because a Princeton professor seemingly didn't understand what game theory actually was.

[BetCheckBet]

Seems pretty obvious this is not prisoners dilemma but more so self interest vs community growth for lack of a better term. Thinking it this way. Someone tells you that you can have five free dollars or get 30 free dollars to give to a stranger. There is no way to rational decide it comes down to personal values. On top of it there is an element of gilt because it’s possible you’ve already benefited but are not aware.

Honestly though no idea why this post got so much attention. It’s a meh issue

These are all elements of psychology, though, not game theory.

As a psychological experiment to demonstrate human greed versus selflessness, it could be interesting. But there's no strategic element to it.

In your example, what would be most interesting would be multiple versions of the same question.

I would take $5 over having $6 awarded to some random, but I would forego the $5 if $10,000 were awarded to a random, even if I wouldn't find out who they are and I couldn't take credit for it.

Where is the point where I'd forego the money to give more to a random? I don't know, and for everyone it's different.

I might run such an exercise in a psych class I'm teaching, but I would never call it game theory.

The tweet got attention because a Princeton professor seemingly didn't understand what game theory actually was.

Fair point that it’s not game theory for sure as this an independent event that has seemingly no relationship to anyone else’s decision .

[podcast]

I dunno, seems like a prisoner's dilemma to me (other than the 2 player thing).

Definition: The prisoner's dilemma is a paradox in decision analysis in which two individuals acting in their own self-interests do not produce the optimal outcome.

No matter what the other player does, you should confess in the prisoner's dilemma, resulting in both players confessing and both getting a medium amount of prison time instead of both cooperating and getting a short sentence.

Kind of the same thing in this, no matter what the other class did you are better off taking the 1 bonus point for yourself. This results in every class taking the 1 point for themselves, whereas if they had "cooperated" every class would get 3 bonus points. Doesn't really matter that the class giving you the 3 points is different from the class you are giving the 3 points to.

[BetCheckBet]

Optimal strategy is prisoners dilemma varies based off of opponents choices. In simplified format tit for tat is optimal.

Believe it or not my fave paper I read in grad school was on prisoners dilemma. It’s talks about noise in communication and that sometimes in real life when it looks like someone is cooperating the message that comes across is that they’ve defected. This results in a negative tit for tat retaliation pattern screwing both parties over. Which means in reality where intent to act is not always the same as the final action you need to adjust tit for tat to allow a small amount of defectors ( be a little nicer and accepting of perceived selfish actions)

[BetCheckBet]

A good example of real life prisoners dilemma is druff handling jewdonk. Druff largely applied a tit for tat retaliation method but also have a slight variation of extra cooperating to mend the relationship and avoid an endless retaliation loop. Someone like hutmaster who cooperated too much was exploited by jewdonk defections.

You could also argue that jewdonks true intent was to cooperate more often but the introduction of copious amounts of alcohol results in communication noise of producing a higher number of defections than intended.