You are currently browsing the tag archive for the ‘Game Theory’ tag.
Fig. – A Symbolical Head (phrenological chart) illustrating the natural language of the faculties. At the Society pages / Economic Sociology web page.
You have much probably noticed by now how Scoop.it is emerging as a powerful platform for those collecting interesting research papers. There are several good examples, but let me stress one entitled “Bounded Rationality and Beyond” (scoop.it web page) curated by Alessandro Cerboni (blog). On a difficult research theme, Alessandro is doing a great job collecting nice essays and wonderful articles, whenever he founds them. One of those articles I really appreciated was John Conlisk‘s “Why Bounded Rationality?“, delivering into the field several important clues, for those who (like me) work in the area. What follows, is an excerpt from the article as well as part of his introductory section. The full (PDF) paper could be retrieved here:
In this survey, four reasons are given for incorporating bounded rationality in economic models. First, there is abundant empirical evidence that it is important. Second, models of bounded rationality have proved themselves in a wide range of impressive work. Third, the standard justifications for assuming unbounded rationality are unconvincing; their logic cuts both ways. Fourth, deliberation about an economic decision is a costly activity, and good economics requires that we entertain all costs. These four reasons, or categories of reasons, are developed in the following four sections. Deliberation cost will be a recurring theme.
Why bounded rationality? In four words (one for each section above): evidence, success, methodology, and scarcity. In more words: Psychology and economics provide wide-ranging evidence that bounded rationality is important (Section I). Economists who include bounds on rationality in their models have excellent success in describing economic behavior beyond the coverage of standard theory (Section II). The traditional appeals to economic methodology cut both ways; the conditions of a particular context may favor either bounded or unbounded rationality (Section III). Models of bounded rationality adhere to a fundamental tenet of economics, respect for scarcity. Human cognition, as a scarce resource, should be treated as such (Section IV). The survey stresses throughout that an appropriate rationality assumption is not something to decide once for all contexts. In principle, we might suppose there is an encompassing single theory which takes various forms of bounded and unbounded rationality as special. cases. As with other model ingredients, however, we in practice want to work directly with the most convenient special case which does justice to the context. The evidence and models surveyed suggest that a sensible rationality assumption will vary by context, depending on such conditions as deliberation cost, complexity, incentives, experience, and market discipline. Beyond the four reasons given, there is one more reason for studying bounded rationality. It is simply a fascinating thing to do. We can mix some Puck with our Hamlet.
Picture – The European Conference on Complex Systems (ECCS’11 – link) at one of the main Austrian newspapers Der Standard: “Die ganze Welt als Computersimulation” (link), Klaus Taschwer, Der Standard, 14 September [click to enlarge – photo taken at the conference on Sept. 15, Vienna 2011].
“Take Darwin, for example: would Caltech have hired Darwin? Probably not. He had only vague ideas about some of the mechanisms underlying biological Evolution. He had no way of knowing about genetics, and he lived before the discovery of mutations. Nevertheless, he did work out, from the top down, the notion of natural selection and the magnificent idea of the relationship of all living things.” Murray Gell-Mann in “Plectics“, excerpted from The Third Culture: Beyond the Scientific Revolution by John Brockman (Simon & Schuster, 1995).
To be honest, I didn’t enjoy this title, but all of us had a fair share with journalists, now and then by now. After all, 99% of us don’t do computer simulation. We are all after the main principles, and their direct applications.
During 5 days (12-16 Sept.), with around 700 attendees the Vienna 2011 conference evolved around main important themes as Complexity & Networks (XNet), Current Trends in Game Theory, Complexity in Energy Infrastructures, Emergent Properties in Natural and Artificial Complex Systems (EPNACS), Complexity and the Future of Transportation Systems, Econophysics, Cultural and Opinion Dynamics, Dynamics on and of Complex Networks, Frontiers in the Theory of Evolution, and – among many others – Dynamics of Human Interactions.
