By Philip Pilkington, a journalist and writer living in Dublin, Ireland
I prefer the company of peasants because they have not been educated sufficiently to reason incorrectly.
– Michel de Montaigne
The so-called ‘games’ depicted in game theory include some particularly disingenuous and evasive intellectual constructions. In times past people devised rationalisations in the most fantastic of ways, summoning up imaginary demons, angels, witches and warlocks to contextualise their psychological and existential difficulties. Today, some turn instead to mathematical games.
Long ago our ancestors constructed mythological systems wherein the world was a place inhabited by mystical and animistic forces. Those who summoned these forces generally lived in close proximity, giving immediate presence to magic and intrigue. Religious systems came later. These were more abstract and impersonal; the leader-figure no longer lived nearby and miracles became increasingly rare and valued.
After some time the more “civilised” and organised societies began to create more rational theologies. By fusing the key structural aspects of the religions with certain metaphysical principles, great systems were constructed that vied with one another for authority.
It was all imagination of course. All of these systems were just ways that men made sense of the world around them. These great mythic, religious and theological systems were just rationalisations used to organise and lend authority to otherwise arbitrary social institutions. The most important and probably most primal among these institutions was marriage, which we will have ample occasion to return to in what follows.
The Age of Enlightenment came next and with it the more fantastic products of the human imagination were dissolved into the bland homogeneity of Reason. The Enlightenment philosophers described social relations primarily in terms of equality – an obvious value judgment on their part. The world was not an equal or a harmonious place, but the Enlightenment philosophers pretended as if it were and concocted their ideological systems accordingly. To look back on the crass oversimplifications these men came up with – many of which continue to haunt us today – is an exercise in both entertainment and caution.
Those that did raise issues of conflict and change – like Hegel and Marx – quickly veneered over them with vast metaphysical and teleological systems that assured us all that society was moving from point X to point Y. The imbalances, they both thought, would be levelled by History – an unstoppable moving force that would reduce all to complete equality and homogeneity.
Today the great leveller is ‘The Market’. And neoclassical economics is the great metaphysical system that supports it. The Market, we are assured, is moving to a state of harmony and equilibrium – an equality in the sense that everybody will be content, of course (they will have their ‘utility maximised’).
What characterised the later metaphysical systems however – modern economics included – was that they banished the relations they were trying to describe to the margin, only to take them up by applying to them pre-constructed systems of reasoning. While marriage – which we take again because it is, as the anthropologists well know, the primal social institution – is absolutely central to social organisation it is completely ignored in the latter day systems. Whereas it occupied a central role in the older systems, it is only dealt with by the later systems by applying pre-established principles (marriage is a utility-maximisation operation; marriage will be unnecessary under Communism etc.).
But every now and then these more primal social relationships bubble up in the new doctrines. And when they do the effort to force fit them into the new construct is far from convincing.
Sexual Politics
Back to game theory. Why do we claim that it is so evasive? Because it nakedly displays the nature of the material it deals with and then denies that it is indeed this material that it is dealing with. Game theory sets up ‘games’ between certain personalities and then claims that the players have no personalities.
The ‘Battle of the Sexes’ game is perhaps the most outlandish example of this. The game, as we will see, is blatantly a psychological construction set up to assuage the theorist that he has a perfectly rational means with which to understand his/her interpersonal relationships. The theorist sets up a game that deals with a difficult interpersonal marital issue and then tries to solve it with various mathematical formulas. When this psychological dimension is pointed out to the game theorist they vigorously deny it, claiming that it is only a ‘pure’ mathematical game and says nothing about his or her conception of their own interpersonal relationships; this even though they have explicitly labelled this game ‘the battle of the sexes’ and allowed the participants’ relationship to be that of an intimate couple.
Here’s how the game works. We have a man and a woman. We assume that they are in a long-term relationship, most likely married. The man wants to go to a football game and the woman wants to go to the opera. They both want to go somewhere together, so they will not simply go their separate ways to their desired event.
Game theorists see two so-called equilibrium points. One is that the couple go to the football game; the other is that they go to the opera. The game theorists then tie themselves in knots trying to figure out probabilities and equitable solutions. They torture themselves when they find that in order for equilibrium to be obtained one ‘player’ has to lose. They introduce bizarre tropes to equalise this, one being a ‘randomising element’ – in other words the couple agree to flip a coin to see which event they will attend.
What a pile of garbage!
Let’s say I’m writing a script or a story wherein such a circumstance arises. Do I break out my game theory book and do some calculations in order to decide how to move the plot forward? Of course not. Instinctively – based largely on empirical experience and common sense – I would argue that there are only two probable outcomes to this so-called ‘game’:
(1) The person assuming the masculine role submits and the couple attend the opera.
(2) The person assuming the masculine role remains recalcitrant, the ‘game’ breaks down and the couple go home.
Yes, there is a certain amount of stereotyping undertaken here. Yes, some people negotiate their relationships differently. And the answer in real life probably depends in particular how extreme one partner’s aversion is to the other’s choice.
But when dealing with this sort of material we’re always dealing with stereotypes and ‘averages.’ Anyway, they can be modified later if we wish. For now, let us just try to tease out some of the underlying structure.
