RIP Lloyd Shapley

I just saw a post on Facebook by economics Nobel prize recipient Vernon Smith with the information that Lloyd Shapley, a giant in game theory, has died. I am sharing a screenshot of the post and the first comment, which is quite something.

Screenshot 2016-03-12 15.21.28.png

On having finished the book Sapiens

I have written here about reading Harari’s book Sapiens. I have finished the book since that post appeared. Now, separated from the time of finishing the book by a few days, in which I spent a lot of time reading graduate student dissertation proposals and writing lecture notes, I thought I would write down my lingering impression from the book.

This impression is bleak. The various revolutions Harari talks about left individual members of Homo Sapiens less well off than before, with the Agricultural Revolution as a prime example. Even worse were the effects of these revolutions (Cognitive, Agricultural, Industrial) on other species on planet Earth.

A secondary impression I got was one of a fundamental tension between Harari’s portrayal of history as proceeding without regard to what is good or bad, for humans or other species, and constantly talking about effects of historical changes as good or bad. I am perfectly content to read normative statements in a book on history (or economics, as a matter of fact), but I want a clearer idea of the author’s ethical convictions. Harari does not elucidate a moral philosophy, but one seems to be in the background, one that I would have liked to be made clearer.

The new Greek agreement

So it looks like the EU “institutions” (the new name for the Troika) have agreed to the Greek list of reforms and, as a result, for the time being Greece avoids a crash. I have read many comments on line. I tend to gravitate to the commentary by Frances Coppola (here is an example). But I don’t have a huge amount to add myself to the debate right now. I will end this brief post by sincerely wishing that SYRIZA will succeed in reducing corruption and tax evasion. Just like so many people from the rest of Europe, however, I remain highly skeptical about this.

Criticism of Varoufakis’s NYT column

So Yanis Varoufakis has been teaching game theory and has some publications in the area. But his New York Times column a little while ago does not impress in terms of his use of game theory. I had a mind to write about this, but I did not make the time to do so yet, and now I have been scooped by Rakesh Vohra. So be it; a couple of typos aside, Vohra has said some some important things and should be read widely.

Greece in the penalty box?

I wrote yesterday about the latest moves in the game between the new Greek government and its European creditors. Today I read this excellent piece by Dan Davies. His point is that the ECB has a lot of power and is not afraid to use it, because it can punish Greece without going all the way to kicking it out of the Euro, which would have some costs for the ECB and Europe in general. It comes down to what is the perception of the game the players think they are playing, as I mentioned yesterday. Davies thinks it’s possible that Varoufakis has misunderstood the real nature of the game: he thought it was brinkmanship, but it may be something else, where the ECB can severely punish Greece without going all the way to a Grexit.

Back in 1967-68, John C. Harsanyi formulated, in an epic three-part paper, the concept of Bayesian Nash equilibrium (for which he shared a Nobel prize). In this formulation of a simultaneous-move game, the players may be uncertain about the players who are in the game, the strategies the players can play, and the other players’ payoffs. Yet it is possible to introduce the concept of a type of a player, that encompasses all these elements of uncertainty, resort to a Bayesian framework with a common prior for all players, and create an equilibrium straight out of Nash equilibrium by treating the types of the players as the players in a new, artificial, game that matches with the original one in important (if too technical for this blog post) ways.

I can imagine that Varoufakis thinks he’s got the right conception of the types of the other players he’s facing in the current game he’s in. But if he’s missed some possibilities, by assigning zero probability to certain strategies, for example, as the moves that Dan Davies describes, then while he thinks he’s playing to a Bayesian Nash equilibrium, he may find that the outcome does not match his expectation, because the strategy followed by his opponents has exploited aspects of the game that he did not realize were there. He may be a professor of game theory, but it may be hubristic of him to think he can outmaneuver the game theorists employed by outfits like the ECB.

When the Finance minister is a game theorist

Events after the election in Greece, related as they are to game theory, have prompted me to revive this moribund blog. I promise I will be updating it more frequently from now on. So here is my first take on this blog of the negotiations that Greece and the Euro zone authorities are engaged in now.

Ever since the election of January 26, 2015 in Greece, an elaborate confrontation has been evolving between the new Greek government and the European Central Bank (ECB) and the International Monetary Fund. Justin Fox, writing in Bloomberg View, said that the new Greek Finance Minister, Yanis Varoufakis (here is his blog), is perhaps playing a Colonel Blotto game with the European governments, spreading the fight in many fronts. Then again, it could just be a classic Prisoner’s Dilemma, as Fox also says.

Today, the ECB released a statement that appears intended to intimidate the Greek side. But Karl Whelan is not too impressed. He points out that the ECB has not been following any clear rules in deciding what to accept as collateral for lending to banks in the Euro zone. Whelan takes this as a nudge towards more productive negotiations to start before February is out. In Whelan’s words:

The ECB is flexing its muscles, letting everyone know that are very close to pulling liquidity from Greece.

I want to point out that applying game theory in a negotiation is a tricky business. You have to assume that the sides to the negotiation have a shared mental model (which is the same as that of the analyst, you) of what the game they are playing is. So you can analyze the situation, make a prediction, have the prediction subsequently falsified by events, and yet blame the predictive failure on the many degrees of freedom in setting up the game for your analysis.

I will be sure to follow developments in this arena closely. There ought to be common ground that the sides of the negotiation can find. Whether the tactics Varoufakis and his colleagues have adopted will find the common ground and lead to a good outcome for all sides remains to be seen. I am not going on a limb to make predictions.