For those who know me (will definitely understand), I was mainly attending those sessions underlined above, the last one (Frontiers in Evolution) being one of my favorites, among all these ECCS years. All in all, the conference had highly quality works (daily, we had about 3-4 works I definitely think should be followed in the future) and to those, more attention should be deserved (my main critics to the conference organization goes in here). Naturally, the newspaper article also reflects on the FuturICT, being historically one of the major scientific European projects ever done (along, probably, with the Geneva LHC), which teams spread across Europe, including Portugal with a representative team of 7 members present on the conference, led by Jorge Louçã, the former editor and organizer on the previous ECCS’10 last year in Lisbon.
Video – “… they forgot to say: in principle!“. Ricard Solé addressing the topic of a Morphospace for Biological Computation at ECCS’11 (European Conference on Complex Systems), while keeping is good humor on.
Let me draw anyway your attention to 4 outstanding lectures: Peter Schuster (link) on the first day, dissected on the source of Complexity in Evolution, battling among – as he puts it – two paradoxes: (1) Evolution is an enormously complex process, and (2) biological evolution on Earth proceeds from lower towards higher complexity. Earlier on that morning – opening the conference -, Murray Gell-Mann (link) who co-founded the Santa Fe Institute in 1984, gave a wonderful lecture on Generalized Entropies. Besides his age, the 1969 Nobel Prize in physics for his work on the theory of elementary particles, gladly turned his interest in the 1990s to the theory of Complex Adaptive Systems (CAS). Next, Albert-László Barabási (link), tamed Complexity on Controlling Networks. Finally, at the last day, closing the conference in pure gold, Ricard Solé (link) addressed the topic of a Morphospace for Biological Computation, an amazing lecture with a powerful topic to which – nevertheless – I felt he had little time (20 minutes), for such a rich endeavor. However – by no means -, he have lost his good humor during the talk (check my video above). Next year, the conference will be held in Brussels, and by just judging at the poster design, it promises. Go ants, go … !
Picture – The European Conference on Complex Systems (ECCS’12 – link) poster design for next year in Brussels.
Figure – Poker final hand rankings. Poker is a typical example of bounded rationality in our daily lives. Without having all the information available, you still have to make a decision. In one of his works, Herbert Simon states: “boundedly rational agents experience limits in formulating and solving complex problems and in processing (receiving, storing, retrieving, transmitting) information“.
[…] Bounded rationality is the idea that in decision making, rationality of individuals is limited by the information they have, the cognitive limitations of their minds, and the finite amount of time they have to make decisions. It was proposed by Herbert Simon as an alternative basis for the mathematical modelling of decision making, as used in economics and related disciplines; it complements rationality as optimization, which views decision making as a fully rational process of finding an optimal choice given the information available. Another way to look at bounded rationality is that, because decision-makers lack the ability and resources to arrive at the optimal solution, they instead apply their rationality only after having greatly simplified the choices available. Thus the decision-maker is a satisfier, one seeking a satisfactory solution rather than the optimal one. Simon used the analogy of a pair of scissors, where one blade is the “cognitive limitations” of actual humans and the other the “structures of the environment”; minds with limited cognitive resources can thus be successful by exploiting pre-existing structure and regularity in the environment. Some models of human behaviour in the social sciences assume that humans can be reasonably approximated or described as “rational” entities (see for example rational choice theory). Many economics models assume that people are on average rational, and can in large enough quantities be approximated to act according to their preferences. The concept of bounded rationality revises this assumption to account for the fact that perfectly rational decisions are often not feasible in practice due to the finite computational resources available for making them. […] In Wikipedia, (link).