The ‘game’ will not stop there. Two new ‘scenes’ arise organically from both these outcomes:
(1) After the opera the couple return home and have an, erm, romantic evening.
(2) The couple have a temporary falling out. The person assuming the masculine role then make amends and possibly gives the person in the feminine role a gift.
Again, these do not follow from each other in some quasi-mathematical or pre-determined manner, but they form a coherent narrative. They are realistic and convincing representations of the so-called ‘game’.
Before moving on to the implicit rules structuring this so-called ‘game’ let us note a few things:
(1) As we will see, the ‘game’ is not based on a relationship of strict equality. We cannot interchange the two ‘players’ with one another without changing the outcome (for a given ‘player’). Their position vis-à-vis one another determines the potential outcomes of the ‘game’. Their positions are thus not static, but relational.
(2) The relational positions of the two players are implicitly structured by ‘external’ rules that must be taken into account before establishing how the ‘game’ will be ‘played’.
(3) The ‘game’ does not have to result in so-called ‘equilibrium’. It can easily break down. We can assume that an infinite and unknowable amount of variables ‘plug in’ to the ‘game’ in order to determine the outcome. Everything from the mood of the two ‘players’ to the nature of the events that they may attend must be taken into account. These elements are 100% unknowable prior to being empirically established and even then they are probably impossible to pin down.
Let us now move onto the implicit rules structuring the so-called ‘game’. Again, we will have to engage in some pretty heavy stereotyping here but such is the nature of the material. Different ‘games’ could easily be constructed taking into account different personal or cultural ‘rules’. Our exposition is merely to try to provide a deeper understanding of what really determines the outcome of such ‘games’ and why game theory is a pointless and absurd device for even trying to approach this sort of thing.
Rules of the Game
So, let’s say that our little narrative appears convincing. Yes, it is almost crassly stereotypical and would not apply to every couple under the sun, but let’s accept it as a ‘model case’ of sorts. What are the structuring principles that lead to the outcome of our so-called game?
When boiled right down this is implicitly an issue of debt. The key rule structuring the ‘game’ is that the player assuming the masculine role is always already in debt to the player assuming the feminine role. There is no need for us here to try to understand why this is the case, we will merely hint that it probably has to do with certain marital/sexual institutions and norms in Western culture (again, and we must stress this: these are not universal).
This ‘indebtedness’ is the key element that leads to the possible outcomes of the game. Either the player in the masculine position plays ‘by the rules’ or the game breaks down. This means that the ‘game’ starts out from a position of inequality rather than equality. Indeed, the implicit argument here is that any ‘game’ that is set up will always start out from a position of inequality and the very idea of a ‘game’ that begins from a position of equality is a mere fantasy. It is the dimension of debt – a burden of debt, as it were – that enforces this inequality at a structural or institutional level.
The ‘game’ thus has no ‘equilibrium’ position. Instead it should be seen as a relationship of force based upon given rules. The rules allow the game to potentially have a sociable but not immediately equitable outcome. Without the rules the ‘game’ would merely break down into bickering. Of course, if force is pursued outside the rules by the player in the masculine position the game will fall apart anyway. So, while the rules give the game a path through which it can be resolved in a fairly sociably acceptable way they do not pre-determine the outcome of the game.
Does this make the game a zero-sum game plain and simple? Does this mean that the player in the masculine position always ‘loses’? No. Recall that we have already alluded to what other outcomes the game would likely have in the medium-run. By reaching a socially equitable compromise – albeit one that they are not particularly pleased with – the player in the masculine position will likely get a ‘payoff’ eventually. It is also likely that said player is quite consciously aware of this.
Because the outcomes of the game run throughout time and are the basis of a relationship that cannot be said to be constantly ‘in balance’, it should not be thought of in terms of equilibrium. There is no static equilibrium. There are simply relations of force structured within social rules or norms. Again, this is not to imply that this is an institutional structure that is completely unstable. Instead it should be seen as an institutional structure that, being constantly out of balance, requires that the participant constantly negotiate various forces that threaten to destabilise the entire structure.
Of course, when you strip away all the fancy language the point being made is rather simple: relationships – all human relationships – never tend toward some sort of fantasy ‘equilibrium’ position. Rather, they are inherently mired in power-imbalances and compromise formations or rules that allow these imbalances to be navigated. Most of the time the compromise formations (or: rules) hold; but sometimes when the power-imbalances become too pronounced and the entire structure of the relationship breaks down.
Arrested at a very primitive level of psychological development game theory cannot even begin to approach this. Indeed, one would be better trying to derive lessons about the real world from a consultation with a fortune-teller or a shaman – they tend to be more subtle psychologists.
Game theorists will, of course, tell me that I have cheated and misrepresented them. “We only play ‘pure’ games,” they will protest. “These are not supposed to be applied to real people with real psychologies in real institutional settings.” If the reader cares to look over the series of ‘games’ that the theorists set up – replete as they are with characters ranging from star-struck lovers to nefarious crooks – and finds the game theorists’ defence to be a credible one then they can happily discount my argument.
For my part, I don’t believe this defence for a second. It is quite obvious what the game theorists seek to do, constructing their little islands of Reason amidst the great sea of indeterminacy and change in which we all exist. Perhaps, indeed, we all need something to hold on to amidst such chaos and tumult. But surely we can do a little better than this crude child’s play that the game theorists have concocted.
