You are currently browsing the monthly archive for August 2011.
Its called Fehr Advice!
Traditional economic research assumes that managers, employees, customers, and suppliers usually make rational decisions and thus do not make systematic decision errors. Behavioral economics, however, has found numerous proofs that people make systematic decision errors that limit their own welfare and also diminish efficiency, perceived fairness, and the profitability of firms.
It was for this reason that FehrAdvice & Partners AG developed the consulting approach BEA™ that is based on empirical insights about the human tendency to make erroneous decisions. This approach systematically includes this knowledge in consulting activities. Our consultants have a “trained eye” in the area of behavioral economics, and the innovative methods of empirical research we use allow us to identify potential areas of improvement in enterprises, markets, and organizations that were previously ignored.
They have a blog (in German), they are on Twitter, and they are hiring.
In the thirteenth century, Italians and Dutch traders went to Champagne not to drink champagne but to trade. They had to travel there and back and worry about theft. There was always the chance that some dispute would arise at the trade fairs. Courts arose to enforce contracts. Did they arise spontaneously in Coasian fashion, created by contracting parties to facilitate trade? Or was their government intervention? The first view is advocated by Milgrom, North and Weingast in a lovely and influential paper. MNW invoke some “stylized facts” about the institution of the “Law Merchant”, They claim the law merchant was a kind of store of the history of past exchanges. The law merchant could substitute for the incomplete knowledge of trading parties themselves and had good incentives to be honest himself to retain his income as a law merchant. The paper is mainly theoretical and has a nice prisoner’s dilemma model with traders changing partners every period.
A new paper by Edwards and Ogilvie challenges the stylized facts that motivate MNW. They claim:
The policies of the counts of Champagne played a major role in the rise of the fairs. The counts had an interest in ensuring the success of the fairs, which brought in very
significant revenues. These revenues in turn enabled the counts to consolidate their political position by rewarding allies and attracting powerful vassals….The first institutional service provided by the counts of Champagne consisted of mechanisms for ensuring security of the persons and property rights of traders. The counts undertook early, focused and comprehensive action to ensure the safety of merchants travelling to and from the fairs..
A second institutional service provided by the rulers of Champagne was contract enforcement. The counts of Champagne operated a four-tiered system of public lawcourts which judged lawsuits and officially witnessed contracts with a view to subsequent enforcement…
A final reason for the success of the Champagne fair-cycle was that it offered an almost continuous market for merchandise and financial services throughout the year, like a great trading city, but without the most severe disadvantage of medieval cities – special privileges for locals that discriminated against foreign merchants
The paper is an interesting read and there are lots of rich details about the Champagne fairs themselves.
“We were worried it was going to be a gripe site,” said the chief executive, Stephen Kaufer. “Who the heck would bother to write a review except to complain?” Instead, the average of the 50 million reviews is 3.7 stars out of five, bordering on exceptional but typical of review sites.
In fact, we can reverse the logic: “Who the heck would bother to write except to praise?”
Imagine you are asked to write a letter of recommendation for someone up for tenure. First, the university asks you if you are willing to write the letter. You mentally measure the amount of time it is going to take to read the papers. Add to that the time to write a clear and comprehensive letter. Are you going o do all that just to say something bad? Probably not. But if you are going to write something nice that gives the candidate a job for life, that might give you the satisfied buzz to counterbalance the cost of writing a letter. So, letters of recommendation will be biased towards the positive.
There is still some information: Bad candidates will get fewer letters than good candidates. Buts is this carefully noted? Is the number of letter writers who refused to write letters even recorded?
Perhaps the main countervailing force is envy. Why does X deserve tenure at highly ranked University A while I the letter writer am at humble University B? It is impossible not to write a letter for University A. If the candidate is bad, you are forthcoming. If the candidate is good, you are begrudging. But the quality of the letter is monotone in the quality of the candidate and information is aggregated. Only the very best universities are the object of envy. The rest have to decode the positive bias in their tenure procedures just like Tripadvisor users.
Usain Bolt was disqualified in the final of the 100 meters at the World Championships due to a false start. Under current rules, in place since January 2010, a single false start results in disqualification. By contrast, prior to 2003 each racer who jumped the gun would be given a warning and then disqualified after a second false start. In 2003 the rules were changed so that the entire field would receive a warning after a false start by any racer and all subsequent false starts would lead to disqualification.