Book cover – Herbert A. Simon. Models of Bounded Rationality, Volume 1, Economic Analysis and Public Policy, MIT Press 1984. The Nobel Prize in Economics was awarded to Herbert Simon in 1978. At Carnegie-Mellon University he holds the title of Professor of Computer Science and Psychology. These two facts together delineate the range and uniqueness of his contributions in creating meaningful interactions among fields that developed in isolation but that are all concerned with human decision-making and problem-solving processes. In particular, Simon has brought the insights of decision theory, organization theory (especially as it applies to the business firm), behavior modeling, cognitive psychology, and the study of artificial intelligence to bear on economic questions. This has led not only to new conceptual dimensions for theoretical constructions, but also to a new humanizing realism in economics, a way of taking into account and dealing with human behavior and interactions that lie at the root of all economic activity. The sixty papers and essays contained in these two volumes are grouped under eight sections, each with a brief introductory essay. These are: Some Questions of Public Policy, Dynamic Programming Under Uncertainty; Technological Change; The Structure of Economic Systems; The Business Firm as an Organization; The Economics of Information Processing; Economics and Psychology; and Substantive and Procedural Reality. Most of Simon’s papers on classical and neoclassical economic theory are contained in volume one. The second volume collects his papers on behavioral theory, with some overlap between the two volumes. (from MIT).
Video lecture by Ken Long (at the Statistics Problem Solvers blog) on Nassim Taleb‘s 4th Quadrant problems [1,2], i.e. a region where statistics not only don’t work but in which statistics are downright dangerous, because they lead you to make predictions as well as control systems that are unprepared for the kinds of systems shocks awaiting you.
“Statistical and applied probabilistic knowledge is the core of knowledge; statistics is what tells you if something is true, false, or merely anecdotal; it is the “logic of science”; it is the instrument of risk-taking; it is the applied tools of epistemology; you can’t be a modern intellectual and not think probabilistically—but… let’s not be suckers. The problem is much more complicated than it seems to the casual, mechanistic user who picked it up in graduate school. Statistics can fool you. In fact it is fooling your government right now. It can even bankrupt the system (let’s face it: use of probabilistic methods for the estimation of risks did just blow up the banking system).”, Nassim Taleb, in .
 Nassim Nicholas Taleb, “The Fourth Quadrant: A Map of the Limits of Statistics“, An Edge Original Essay, Set., 2008. (link)
 Nassim Nicholas Taleb,”Convexity, Robustness, and Model Error inside the Fourth Quadrant“, Oxford Lecture (Draft version), Oxford, July 2010. [PDF paper]
“It is not the strongest of the species that survives, nor the most intelligent that survives. It is the one that is the most adaptable to change“. Charles Darwin (On the Origin of Species, Nov. 1859)
During the Victorian era where high prudery and morality were constant, it would be hard to imagine seeing Charles Darwin wearing a Scottish-kilt. In fact, men’s formal clothing was less colourful than it was in the previous century, while women’s tight-fitting jersey dresses of the 1880s covered the body, leaving little to the imagination (source). There is however, one beautiful – as in strict sense of delighting the senses for exciting intellectual or emotional admiration – reason, I think he should have done it (!), regardless the obvious bearing consequences of a severe Victorian society. Surprisingly, some how, that reason is linked to cheetahs chasing gazelles, among many other things…
As the image of Charles Darwin wearing a kilt, you will probably find these awkward too, but when a cheetah chases a gazelle, banded tartan Scottish-kilt woven textile like patterns soon start to pop-up everywhere. Not at the ground terrain level, of course. Instead, they appear as a phenotype-like map between your present and the past. You may think that this banded tartans will have no significance for your life, but do mind this: crying babies do it all the time with their mommy’s and fathers, companies do it with other companies in their regular business, people commuting in large cities do it over large highways, human language, literature and culture does it, friends do it, PC virus and anti-virus software do it, birds singing do it also, … and even full countries at war do it.
One extreme example is the Cold War, where for the first time on our Human history, co-evolutionary arms-race raised to unprecedented levels. Co-Evolution is indeed the right common key-word for all these phenomena, while large white banded strips punctuated by tiny black ones (bottom-left woven kilt above), would be the perfect correspondent tartan pattern for the case of the Cold War example mentioned. But among these, there is of course, much more Scottish-kilt like patterns we could find. Ideas, like over this TV ad above, co-evolve too. Here, the marketeer decided to co-evolve with a previous popular famous meme image: Sharon Stone crossing his legs at the 1992 ‘Basic Instinct‘ movie. In fact, there is an authentic plethora of different possible behavioural patterns. Like a fingerprint (associated with different Gaelic clans), each of these patterns correspond to a lineage of current versus ancestral strategies, trying to solve a specific problem, or achieving one precise goal. But as the strategic landscape is dynamically changing all the time, a good question is, how can we visualize it. And, above all, what vital information and knowledge could we retrieve from this evolutionary Scottish-kilts maps.