74 comments
Comments are closed.
http://annales.ensae.fr/anciens/n61/vol61-04.pdf
http://www.jstor.org/pss/20619937
Anywho I hate games as I have a problem with *severe empiric abnormality’s*.
Skippy… irrational emotions ba da ding! BTW whats in your cigar, all natural? JK.
i dont pretend to grasp all, but this is a fascinating attack on game theory assumptions that undergird (or do they) modern economics.
this commentary flows in line with the interview with the prof (emeritus) from univ of tenn.
i dont have the detailed knowledge to evaluate either set of claims, but i am the type of interested observer of economics who is delighted to see underlying assumptions challenged.
what thoughtful undergraduate econ student has not listened to assumptions about, e.g., all participating members in a market having equally perfect knowledge, and thought,
“in what world other than mine.”
I really hope this isn’t an attack on ALL game theory. The example posted is obviously crude and stupid. What about, for example, the Prisoner’s Dilemma (particularly in iteration)? There’s a perfect example of how the simulation leads to questions and narratives about pre-existing arrangements and trust.
PD games aren’t that great either, and used/referred to/relied on way too easily. They predict reasonably well how people in life/death/chance for imprisonment situations could reason (when considerations about what picking a specific option might mean for (future) social relations are pushed entirely into the background because of the direness of the threat to life and liberty that is at hand), but it in no way follows from that that they can therefore be used to model all situations.
I view it as more of a parable, personally.
Actually, iterated PDs work fairly well for a large range of situations. So does Free-Rider-Problem (tragedy of commons).
But of course, having a hammer doesn’t mean one should treat the whole world as a nail.
That’s the problem with the article – it saying “See, your hammer doesn’t work on my screw! Your hammer is wrong!”
The only problem with the “tragedy of the commons”, is that it doesn’t occur in real life, except under extreme circumstances (such as when countries start to subsidize overfishing). It was used as a propaganda device to justify a land grab by the nobility and aspiring capitalists.
See further, Boyle’s book The Public Domain, and
Garrett Hardin, The Tragedy of the Commons, SCIENCE, Dec. 13, 1968, at 1243.
(on contemporary land use) ROBERT C. ELLICKSON, ORDER WITHOUT LAW: HOW NEIGHBORS SETTLE DISPUTES (1991);
ELINOR OSTROM, GOVERNING THE COMMONS: THE EVOLUTION OF INSTITUTIONS FOR COLLECTIVE ACTION (1991).
I saw tragedy of commons occuring more than once even in the recent few months (say simple clearing-snow-from-your-pavement stuff to give one example). But I guess on how the definition of ToC you use.
Also, ToC is just a subset of FRP – and for other examples of FRP look no further than OECD (and how it doesn’t really work). Climate change and carbon emissions are another FRP, which is why we need stuff like Kyoto/Copenhagen.
As far as GT goes, we could also look at the auction theory , which is, arguably, the most practically relevant application of GT.
do you have agreements with your neighbors to clean the snow away in turn or is this agreement “tacit”? ;)
It’s not ‘your hammer doesn’t fit my nail’. It’s more like: ‘your hammer is a crude and silly instrument that looks like it was built with a caveman, go buy a new hammer’.
Game theory implicitly deals with psychology but it applies principles of psychology that only an underdeveloped child would adhere to.
And therefore a perfect fit for Homo Economicus as Our Betters would like us to be.
All game theory strikes me as being a very unusual approach toward looking at the world. The games are, to call a spade a spade, fantasies put into mathematical form. People come up with imaginary or fantasy scenarios (that don’t exist in the real world) and then try to cast them in mathematical form. But all the fantasies rest on at least two strongly unrealistic assumptions:
(1) Rationality on the part of the actors.
and
(2) Equality between the actor.
Both are impossible to fulfill in any real world situations.
So, I see this whole thing as a rather crude product of the human imagination. Not unlike myths or religious-ethical systems, but far more crude.
Err. No?
Game theory doesn’t hold anything about rationality of actor (it may say that what the actor _should_ do if they revealed payoffs are really they revealed payoff, but nothing at all what they _will_ do. In fact, big part of GT is to how to get the game so _no matter_ what the other player(s) do, what _should_ do to achieve your preference – if you know it), and even more so, absolutely nothing about “equality”.
Game theory is, in simple words, set of players, moves and payoffs. Any and all of them are arbitrary. You can have payoffs that move with time following no “logic”. You can have moves that only one player can do.
The thing is, once you define them, you operate within that defined space. If you introduce a new move/player/payoff, it’s an entirely new “game”. Indeed, one of things around GT is to show that often you need to introduce a new move to get “good” outcome (say in FRP it’s introduction of regulator/arbiter, in PD it’s durable reputation and repeatability).
And, of course, if your moves/players and payoff are complex, you get into a problem (both with saying what these are and figuring out how the system works). How’s that different from say three-bodies problem in physics?
That’s actually one of the reasons why easiest applicability of GT is mostly around auctions, as there you can control the players and moves (as the auctioner), and have some reasonable idea of payoffs (as the one doing the buying).