Let’s start with the premise that an indispensible requirement of sprint competition is that all racers must start simultaneously. That is, a sprint is not a time trial but a head-to-head competition in which each competitor can assess his standing at any instant by comparing his and his competitors’ distance to a fixed finished line.
Then there must be penalty for a false start. The question is how to design that penalty. Our presumed edict rules out marginally penalizing the pre-empter by adding to his time, so there’s not much else to consider other than disqualification. An implicit presumption in the pre-2010 rules was that accidental false starts are inevitable and there is a trade-off between the incentive effects of disqualification and the social loss of disqualifying a racer who made an error despite competing in good faith.
(Indeed this trade-off is especially acute in high-level competitions where the definition of a false start is any racer who leaves less than 0.10 seconds after the report of the gun. It is assumed to be impossible to react that fast. But now we have a continuous variable to play with. How much more impossible is it to react within .10 seconds than to react within .11 seconds? When you admit that there is a probability p>0, increasing in the threshold, that a racer is gifted enough to reach within that threshold, the optimal incentive mechanisn picks the threshold that balances type I and type II errors. The maximum penalty is exacted when the threshold is violated.)
Any system involving warnings invites racers to try and anticipate the gun, increasing the number of false starts. But the pre- and post-2003 rules play out differently when you think strategically. Think of the costs and benefits of trying to get a slightly faster start. The warning means that the costs of a potential false start are reduced. Instead of being disqualified you are given a second chance but are placed in the dangerous position of being disqualified if you false start again. In that sense, your private incentives to time the gun are identical whether the warning applies only to you or to the entire field. But the difference lies in your treatment relative to the rest of the field. In the post-2003 system that penalty will be applied to all racers so your false start does not place you at a disadvantage.
Thus, both systems encourage quick starts but the post 2003 system encouraged them even more. Indeed there is an equilibrium in which false starts occur with probability close to 1, and after that all racers are warned. (Everyone expects everyone else to be going early, so there’s little loss from going early yourself. You’ll be subject to the warning either way.) After that ceremonial false start the race becomes identical to the current, post 2010, rule in which a single false start leads to disqualification. My reading is that equilibrium did indeed obtain and this was the reason for the rule change. You could argue that the pre 2003 system was even worse because it led to a random number of false starts and so racers had to train for two types of competition: one in which quick starts were a relevant strategy and one in which they were not.
Is there any better system? Here’s a suggestion. Go back to the 2003-2009 system with a single warning for the entire field. The problem with that system was that the penalty for being the first to false start was so low that when you expected everyone else to be timing the gun your best response was to time the gun as well. So my proposal is to modify that system slightly to mitigate this problem. Now, if racer B is the first to false start then in the restart if there is a second false start by, say racer C, then racer C and racer B are disqualified. (In subsequent restarts you can either clear the warning and start from scratch or keep the warning in place for all racers.)
Here’s a second suggestion. The racers start by pushing off the blocks. Engineer the blocks so that they slide freely along their tracks and only become fixed in place at the precise moment that the gun is fired.
(For the vapor mill, here are empirical predictions about the effect of previous rule-regimes on race outcomes:
- Comparing pre-2003, under the 2003-2009 you should see more races with at least one false start but far fewer total false starts per race. The current rules should have the least false starts.
- Controlling for trend (people get faster over time) if you consider races where there was no false start, race times should be faster 2003-2009 than pre-2003. That ranking reverses when you consider races in which there was at least one false start. Controlling for Usain Bolt, times should be unambiguously slower under current rules.)
There are four major providers of national cellphone service in the U.S – Verizon, AT&T, Sprint and T-Mobile. Two of them are proposing to merge. What impact will the merger have on consumers? Senator Herbert Kohl (of Kohl’s stores fame) says:
“According to Consumer Reports, ‘T-Mobile plans typically cost $15 to $50 per month less than comparable plans from AT&T.” Removal of such a maverick price competitor from such a highly concentrated market – a competitor that disciplines price increases from all three other national cell phone competitors, not only At&T – raises a substantial likelihood that prices will rise following this merger.”
Someone can take the opposite view to ATT-TM LT to DOJ and FCC but at least it is coherently argued.