Fig. – The frontispiece drawing to the English edition of Ernst Haeckel‘s Evolution of Man (trans. 1903) presents a skull labelled “Australian Negro” as an intervening evolutionary stage between the “Mediterranean” skull and those of the lower primates (from the 1891 ed. of the Anthropogenie).
In nature, organisms and species coexist in an ecosystem, where each species has its own place or niche in the system. The environment contains a limited number and amount of resources, and the various species must compete for access to those resources, where successive adaptations in one group put pressure on another group to catch up (e.g., the coupled phenomena of speed in the cheetah and evasive agility in the gazelle). Through these interactions, species grow and change, each influencing the others evolutionary development . This process of bi-adaptive relationship (in some cases can also assume a form of cooperation and mutualism) or reciprocal adaptation is know as Co-evolution, i.e. the evolution of two or more competing populations with coupled fitness.
The phenomena has several interesting features that may potentially enhance the adaptive power of artificial evolution , or other types of bio-inspired learning systems. In particular, competing populations may reciprocally drive one another to increasing levels of complexity by producing an evolutionary “arms race”, where each group may become bigger, faster, more lethal, more intelligent, etc. Co-Evolution can then happen either between a learner (e.g., single population) and its environment (i.e. based on competitions among individuals in the population) or between learning species (two populations evolving), where the fitness of individuals is based on their behaviour in the context of the individuals of the other population . This latter type of co-evolutionary search is often described as “host-parasite”, or “predator-prey” co-evolution. A good example and application of co-evolutionary learning include the pioneering work by Hillis in 1990  on sorting networks.
It can occur at multiple levels of biology: it can be as microscopic as correlated mutations between amino acids in a protein, or as macroscopic as co-varying traits between different species in an environment. Being biological Co-Evolution, in a broad sense, “the change of a biological object triggered by the change of a related object” , his visualization however, could be profoundly hard. In fact, attempting to define and monitor “progress” in the context of Co-Evolutionary systems can be a somewhat nightmarish experience , as stated in . It’s exactly here where Scottish-kilts come into play.
In 1995 , two researchers had a simple, yet powerful idea. In order to monitor the dynamics of artificial competitive co-evolutionary systems between two populations, Dave Cliff and Geoffrey Miller [3,4,5] proposed evaluating the performance of an individual from the current population in a series of trials against opponents from all previous generations. while visualizing the results as 2D grids of shaded cells or pixels: qualitative patterns in the shading can thus indicate different classes of co-evolutionary dynamic. Since their technique involves pitting a Current Individual (CI) against Ancestral Opponents (AO), they referred to the visualizations as CIAO plots (fig. above ).
Important Co-Evolutionary dynamics such as limited evolutionary memory, “Red Queen” effects or intransitive dominance cycling, will then be revealed like a fingerprint as certain qualitative patterns. Dominance cycling, for instance, it’s a major factor on Co-Evolution, wish could appear or not, during the entire co-evolutionary process. Imagine, for instance, 3 individuals (A,B,C) or strategies. Like over the well known “Rock, Paper, Scissors” game, strategy B could beat strategy A, strategy C could beat B, and strategy A could beat C, over and over in an eternal cycling, where only “arms race” specialized learning will emerge, at the cost of a limited learning generalization against a possible fourth individual-strategy D. If you play poker, you certainly know what I am talking about, since 2 poker players are constantly trying to broke this behavioural cycle, or entering it, depending on their so-far success.
Above (left and right figures – ), two idealised typical CIAO plot patterns can be observed, where darker shading denotes higher scores. On the left figure, however, co-evolutionary intransitive dominance cycling is a constant, where current elites (population A elites) score highly against population B opponents from 3, 8 and 13 generations ago, but not so well against generations in between. On the other hand (right figure), the behavioural pattern is completely different: over here we do observe limited evolutionary memory, where the current elites do well against opponents from 3,4 and 5 generations ago, but much less well against more distant ancestral opponents.