“That’s actually one of the reasons why easiest applicability of GT is mostly around auctions, as there you can control the players and moves (as the auctioner), and have some reasonable idea of payoffs (as the one doing the buying).”
Yes, game theory is mainly about controlling the players. You’re quite right in that. The control is undertaken by making restrictive assumptions in the theorems.
Then when the theorems are applied to policy or applied social science, these restrictions are attempted to be ‘made good’ in reality.
This is why I compare it to a religious or ethical doctrine. First the priest reasons as to what is necessary for a ‘perfect outcome’ then they take their doctrine and try to get people in the real world to conform so that their ethical rules are followed.
Indeed, a big part of any book/GT course is that even with a reasonably well defined and simple rules, players and payoffs, AND a known winning strategy, if the winning strategy relies on the other players to do the optimal moves, in real life it’s very likely to fail – it may be just too “computationally” intensive to get it right (not to mention counter intuitive).
Take Go it’s guaranteed to have at least not-losing strategy for the first mover (assuming equal positions), but I suspect even if we knew it, no human could really apply it.
Yes, this is the ‘pure mathematics’ defence that I mentioned in the article. (Knew this would happen). Game theorists say “Oh well, it’s just a pure game etc. etc. it doesn’t have any application in reality.”
Nonsense. First of all, why learn it then? Secondly, it is applied in reality — in political science, in economics, in negotiation strategy etc. And it usually makes the assumption mentioned above. (Especially in economics).
Game theory was invented to be applied in the social sciences using very restrictive assumptions (taken mainly from neoclassical economics). To say that this is a misappropriation is a very weak defence. It reminds me of Ken Arrow complaining that his silly mathematical model of the economy was used in policymaking. Eh… duh, Ken!
Why learn it? It is the same with all theoretical sciences — because of the asthetic beauty of it. Why learn number theory? Not because it has practical applications (it has none until the late 1960s among secret services and until 1970s for the general public, with the invention of public cryptography), but because it is beautiful. The same goes with just about every branch of mathematics out there. I suggest you abandon the utilitarian point of view when you come to criticising mathematics — history has proved that pure mathematicians, not seeking practical applications, lay the foundation for applications centuries ahead, for example, Pascal’s invention of binary number system as foundation for modern computing, RSA encryption is based on Euler-Fermat, 19th century non-Euclidean geometry was the groundwork for Einstein’s general relativity, zeros of the Riemann zeta function resemble closely to the energy level of uranium.
Game theory was not invented by social scientists, but by mathematicians (Cournot, Borel, von Neumann are all mathematicians), unless you consider Thomas Hobbes as the inventor! As usual, it is a tool that may or may not be applicable in a specific scenario, and it is the person who is using the tool that makes the decision to use the tool without checking the premises (didn’t we hear that before? Central limit theorem, static vs dynamic modelling, etc).
This article doesn’t really make any sense. As presented, the “Battle of the Sexes” might be true – if it is actually a game. But in mathematics there are equations that are trivial and there are equations with no solution. That doesn’t invalidate all of mathematics.
In game theory we have problems such as the “Prisoners Dilemma” which may not be earth shattering, but which result in a non-trivial solution/strategy.
Even the “Battle of the Sexes” probably has a “solution” if you can quantify just how badly the masculine party wants to get laid.
Thank you for the preface by de Montaigne “I prefer the company of peasants because they have not been educated sufficiently to reason incorrectly”, which I receive as license to participate in the discussion.
Peasant that I am, my principal take-away is condensed in the following “… one would be better trying to derive lessons about the real world from a consultation with a fortune-teller or a shaman”. I shall now ask my wife to read the tarot in order to decide where to go this evening. I predict the tarot will suggest opera as the far better choice. An undefinable danger lurks behind the football choice, whilst opera attendance will be fortuitous, leading to abundance, fecundity and excellent crops.
I also predict (yes, I’m a decent tarot-reader myself) that my peasant wife will casually invite me to look over her shoulder over coffee, at a web-page displaying the 26-year old leading lady’s décolleté.
The REAL issue with game theory, and humanity, is that it does NOT comprehend that winning is losing. What it defines as winning is empty because it severs relationship.
Well put!
As a fellow peasant, let me say that the key error in the “classical” approach to the “Battle of the Sexes” is that it imposes the wrong value system on the question. The key metric is not personal satisfaction from being able to attend one’s favored event; the key metric is marital felicity for both. In that light, the decision is very straightforward, and there is no dilemma: the opera wins out. Because, as a non-economist once said, when the missus is happy, everyone is happy, when the missus is unhappy, no one is happy.
As above… Game theory is for providing theorem’s, not for playing games.
Rationality and Emotions in Quantum Bargaining…. Rulliere.
http://annales.ensae.fr/anciens/n61/vol61-04.pdf
Skippy… The Journal of Philosophy anyone, Uni – Turin?
Yes, that’s the ‘pure mathematics’ argument. If you buy it fine. I don’t. It’s clear that they want to use these games for real-world situations.
http://en.wikipedia.org/wiki/Game_theory#General_and_applied_uses
As a tool specifically used in academic macro observations, uncovering – more – questions, I’m good with that. Although I agree with you on the point of human trials.