From security analysts Raelynn Hillhouse and her blog The Spy Who Billed Me:
Sources in the intelligence community tell me that after years of trying and one bureaucratically insane near-miss in Yemen,the US government killed OBL because a Pakistani intelligence officer came forward to collect the approximately $25 million reward from the State Department’s Rewards for Justice program.
The informant was a walk-in.
The ISI officer came forward to claim the substantial reward and to broker US citizenship for his family. My sources tell me that the informant claimed that the Saudis were paying off the Pakistani military and intelligence (ISI) to essentially shelter and keep bin Laden under house arrest in Abbottabad, a city with such a high concentration of military that I’m told there’s no equivalent in the US.
The CIA and friends then set about proving that OBL was indeed there. And they did.
Next they approached the chiefs of the Pakistani military and the ISI. The US was going to come in with or without them. The CIA offered them a deal they couldn’t refuse: they would double what the Saudis were paying them to keep bin Laden if they cooperated with the US. Or they could refuse the deal and live with the consequences: the Saudis would stop paying and there would be the international embarassment…
The ISI and Pakistani military were cooperating with the US on the raid.
This is turning into a Tom Clancy novel
You might have thought it obvious that the stock market would go down after S&P downgraded US government debt. The bad news about US debt made investors worry, and worried investors are usually less enthusiastic about holding stocks.
But there is something wrong with this view. Ask yourself, when fearful investors sell their stocks, what do they buy? They sell their stocks for cash, of course; and then, being fearful, they typically want to keep the proceeds in the nearest thing to cash that pays interest: US government debt. Thus, as investors’ demand for stocks goes down, their demand for dollars and other US liabilities goes up. Such a surge in demand for US government debt would cause the price of US bonds to go up, which means that the interest rate on US debt would go down. Doesn’t it seem paradoxical, that a downgrading of US government debt could cause demand for this debt to increase?
But sure enough, the New York Times described with some surprise that the United States Treasury was actually a beneficiary of the market shifts today (Aug 8), despite the downgrade of its debt, as 10-year yields fell to 2.32 percent from 2.56 percent, and the yield on the two-year Treasury note hit a record low. Those who are worried about inflation should also notice that the decline in the stock market means that any given amount of dollars can actually buy more shares of the Dow Industrials.
So it is very difficult to see investors’ behavior today as a reaction to fears about inflationary deficits or a default on US government debt. If serious investors were actually worried about the real value of US government liabilities then they should tend to move out of bond markets and into real investments like the stock market, which should drive stock prices up. Such a move would help get real investment started, which would help get people back to work; but that is not what we saw today.
To understand what is really happening, we need to think more carefully about the risks that S&P was assessing. Of course the US government as a whole cannot be incapable of paying the dollars that it owes, because the US government has the ability to print dollars itself. So how can the S&P bond raters have any legitimate concerns about a possibility of the US government defaulting on its debt obligations? The answer is that a default could happen only if one part of the US government prevented other parts from paying the bills.
The bills of the US government are paid by the Treasury Department, but the ability to print dollars is vested in the Federal Reserve. The Treasury can get new dollars by issuing bonds that are purchased by the Federal Reserve. But, although the bonds in such a transaction would be held by the government itself, they would still be counted in the aggregate debt that is restricted by Congress’s debt limit. So when the US Congress refused to raise the debt limit, it was threatening to prevent the Treasury and Federal Reserve from working together to pay for the government’s budgeted expenses and debt obligations. In a situation like that, the President would indeed have to choose between cutting budgeted government expenses or asking the bond holders to wait.
We have seen, however, that the investors’ movement from stocks to bonds today is very hard to reconcile with fears of default on these bonds. So to explain the stock market decline, we have to look at the other side of the story, the very real possibility that a politically constrained US government might have to cut expenses for essential government services. Broad fears of a crippled US government that is unable to enforce laws or invest in infrastructure could do very serious damage to investment and economic growth in America. Indeed the possibility of such government paralysis could be far more economically damaging than any marginal increase in taxes.
Thus, there is every reason to believe that investors are reacting, not to fears of too much government debt, but to fears of too little government spending where it is needed. Investors are expressing fears that the US government may become unable to do its essential part in maintaining the strength of this country. Pundits and congressmen should take note.
All uncles are weird. But being an uncle is an exogenous event independent of any quality you have, weirdness or otherwise. So statistically it looks like a monumental fluke that all the weird people turn out to be uncles.