“For to win one hundred victories in one hundred battles is not the acme of skill. To subdue the enemy without fighting is the acme of skill.” ~ Sun Tzu
Of course, in increasingly complex real-world situations Scottish-kilt like CIAO plots are much noisy than this (fig. above -) where banded tartans could be less prominent, while the same could happen in irregular dominance cycling as elegantly showed by Cartlidge and Bullock in 2004 . Above, some of my own experiences can be observed (submitted work). Over here I decided to co-evolve a AI agent strategy to play against a pool of 15 different strategies (6 of those confronts are presented above), and as a result, 6 different behavioural patterns emerged between them. All in all, the full spectrum of co-evolving dynamics could be observed, from the “Red Queen” effect, till alternate dominant cycles, and limited or long evolutionary memory. Even if some dynamics seem counter-productive in one-by-one confronts, in fact, all of these dynamics are useful in some way, as when you play Poker or the “Rock, Paper, Scissors” game. A typical confront between game memory (exploitation) and the ability to generalize (exploration). Where against precise opponents limited evolutionary memory was found, the same effect produced dominant cycles or long evolutionary memory against other strategies. The idea of course, is not to co-evolve a super-strategy to win all one-by-one battles (something that would be rather impossible; e.g. No free Lunch Theorem) but instead to win the whole round-robin tournament, by being highly adaptive and/or exaptive.
So next time you see someone wearing a banded tartan Scottish-kilt do remind yourself that, while getting trapped in traffic, that precise pattern could be the result of your long year co-evolved strategies to find the quickest way home, while confronting other commuters doing the same. And that, somewhere, somehow, Charles Darwin is envying your observations…
 W. Daniel Hillis (1990), “Co-Evolving Parasites improve Simulated Evolution as an Optimization Procedure”, Physica D, Vol. 42, pp. 228-234 (later in, C. Langton et al. (Eds.) (1992), Procs. Artificial Life II, Addison-Welsey, pp. 313-324).
 Yip et al.; Patel, P; Kim, PM; Engelman, DM; McDermott, D; Gerstein, M (2008). “An integrated system for studying residue Coevolution in Proteins“. Bioinformatics 24 (2): 290-292. doi:10.1093/bioinformatics/btm584. PMID 18056067.
 Dave Cliff, Geoffrey F. Miller, (1995), “Tracking the Red Queen: Methods for measuring co-evolutionary progress in open-ended simulations“. In F. Moran, A. Moreno, J. J. Merelo, & P. Cachon (Eds.), Advances in artificial life: Proceedings of the Third European Conference on Artificial Life (pp. 200-218). Berlin: Springer-Verlag.
 Dave Cliff, Geoffrey F. Miller, (2006), “Visualizing Co-Evolution with CIAO plots“, Artificial Life, 12(2), 199-202
 Dave Cliff, Geoffrey F. Miller (1996). “Co-evolution of pursuit and evasion II: Simulation methods and results“. In P. Maes, M. J. Mataric, J.-A. Meyer, J. Pollack, & S. W. Wilson (Eds.), From Animals to Animats 4: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior (pp. 506-515). Cambridge, MA: MIT Press.
 Cartlidge, J. and Bullock S., (2004), “Unpicking Tartan CIAO plots: Understanding irregular Co-Evolutionary Cycling“, Adaptive Behavior Journal, 12: 69-92, 2004.
 Ramos, Vitorino, (2007), “Co-Cognition, Neural Ensembles and Self-Organization“, extended abstract for a seminar talk at ISR – Institute for Systems and Robotics, Technical Univ. of Lisbon (IST), Lisbon, PORTUGAL. May 31, 2007.