Skippy… same – same with derivatives, not enough computational power on the planet to do a proper run. So why the hell are we doing testing on such scale and with such speed. 10% risk inverts to 100%+ in the blink of an eye, seen it on the battlefield, death, the sight of it is a hand brake. Wish more got that or put it in the formula.
I thought that the lesson to be gleaned from game theory is the omnipresence of collective action problems in society. That when everyone tries to be number one pretty much everyone is worse off except the select few who claw their way to the edge of the bucket on everybody elses’ heads.
Government, its monopoly on violence and its ability to spend and tax, are supposed to mitigate such inequalities but when said government is up for sale to the highest bidder, the collection action problems between the 1% who can afford to play and the internecine battles that ensue simply exacerbate the collateral damage that is suffered by the rest. Too bad that the generation that gave us Sesame Street’s segments on co-op-er-a-tion could not be bothered to live by their own maxim.
What, no telepathy?
Actually, they’re working on it as we speak:
http://ibmresearchnews.blogspot.com/2011/12/mind-reading-is-no-longer-science.html
http://www.homelandsecuritynewswire.com/dhs-completes-tests-mind-reading-technology
Curious as to how many from the global power elites will fold up and go home knowing Nash and Game Theory have been spiked, staked, and burnt for the buzzards to pick and they now have to fall back on the bullshitting skills of the Grand Bernank and Oboomer the Splendificent. None, I figure.
WTF? Seriously. I’m in the “I don’t think that word means what you think it means” camp.
Game Theory was introduced into economics to deal with the problem of individuals and firms that react to the actions of other individuals and firms. That economic agents react to the actions of other economic agents seems like common sense but it generally gets relegated to a chapter on Oligopoly at the end of introductory texts. Economists wanted an oligopolistic equivalent to the neat equilibrium of perfect competition where everyone was an atomistic price taker. Game Theory was it, and it should come as no surprise that it was manipulated and twisted to ensure that it demonstrated the same basic results as perfect competition.
Very sorry if I didn’t read everything. PP,are you calling “marriage” the chicken and “gender” the egg?
Ha, ha, got dinged in Competitive Strategy in business school (U of Chicago) for basically calling BS on game theory in class on morality and reciprocation principles(professor hated my guts for that and remembered me at the time grades were distributed). Glad to see I’m not alone.
Game theory is an easy target – but I think you totally mis-represent any reasonable usage of it. Unlike classic market economics, in which nobody reacts to anything you do, game theory looks to abstract out the dynamics when each parties’ action can affect the others’ outcome. One can tell ridiculous stories with such abstractions (the straw man you tear down), but one can also look for deeper understanding – and when one does so it can be quite enlightening.
I studied a game theoretic paper in grad school which modeled “why birds sing to each other”. It highlighted this dynamic: by investing energy which shows no purpose except to signal “we have a relationship” it establishes a system where the standard of behavior is to cooperate altruistically when it matters (e.g. to squawk when a predator approaches, to warn your friends).
Likewise we could look at the dynamics of the two-party system from a game theoretic system, and easily understand why our current government is so incredibly dysfunctional.
You don’t need a game theory model to figure out stuff like why birds sing to each other or why the government is crap. If you can’t figure these out without a game theory model I suggest you stop trying to figure them out at all.
The models are pointless when they make sense and senseless when they try to make any real point.
Game Theory is not a living entity. The reason this article sucks is because it blames all problems on an amorphous entity that is called “game theory”.
I don’t think Phillip normally makes this mistake, but he did here. It’s twenty paragraphs of rambling about game theory with no substance.
Dear Philip –
[1] Internet packet routing and AI multi-agent systems run on computational/algorithmic game theory principles. In the reality-based community of electrical engineers and computer scientists, etc, this is a simple fact.
See for a few random examples —
‘COMPUTATIONAL GAME THEORY: A TUTORIAL
Neural Information Processing Systems ‘
http://www.cis.upenn.edu/~mkearns/nips02tutorial/
‘A Game Theory-Based Obstacle Avoidance Routing Protocol for Wireless Sensor Networks’
http://www.mdpi.com/1424-8220/11/10/9327/
‘Algorithmic Game Theory and Applications’
http://www2.cs.ucy.ac.cy/~viki/papers/Games.pdf
‘Swarm Intelligence’
http://en.wikipedia.org/wiki/Swarm_intelligence
There’s tens of thousands more papers out there. Very simply, it’s increasingly the dominant model used to build and model multi-agent systems of all kinds. Because it works.
That game theories can make robots act like robots is minorly impressive. Apply it to the human sciences and you get nonsense.
You seem to believe in some unobservable quality known as “human nature” to hang all your arguments on. But what happens if humans prove to to be virtually infinitely pliable and “programmable” as to “values” – so elastic as to produce cultures built around the torturing of captives wherein the torture is a measure of the captives’ strength and courage (as in parts of Native North America)? Or ours today, one that merrily bombs anyone else with a foreign tongue or skin-colour, so long as its a one-way drubbing?
Compare a typical day in 2012 with typical day in 1912 or 1812 and ask who is winning, the social engineers, or “freedom,creativity, independence, maturity” etc.
Not ‘human nature’. Just ‘the indeterminably of human decision making’. Simple.