In reality it’s the uncleness that causes the weirdness. And more than that: even totally normal people who happen also to be uncles turn onto weirdos precisely when they are around their nieces and nephews and here’s why.
An uncle is basically an anti-parent. Not anti- in the sense of anti-aircraft or anti-American. More like antimatter. Because an uncle looks at his brother or sister’s family and basically sees everything that his own family is not. Good or bad. Especially in terms of the children.
And just like boys socialize by playing up and exaggerating differences as a mechanism for toughening up each others’ weak spots, the uncle can’t help but play that same role with the nieces and their parents. Niece is too sensitive because she is too sheltered, uncle brings the missing risk to the party. Nephew thinks he’s tough because he can push around his younger sisters, uncle gives him a taste of his own medicine.
Then there’s spy mode. Uncle wants the inside scoop on his brother/sister and spouse so he asks neice/nephew strange leading questions. Your uncle is an outsider who has just enough of an inside track to feel comfortable poking around where he doesn’t belong.
But whatever mischief the uncle is up to, the actual effect is that he comes across as a weirdo to his neices and nephews.
- Stingrays hurt.
- A lot.
- I could use a pedicure.
Related to my earlier post, where I suggested there as advantage to each chamber from delaying debt limit proposal as long as possible, Pelosi says:
[I]f we had days, instead of backed up to hours, we could have said ‘you don’t have the votes, let’s go back in and how do we move this way in order to cut some of those cuts and have a better bill and get the votes.’ So I think we could’ve done better. I think they were successful at just prolonging it to the last minute so that we didn’t have that option and it was default or no default.
The Guardian has gotten hold the U.K. interrogation policy:
In deciding whether to give permission [for overseas interrogation], senior MI5 and MI6 management “will balance the risk of mistreatment and the risk that the officer’s actions could be judged to be unlawful against the need for the proposed action”.
At this point, “the operational imperative for the proposed action, such as if the action involves passing or obtaining life-saving intelligence” would be weighed against “the level of mistreatment anticipated and how likely those consequences are”.
Here is a quote from Robert Parker
“Any serious introspection of the global wine market for Bordeaux over the last two years has to include the fact that it is impossible to determine the amount of 2009 Bordeaux futures (and in a few months, 2010 Bordeaux futures) that have actually been sold to consumers. Throughout Bordeaux there is talk of the massive market in Asia, and the increasing significance of the English wine investment firms, but there are those (and I wouldn’t dismiss their opinions) who tend to think that such assertions are grossly inflated. Moreover, they argue that there is a real bubble that is in danger of bursting if the right external influences unfold. One theory is that the Big Eight (which includes all the first growths of the Médoc as well as Haut-Brion and the trifecta of unofficial first growths of the Right Bank, Petrus, Cheval Blanc and Ausone) are actually hoarding huge inventories of their wines to inflate prices. This theory also suggests that the super seconds and many of the other cherished names in Bordeaux are doing the same thing. Why? They are trying to manipulate the market price. The appearance of little or no appreciable quantities of wine from two great vintages equals higher and higher prices. Is there a falsification of the demand from Asian consumers? The fact is, no one seems to know the answer. While some 2009s have not held their initial opening prices because they were too high, many have. If much of the 2009s, as well as the 2010s, are not sold through to wine consumers, who are the true marketplace since they actually drink these wines, and then tend to replenish their stock, buttressing the marketplace, then this is a bubble. Despite huge warehouses filled with reserve stocks of great vintages, prices could be set for a major adjustment, just as we have seen in the United States with the real estate market. What, if any of this, is true?
I raise this issue only because it is a possibility. The fact that no one can (or wants to) provide the actual sales figures of how much 2009 (or over the next six months, how much 2010) is actually being sold through to consumers is astonishing. If most of the stocks of these two vintages are held by importers, négociants, wholesalers, or on paper by investment firms, then it is obvious the consumers have not purchased 2009 and eventually 2010. In any event, I think this scenario has to be raised, given the overheated marketplace and the sometimes absurd rhetoric about how popular these wines are at prices of $1000 or more a bottle.” Robert Parker, May 3rd 2011
When the debt limit increase finally passed, the law included the creation of a 12 person committee, the SuperCongress, which will negotiate the next round of spending cuts and tax increases. If they fail to agree, automatic spending cuts go into effect. These spending cuts include elements that are painful to both parties and this punishment is meant to help the committee members compromise. Also, the automatic spending cuts that kick in if there is no compromise are not as painful for the economy as a failure to increase the debt limit. This seems to be the idea. I have a couple of points.