On Bilateral Monopolies: […] Mary has the world’s only apple, worth fifty cents to her. John is the world’s only customer for the apple, worth a dollar to him. Mary has a monopoly on selling apples, John has a monopoly (technically, a monopsony, a buying monopoly) on buying apples. Economists describe such a situation as bilateral monopoly. What happens? Mary announces that her price is ninety cents, and if John will not pay it, she will eat the apple herself. If John believes her, he pays. Ninety cents for an apple he values at a dollar is not much of a deal but better than no apple. If, however, John announces that his maximum price is sixty cents and Mary believes him, the same logic holds. Mary accepts his price, and he gets most of the benefit from the trade. This is not a fixed-sum game. If John buys the apple from Mary, the sum of their gains is fifty cents, with the division determined by the price. If they fail to reach an agreement, the summed gain is zero. Each is using the threat of the zero outcome to try to force a fifty cent outcome as favorable to himself as possible. How successful each is depends in part on how convincingly he can commit himself, how well he can persuade the other that if he doesn’t get his way the deal will fall through. Every parent is familiar with a different example of the same game. A small child wants to get her way and will throw a tantrum if she doesn’t. The tantrum itself does her no good, since if she throws it you will refuse to do what she wants and send her to bed without dessert. But since the tantrum imposes substantial costs on you as well as on her, especially if it happens in the middle of your dinner party, it may be a sufficiently effective threat to get her at least part of what she wants. Prospective parents resolve never to give in to such threats and think they will succeed. They are wrong. You may have thought out the logic of bilateral monopoly better than your child, but she has hundreds of millions of years of evolution on her side, during which offspring who succeeded in making parents do what they want, and thus getting a larger share of parental resources devoted to them, were more likely to survive to pass on their genes to the next generation of offspring. Her commitment strategy is hardwired into her; if you call her bluff, you will frequently find that it is not a bluff. If you win more than half the games and only rarely end up with a bargaining breakdown and a tantrum, consider yourself lucky.
Herman Kahn, a writer who specialized in thinking and writing about unfashionable topics such as thermonuclear war, came up with yet another variant of the game: the Doomsday Machine. The idea was for the United States to bury lots of very dirty thermonuclear weapons under the Rocky Mountains, enough so that if they went off, their fallout would kill everyone on earth. The bombs would be attached to a fancy Geiger counter rigged to set them off if it sensed the fallout from a Russian nuclear attack. Once the Russians know we have a Doomsday Machine we are safe from attack and can safely scrap the rest of our nuclear arsenal. The idea provided the central plot device for the movie Doctor Strangelove. The Russians build a Doomsday Machine but imprudently postpone the announcement they are waiting for the premier’s birthday until just after an American Air Force officer has launched a unilateral nuclear attack on his own initiative. The mad scientist villain was presumably intended as a parody of Kahn. Kahn described a Doomsday Machine not because he thought we should build one but because he thought we already had. So had the Russians. Our nuclear arsenal and theirs were Doomsday Machines with human triggers. Once the Russians have attacked, retaliating does us no good just as, once you have finally told your daughter that she is going to bed, throwing a tantrum does her no good. But our military, knowing that the enemy has just killed most of their friends and relations, will retaliate anyway, and the knowledge that they will retaliate is a good reason for the Russians not to attack, just as the knowledge that your daughter will throw a tantrum is a good reason to let her stay up until the party is over. Fortunately, the real-world Doomsday Machines worked, with the result that neither was ever used.
For a final example, consider someone who is big, strong, and likes to get his own way. He adopts a policy of beating up anyone who does things he doesn’t like, such as paying attention to a girl he is dating or expressing insufficient deference to his views on baseball. He commits himself to that policy by persuading himself that only sissies let themselves get pushed around and that not doing what he wants counts as pushing him around. Beating someone up is costly; he might get hurt and he might end up in jail. But as long as everyone knows he is committed to that strategy, other people don’t cross him and he doesn’t have to beat them up. Think of the bully as a Doomsday Machine on an individual level. His strategy works as long as only one person is playing it. One day he sits down at a bar and starts discussing baseball with a stranger also big, strong, and committed to the same strategy. The stranger fails to show adequate deference to his opinions. When it is over, one of the two is lying dead on the floor, and the other is standing there with a broken beer bottle in his hand and a dazed expression on his face, wondering what happens next. The Doomsday Machine just went off. With only one bully the strategy is profitable: Other people do what you want and you never have to carry through on your commitment. With lots of bullies it is unprofitable: You frequently get into fights and soon end up either dead or in jail. As long as the number of bullies is low enough so that the gain of usually getting what you want is larger than the cost of occasionally having to pay for it, the strategy is profitable and the number of people adopting it increases. Equilibrium is reached when gain and loss just balance, making each of the alternative strategies, bully or pushover, equally attractive. The analysis becomes more complicated if we add additional strategies, but the logic of the situation remains the same.