Then write a different piece.
[2] The fact that you don’t know that without computational game theory possibly nobody would be reading your little internet screed in the first place — that algorithmic game theory is HOW THE INTERNET WORKS — might be embarrassing, I suppose.
Still, we’re all ignorant some of the time. The fact that you didn’t do the basic fact-checking/self-questioning/research to catch your ignorance is more damning.
You quoted Philip Mirowski with approval a month or so ago, for instance. All this material is referenced in Mirowski’s book MACHINE DREAMS, IIRC —
http://www.amazon.com/Machine-Dreams-Economics-Becomes-Science/dp/0521775264
While Mirowski is dismissive of Nash, noting that John von Neumann famously told Nash that the latter’s work was, essentially, trivial — and from von Neumann’s POV, one can see that he might consider Nash equalibria (NEs) as being implicit in and a mere restatement of his basic minimax theorem — Mirowski does understand this stuff. Start there if you like and read carefully. Because right now you’re bloviating about stuff you clearly don’t understand.
[3] Am I being unfair? Well, take your beginning statement:
‘Back to game theory. Why do we claim that it is so evasive? Because it nakedly displays the nature of the material it deals with and then denies that it is indeed this material that it is dealing with. Game theory sets up ‘games’ between certain personalities and then claims that the players have no personalities.’
No. You couldn’t be more wrong. The very point of game theory is that it does not set up games between certain personalities.
Game theory is essentially an attempt to model mathematically the strategic payoffs available to actors in a given contextual situation if vagaries of psychology and personality are deliberately and explicitly removed from the analysis. That’s it. Any prescriptive assumptions that an analyst brings to an analysis beyond that merely amount to that analyst’s prescriptive gingerbread.
[4] What’s that you say? That in reality vagaries of psychology and personality are never removed from any strategic analysis?
Well, firstly, that’s not necessarily true in the case of either animals/evolutionary game theory or multi-agent AI systems.
Secondly, however, game theorists have developed ways to analyze the strategic payoffs available to actors in contexts where other actors act irrationally or unpredictably. See for a few initial examples —
‘Did Thomas C. Schelling Invent the Madman Theory?’
http://hnn.us/articles/17183.html
Trembling hand perfect equilibrium
http://en.wikipedia.org/wiki/Trembling_hand_perfect_equilibrium
They all oversimplify. They don’t take in any context. They’re autistic. They could not deal with the complex ‘rules’ and ‘relations between players’ that I dealt with above. These relations are too nuanced and too complex to mathematically model.
Maybe you should read my piece and try to see where I’m coming from rather than shouting and citing papers like a devotee or a cult member.
Pot, meet kettle. If this was my first visit to Naked Capitalism, after I read the article I’d be tempted never to come back: Your answers to people trying to show you how much of an extremist your POV came out like just made it worse.
Ultimately, what you have here is a fallacious argument, quite similar to many we hear in talk radio.
Global warming does not cover the complexities of climate, therefore it cannot be right.
Since I cannot come up with a way for an eye to evolve all at once, it must be impossible, so evolution is impossible.
Argue that a very specific model does not work in a specific situation, but claiming that modeling can’t work is just an ego trip.
Haha! No… you cannot model human behavior. Of course you can’t. Stop being absurd.
You say that to say that no models work is ‘egotistic’. Here’s a proposition — one that my psychiatrist friends will fully adhere to: if you truly think you can model your fellow humans… I mean REALLY model them… then you’re psychotic and you need to be treated.
How about that? It’s a diagnostic criteria for paranoid schizophrenia. Look it up before you truck out your Freud!
You’re playing this out just as I thought:
“Game theory sets up ‘games’ between certain personalities and then claims that the players have no personalities.”
As I said at the end — read carefully, brother:
“Game theorists will, of course, tell me that I have cheated and misrepresented them. “We only play ‘pure’ games,” they will protest. “These are not supposed to be applied to real people with real psychologies in real institutional settings.” If the reader cares to look over the series of ‘games’ that the theorists set up – replete as they are with characters ranging from star-struck lovers to nefarious crooks – and finds the game theorists’ defence to be a credible one then they can happily discount my argument.”
If you believe the ‘pure maths’ defence, by all means have at it. But don’t come running back to me when you find some lab-coat kook applying it in a human sciences department…
Again, that it can apply to computational system doesn’t make it impressive.
While I like Mirowski I don’t believe in his dreams of a cyborg science of Man. Speaking with him personally I’m not sure how much he really buys into it either.
Imperialist game theory http://img.photobucket.com/albums/v256/LowlanderGSU/04_smuggling_1.jpg
Dilbert and the prisoner’s dilemma http://www2.owen.vanderbilt.edu/mike.shor/Humor/gametheorycartoons/images/SIMprison.jpg
The grassy knoll theory http://www.cartoonstock.com/newscartoons/cartoonists/jdu/lowres/jdun599l.jpg
My offtopic two cents
The introduction wastes time and possible audience on attacking religion. I am not your everyday zealot, but calling BS on religion makes one consider reading something else. This is something a preacher on a mission (such as yourself) should always consider.