First, the Democrats partly caved this time around because they feared the Tea Party wing of the Republican House members were “crazy types” who were willing to destroy ratings of American Treasury bonds to get dramatic spending cuts. Next time around, the threat of disagreement has to police the Tea Partiers. Do they really care about defense? If they care about small government and less foreign intervention, they may actually want defense spending cuts. To get these guys to compromise, the disagreement point should have included libertarian-unfriendly policies. Federally mandated rules that everyone should brush their teeth twice a day, extra additives in drinking water, gun control laws and perhaps a constitutional amendment to eliminate the right to bear arms. Stuff like that should have been in the disagreement point.
Second, the parties face a choice of whom to put onto the SuperCongress. Each party will have six members, three drawn from each chamber. The strategic problem is fairly familiar as it resembles Schelling’s discussion of delegated bargaining. Each party has the incentive to appoint extreme members with tough bargaining stances so the other side will be more likely to give in. Paul Ryan is an obvious choices for the Republicans. Dick Durbin is an obvious choice for the Democrats. If both parties pursue this strategy, there will be deadlock and the disagreement point will come out of the SuperCongress (another example of Prisoner’s Dilemma everywhere).
This analysis is normative – it does not account for strategic errors. And there have been plenty of those. During the health care negotiations we had Max Baucus fruitlessly pursuing his Republican friends trying to get them to sign on. Olympia Snowe got a lot of one-one-one face time with the President. As Krugman points out (see also Jon Stewart a couple of nights ago!), the debt limit extension could have been folded into the extension of the Bush tax cuts last December but the President believed John Boehner at his word. So, my guess is that while Nancy Pelosi and the Republicans will follow the rational choice predictions because they are quite clearheaded, Harry Reid will try to forge a bipartisan compromise. Max Baucus is on (and Kent Conrad would have been if he were not retiring). Ben Nelson is a maybe. Why not go take the extra step Harry and nominate Susan Collins or Olympia Snowe to show that you mean well? With that kind of strategy, the Bush tax cuts may get extended again as part of the SuperCongress compromise and remarkably the Democrats might be forced to implement a compromise that is worse for them than the disagreement point.
If you are programming a robot to vacuum your floors here’s one thing you would never consider doing: endow the robot with feelings of happiness and sadness and teach it to be happy when the floor is clean and unhappy when it is dirty.
But evolution led us to a state of affairs where emotions are what motivate us to do our jobs. How could such a kludge arrive.
Here is a story. Primitive organisms are reproduction machines. They need a certain chemical in the environment, and when they can obtain that fuel they can reproduce.
So the most successful primitive organisms are those that are the best at finding fuel. Natural selection favors those that seek fuel.
Next the organisms get more complicated. They have to make decisions that involve more than just immediate reproduction. They have intertemporal tradeoffs, multi-dimensional consumption, etc.
There is infrastructure in place to simplify this transition. The organisms that have survived to this stage are the organisms that seek fuel. They have built and learned systems for doing what is necessary to get fuel. So fuel is a simple and effective incentive mechanism.
The organism could evolve a mechanism for storing and later releasing fuel. Fuel is released when the organism takes certain actions. Fuel-seeking organisms will take those actions. Natural selection will favor the organisms that release fuel for the right actions.
Now remember that fuel is the energy needed for the most primitive functions of the organism. When this fuel is released the organism gets a boost of that energy.
A boost of energy is a big part of what we call happiness.
(subject to the usual disclaimer, this is based on some conversations with Balasz Szentes.)
From Paul Kedrosky, via Mallesh Pai:
In the game of Scrabble, letter tiles are drawn uniformly at random from a bag. The variability of possible draws as the game progresses is a source of variation that makes it more likely for an inferior player to win a head-to-head match against a superior player, and more difficult to determine the true ability of a player in a tournament or contest. I propose a new format for drawing tiles in a two-player game that allows for the same tile pattern though not the same board to be replicated over multiple matches, so that a players result can be better compared against others, yet is indistinguishable from the bag-based draw within a game. A large number of simulations conducted with Scrabble software shows that the variance from the tile order in this scheme accounts for as much variance as the different patterns of letters on the board as the game progresses. I use these simulations as well as the experimental design to show how much various tiles are able to affect player scores depending on their placement in the tile seeding.