This particular example of bilateral monopoly is relevant to one of the central disputes over criminal law in general and the death penalty in particular: Do penalties deter? One reason to think they might not is that the sort of crime I have just described, a barroom brawl ending in a killing more generally, a crime of passion seems to be an irrational act, one the perpetrator regrets as soon as it happens. How then can it be deterred by punishment? The economist’s answer is that the brawl was not chosen rationally but the strategy that led to it was. The higher the penalty for such acts, the less profitable the bully strategy. The result will be fewer bullies, fewer barroom brawls, and fewer “irrational” killings. How much deterrence that implies is an empirical question, but thinking through the logic of bilateral monopoly shows us why crimes of passion are not necessarily undeterrable. […]
Note – Further reading should include David D. Friedman’s “Price Theory and Hidden Order“. Also, a more extensive treatment could be found on “Game Theory and the Law“, by Douglas G. Baird, Robert H. Gertner and Randal C. Picker, Cambridge, Mass: Harvard University Press, 1994.
From left to rigth, Nelson Minar, JJ Merelo (one of my co-authors), Manor Askenazi and Chris Langton (founding father of Artificial Life) at the El Farol Bar, Santa Fe, New Mexico, during summer 1995. At the same year, Chris was the editor of the well-know Artificial Life book, by MIT Press, and JJ for the 3rd European Conference on Artificial Life, Granada, Spain.
In case you do not have a clue what the El Farol Bar meant to the Santa Fe Institute (SFI), have a read here to Brian Arthur‘s paper “Inductive Reasoning and Bounded Rationality: The El Farol bar problem“, American Economic Review, 84, 406-411, 1994 (or check previous posts). I am happy to say that I was also there, visiting Santa Fe back in 2000, speaking with, among other people with Cosma Shalizi, as well as having a cigar and a beer at the El Farol. Much probably at this table, which was near the front door window, one of my favourite ones during my two week stay.
Finally, and in what regards the ongoing present financial world crisis, here’s a quote from 1994’s Arthur’s paper:
[…] Economists have long been uneasy with the assumption of perfect, deductive rationality in decision contexts that are complicated and potentially ill-defined. The level at which humans can apply perfect rationality is surprisingly modest. Yet it has not been clear how to deal with imperfect or bounded rationality. From the reasoning given above, I believe that as humans in these contexts we use inductive reasoning: we induce a variety of working hypotheses, act upon the most credible, and replace hypotheses with new ones if they cease to work. Such reasoning can be modeled in a variety of ways. Usually this leads to a rich psychological world in which agents’ ideas or mental models compete for survival against other agents’ ideas or mental models–a world that is both evolutionary and complex. […]
[…] The type of rationality we assume in economics – perfect, logical, deductive rationality–is extremely useful in generating solutions to theoretical problems. But it demands much of human behavior – much more in fact than it can usually deliver. If we were to imagine the vast collection of decision problems economic agents might conceivably deal with as a sea or an ocean, with the easier problems on top and more complicated ones at increasing depth, then deductive rationality would describe human behavior accurately only within a few feet of the surface. For example, the game Tic-Tac-Toe is simple, and we can readily find a perfectly rational, minimax solution to it. But we do not find rational “solutions” at the depth of Checkers; and certainly not at the still modest depths of Chess and Go.