Also, tongue-in-cheek, your blind smirk at shamanism puts you at odds with humanities, whose territory you are trying to border (peacefully I presume). Let me just tell you that one does not create whole cultures around shamanism if the shaman is just a bag of hot air.
On a bigger picture, I like what you are doing, though you still seem to beat around the bush. Stop trying to hit me and hit me:)
To dismiss the accomplishments of Enlightenment and Reason in a manner pretending that, with the benefits of a few centuries of hindsight, our contemporary half-assery is superior to the thoughts of minds that were very much entrapped in the societies and attitudes of their unenlightened times, yet, unlike us, took upon themselves the hard work of recognizing their limitations and overcoming them, is bloody well as unenlightened as it gets.
I am somewhat tired of all the convenient Malthus and Club of Rome bashing, and I am tired of seeing the foundational ideas of Enlightenement and Reason misrepresented by the merely lit.
The same is true for a la mode dismissal of game theory. There is nothing wrong with doing math with rational actors, it is the dishonest or plain incompetent *unreasoned* use of the results applied to systems without a lot of rationality that is *unenlightening* in the extreme – or worse, serves to sabotage any reasoned discussion of e.g. economics. One cannot fight lack of reason by unreasonableness in the opposite direction.
If one feels the overwhelming need to debate Descartes, debate Descartes. Kindly appreciate the Enlightenment for what it strove to do, despite the fallability of man. It has that much in common with science, and despite its “bland homogeneity of Reason”, it has served us better than the less bland alternatives.
I’m just summing up the history. I’m not dismissing them anymore than I’m dismissing the old religious systems. They had their time and their place, but they look silly now.
By the way, I don’t think that the Enlightenment were “very much entrapped in the societies and attitudes of their unenlightened times”. One of the key features of Enlightenment was an implicit accepting of the idea that all is equal. Or, at the very least, that all things could be understood by being compared and contrasted with each other in a simple manner (A=A).
That doesn’t look like an attitude that reflected the prejudice of the times at all. The Enlightenment brought us something quite new there. It is, however, something that reflects rather brutally the prejudices of our own times. And its one that game theory continues in like manner.
Sorry, that should read (A=B). ‘A=A’ sounds like something Ayn Rand would say! ;)
Why not go there then? Your griping about mathematics and utility already sounds like something von Mises would say.
Yes, one of the times that Mises got it right…
That’s really funny, because on my first read last night, I wrote this in response, then decided not to post to give you a break. You dump on the entire history of thought, yet believe in only 1 answer to everything – more “growth”, more “growth”, more “growth” right up to last tree standing. Who is naive, and who enlightened?:
“Those foolish Chinese, Egyptian, East Indian, Greek, Roman, New World Indian and other Ancients and Lord knows how many lines of learned thinking linking them through to moderns right up to WWII with their silly notions and strivings for balance, equity, harmony, peace, natural and social order, processes and laws – the stupefying waste of effort as their tiny minds smashed apart on the Rocks of Unlimited Growth and Power, their puny human dreams not yet matured, just not up to Godhood.” Roughly 6:00 am, Fed 28.
I know this is a nitpick and not your primary subject, but I strongly suspect the origin of religion lies with the near death experience rather than pure imagination. And we know they occured even in antiquity due to the account recorded by Plato. It wouldn’t take many NDEs to convince early man that there was an afterlife. Of course, the question is still open whether NDEs reflect a afterlife reality or not.
Mathematics is language invented by humans for the purpose of communicating with other humans. It is no fault of mathematics or its users that you don’t understand what it means or how to use it. Your argument is like criticizing Greek philosophy because it’s written in Greek.
I didn’t write a piece entitled ‘The Absurdity of Mathematics’. It’s about Game Theory specifically. For some things mathematics is great. For human activity… it’s usually as bad as reading goat intestines. Literally.
P.S. If you want to go down this route, the nature of the language of mathematics makes it inappropriate to apply to humans. People inhabit a world of meaning and nuance. Mathematics, by its nature, tries to exclude this. When you combine the two… well… you get stuff like game theory and other pseudo-rational garbage.
Speaking of nuance, you do realize that “Battle of the Sexes” is just a metaphor, meant to provide a memorable label and example, and not meant to describe any actual gender relations in the real world, right?
Also, you missed an equilibrium in your description, suggesting you don’t even understand the model that you’re criticizing.
Do you also criticize fairy tales and religious parables for not being totally realistic models of the world?
“Speaking of nuance, you do realize that “Battle of the Sexes” is just a metaphor, meant to provide a memorable label and example, and not meant to describe any actual gender relations in the real world, right?”
Sure it is. I believe you.
Seriously though, the point is that game theory IS applied to real world situations so you can’t defend it on the basis of ‘pure theory’.
“Also, you missed an equilibrium in your description, suggesting you don’t even understand the model that you’re criticizing.”
There’s two Nash equilibria. There’s some other one that has to do with running the game a few times or some such thing. I wasn’t going to go into that. Might put the reader to sleep. It wasn’t relevant to the analysis of the structure.
“Do you also criticize fairy tales and religious parables for not being totally realistic models of the world?”
If they were being applied in serious university human science faculties then, yes, controversial as that might seem…
Sorry two ‘pure strategy Nash equilibria’. The other is a mixed strategy. Again, not interesting for the sake of my piece.