Alternatively you could just let the market prices tell you.
The right to remain silent is not necessarily a blessing to a defendant. Because having a choice is not necessarily a good thing. Unless the decision to testify is uncorrelated with guilt, that decision by itself will convey information to the jury. (I know juries are instructed not to infer anything. But that is impossible.) So for example, if those who take the fifth are more likely to be guility (as I would guess. Are there data on this?), then an innocent person’s “right” to remain silent is actually a right to partially incriminate himself
A prohibition against defendants testifying on their own behalf is worth considering. If the goal is to protect defendants from incriminating themselves, then the above benefit offsets the obvious cost.
And if that is too extreme there are middle grounds to consider. For example, since the defense puts on its case last, the defendant does not make his decision until after the prosecution has introduced evidence. A defendant might want to commit in advance of that evidence being revealed that he will not testify so that nothing can be revealed by his decision being contingent on the evidence. As far as I know this commitment is not possible under current law.
In general the earlier you commit not to testify the less can be inferred from this. So we should allow citizens to register their commitments before they are charged with any crime. When you register to vote you also check a box that says whether you or not you will testify in the event you are ever charged with a crime.
Stan Reiter had a standard gripe about statistics/econometrics. Imagine you there is a cave in front of you and you want to map out its dimensions. There are many ways you could do it. One thing you could do is go inside and look. Another thing you could do is stand outside and throw into the cave a bunch of super bouncy balls and when they bounce out, take careful note of their speed and trajectory in order to infer what walls they must have bounced off of and where. Stan equated econometrics with the latter.
That’s not what I am going to say but it is a funny story and its the first thought that came to my mind as I began to write this post.
But I do have something, probably even more heretical, to say about econometrics. Suppose I have a hypothesis or a model and I collect some data that is relevant. If I am an applied econometrician what I do is run some tests on the data and report the results of the tests. I tell you with my tests how you should interpret the data.
My tests don’t contain any information in them that isn’t in the raw data. My tests are just a super sophisticated way to summarize the data. If I just showed you the tables it would be too much information. So really, my tests do nothing more than save you the work of doing the tests yourself.
But I pick the tests. You might have picked different tests. And even if you like my tests you might disagree with the conclusion I draw from them. I say “because of these tests you should conclude that H is very likely false.” But that’s a conclusion that follows not just from the data, but also from my prior which you may not share.
What if instead of giving you the raw data and instead of giving you my test results I did something like the following. I give you a piece of software which allows you to enter your prior and then it tells you what, based on the data and your prior, your posterior should be? Note that such a function completely summarizes what is in the data. And it avoids the most common knee-jerk criticism of Bayesian statistics, namely that it depends on an arbitrary choice of prior. You tell me what your prior is, I will tell you (what the data says is) your posterior.
Pause and notice that this function is exactly what applied statistics aims to be, and think about why, in practice, it doesn’t seem to be moving in this direction.
First of all, as simple as it sounds, it would be impossible to compute this function in all practical situations. But still, an approach to statistics based on such an objective, and subject to the technical constraints would look very different than what is done in practice.
A big part of the explanation is that statistics is a rhetorical practice. The goal is not just to convey information but rather to change minds. In an imaginary perfect world there is no distinction between these goals. If I have data that proves H is false I can just distribute that data, everyone will analyze it in their own favorite way, everyone will come to the same conclusion, and that will be enough.
But in the real world that is not enough. I want to state in clear, plain language terms “H is false, read all about it” and have that statement be the one that everyone focuses on. I want to shape the debate around that statement. I don’t want nuances to distract attention away from my conclusion. In the real world, with limited attention spans, imperfect reasoning, imperfect common-knowledge, and just plain old laziness, I can’t get that kind of focus unless I push the data into the background and my preferred intepretation into the foreground.
I am not being cynical. All of that is true even if my interpretation is the right one and the most important one. As a practical matter if I want to maximize the impact of the truth I have to filter it.
Still it’s useful to keep this perspective in mind.
I read the transcript and it is a very eloquent clarification of his views on game theory’s role and even the game theorist’s role. Worth checking out.