There are two reasons for perfect or deductive rationality to break down under complication. The obvious one is that beyond a certain complicatedness, our logical apparatus ceases to cope – our rationality is bounded. The other is that in interactive situations of complication, agents can not rely upon the other agents they are dealing with to behave under perfect rationality, and so they are forced to guess their behavior. This lands them in a world of subjective beliefs, and subjective beliefs about subjective beliefs. Objective, well-defined, shared assumptions then cease to apply. In turn, rational, deductive reasoning–deriving a conclusion by perfect logical processes from well-defined premises – itself cannot apply. The problem becomes ill-defined.
As economists, of course, we are well aware of this. The question is not whether perfect rationality works, but rather what to put in its place. How do we model bounded rationality in economics? Many ideas have been suggested in the small but growing literature on bounded rationality; but there is not yet much convergence among them. In the behavioral sciences this is not the case. Modern psychologists are in reasonable agreement that in situations that are complicated or ill-defined, humans use characteristic and predictable methods of reasoning. These methods are not deductive, but inductive. […] The system that emerges under inductive reasoning will have connections both with evolution and complexity. […]
This is something that you face it in everyday life, be it on bars, restaurants, supermarket cues or in highways. Buying a house or selling it. So, have a look and decide for yourself! The problem is as follows: There is a particular, finite population of people. Every Thursday night, all of these people want to go to the El Farol Bar. However, the El Farol is quite small, and it’s no fun to go there if it’s too crowded. So much so, in fact, that the following rules are in place:
- If less than 60% of the population go to the bar, they’ll all have a better time than if they stayed at home.
- If more than 60% of the population go to the bar, they’ll all have a worse time than if they stayed at home.
Unfortunately, it is necessary for everyone to decide at the same time whether they will go to the bar or not. They cannot wait and see how many others go on a particular Thursday before deciding to go themselves on that Thursday.
One aspect of the problem is that, no matter what method each person uses to decide if they will go to the bar or not, if everyone uses the same method it is guaranteed to fail. If everyone uses the same deterministic method, then if that method suggests that the bar will not be crowded, everyone will go, and thus it will be crowded; likewise, if that method suggests that the bar will be crowded, nobody will go, and thus it will not be crowded. Often the solution to such problems in game theory is to permit each player to use a mixed strategy, where a choice is made with a particular probability. In the case of the El Farol Bar problem, however, no mixed strategy exists that all players may use in equilibrium.
[…] Consider now a problem I will construct to illustrate inductive reasoning and how it might be modeled. N people decide independently each week whether to go to a bar that offers entertainment on a certain night. For concreteness, let us set N at 100. Space is limited, and the evening is enjoyable if things are not too crowded–specifically, if fewer than 60% of the possible 100 are present. There is no way to tell the numbers coming for sure in advance, therefore a person or agent: goes–deems it worth going–if he expects fewer than 60 to show up, or stays home if he expects more than 60 to go. (There is no need that utility differ much above and below 60.) Choices are unaffected by previous visits; there is no collusion or prior communication among the agents; and the only information available is the numbers who came in past weeks. (The problem was inspired by the bar El Farol in Santa Fe which offers Irish music on Thursday nights; but the reader may recognize it as applying to noontime lunch-room crowding, and to other coordination problems with limits to desired coordination.) Of interest is the dynamics of the numbers attending from week to week.
Notice two interesting features of this problem. First, if there were an obvious model that all agents could use to forecast attendance and base their decisions on, then a deductive solution would be possible. But this is not the case here. Given the numbers attending in the recent past, a large number of expectational models might be reasonable and defensible. Thus, not knowing which model other agents might choose, a reference agent cannot choose his in a well-defined way. There is no deductively rational solution–no “correct” expectational model. From the agents’ viewpoint, the problem is ill-defined and they are propelled into a world of induction. Second, and diabolically, any commonalty of expectations gets broken up: If all believe few will go, all will go. But this would invalidate that belief. Similarly, if all believe most will go, nobody will go, invalidating that belief. Expectations will be forced to differ.
At this stage, I invite the reader to pause and ponder how attendance might behave dynamically over time. Will it converge, and if so to what? Will it become chaotic? How might predictions be arrived at? […]