Mark P’s criticisms aren’t relevant to the piece. He cites the sources to ‘prove’ stuff like that game theory is used for computer games and all that. I knew this. I’m talking about its use in applied human sciences.
I knew the game theory thing would make some people livid. It’s popular outside of neoclassical economics. Very popular. Same disease though.
Disease? Do you talk that way to people’s faces? Please, show some civility.
Watch the Adam Curtis film, Philip. It’s just what you’re saying. ;-)
YES! It is! Thank you. I love Adam.
Mathematical modelling of reality will always include estimations. Incorrect estimations will give incorrect answers even if the equations are solved correctly. In that respect there is no difference between algebra or game theory
Should we discard algebra as algebra has been used in economic models that turned out to be completely useless?
I’d recommend to continue to use algebra and direct criticism towards the model builders and their incorrect assumptions.
Depends on what level you’re modeling at. If you try and fit most microeconomic models into algebra you’re probably going down the wrong path. If applied at a high level of macroeconomic model building it can have limited usefulness. Not a huge amount, if you ask me. It’s just a tiny bit neater. And then this neatness always leads to ‘closed models’ that leave things out. I wouldn’t be a huge fan. Still better than game theory in that they’re not completely weird!
Philip,
Yes, game theory has been misused by some to push ideological agendas.
Yes, the games are inherently simplistic. But that’s what makes them powerful! They encapsulate simple models of one aspect of a very complex world. In your example you take a game – 2 players A and B, a rule for decision making, a payoff matrix – and then attack a complete non-sequitor, which is the name that some people apply to the game to make it easier for some people to understand the game in a more user-friendly context. Your attacking one set of arbitrary words that some people use to describe this.
What game theorists do is to look at the results and then almost always move immediately to asking, what makes the real world different?
Like with PD, we don’t observe the Nash equilibrium in real life, except in certain circumstances, and so we know there must be something more underlying human interaction. What else do we need to adjust in order to get something that roughly approximates the real world?
Unfortunately, with this post you’ve joined the camp of neo-classical economists in that you are now pushing an ideological agenda rather than arguing from the real world.
I agree with a lot of your views; I teach MMT to my grad students and others; I’m on your side; but in this case you’re wrong.
Did you write this article just to create an opportunity to argue and put other people down? Maybe you can find a more constuctive use for your time.
Adam Curtis’s film series THE TRAP addresses game theory marvelously.
Part 1
http://www.youtube.com/watch?v=gZt2HhFXB3M
Game theory is a mathematical field, and frankly your “criticisms” don’t apply to game theory proper.
Game theory in economics is another matter; it’s useful, but it is extremely primitive.
However, it’s the field which provided the experimental data which proved once and for all that people do *not* maximize short-term expected utility. So it’s done very useful work.
The economic game theorists I’ve met are all about seeing how people *actually* behave, experimentally, when given the super-simplified ‘games’ to play. People naturally follow strategies other than the ones initially predicated by Nash equilibrium, etc., so the economic game theorists tried to come up with coherent hypotheses which predicted the *actual* behavior of people.
In other words, they did actual scientific work. Experiments. Have you ever done any actual scientific work, Philip? Because your attack here *does not constitute an attack on the game theorists who do experimental work*. It simply doesn’t actually attack them. Are you merely attacking economic game theorists who don’t do experimental work? If so, attack them for *not doing experimental work*.
I’m intrigued that no one has commented on this post. I originally read it on my smartphone, which is useless for leaving comments so I left it, but I’ve come back to see how people responded and see nothing!
I think you’ve really missed the point of toy games. It’s all about setting up various simple structures of games to reflect different kinds of situations – the Battle of the Sexees is about co-ordination where payoffs are not equal (as opposed to the Driving Game, in which you and your opponent choose driving on the left or right – if you pick opposites you crash, so payoffs are equal). The story used is just a silly story and easy to modify so it’s not so un-PC and ridiculous. But either way the specific story isn’t the point.
Ken Binmore explains why “toy games” are used really well in the first chapter of “Playing for Real”. It’s a bit of a tome, but if you’re really interested in learning about game theory in a richer context I thoroughly recommend it, especially because it’s about more than economics (disclaimer – he was my game theory lecturer when I was an undergrad).
Ok that’s weird – when I arrived on this page I didn’t see any comments – ignore the first paragraph of the comment I just posted as now I can see!
I am fully aware of game theory, but with regard to current discussions of T.I.N.A (There is no alternative), how come when criticising the battle of the sexes game, rather than delve into some dodgy stereotypes about gender, did you not just say, well the couple do have the option of doing something else? (and yes this is also the limitation of prisoner’s dilemma and tragedy of the commons – yes tragedy of the commons (also re TOC as it had no or little precedent before enclosure acts but does occur now is that because self-interest only became the sole element of human rationality at that point, or perhaps did the enclosure acts create the conditions for the increased likelihood of its reoccurence – just one for those who can’t get their heads around the mediated aspects of human nature))
Although re: gender stereotypes – yes I realise it was a trope for the next bit on debt. But if the game is football (soccer, for US readers) and you want to change game, or even do some non-competitive exercise, get rid of the referee! If you accept the game theory pitch a game theory ref will be appointed.