Readings: Games, Brexit, qMRI, Fund Managers, etc.

While game theory is (at least within the academic community) largely considered a failure, that doesn't mean it is without its uses. I was thinking of it this morning in reading a recent FT piece about misunderstandings between the European Union and Brexiteers in Britain.

Here is Simon Kuper:
Both the EU and the British government keep making the same mistake about each other, notes Douglas Webber of Insead business school, author of European Disintegration? (2016): each side thinks the other will cave to avoid an economically damaging no-deal Brexit.
In fact, says Webber, both sides regard short-term economics as secondary. Johnson’s government prioritises achieving Brexit. Europeans prioritise preserving the rules of the single market and standing by Ireland.
Source: How Europe views the Brexit endgame, Simon Kuper

This is, of course, a kind of prisoner's dilemma, where both parties think that the other side is so worried about its counterpart defecting -- embrace a no-deal Brexit -- that the other side will "see sense" and negotiate in good faith toward a new post-Brexit arrangement.
The trouble with this sort of symmetric game is not the symmetry, however. Among the useful things we have learned from game theory, in general, and the prisoner's dilemma, in particular, especially from competitions based on game theory, is that it matters whether the game (rightly or wrongly) is seen as repeated or one-off. If a game is a one-off, then vengeful and retributive strategies can do surprisingly well. After all, who cares how much you maximize your own gains at another party's expense if you never have to deal with them again? Defect away!

Despite appearances, however, Brexit is not a one-off game. While Britain admittedly won't be Brexit-ing in future on a yearly basis, it will still have to deal with the EU constantly. Its trade, immigration, and political relationships are too strong and important. If anything, post-Brexit, it would be talking with the EU almost as much, albeit with a different focus and status. This is more like a repeated game, in prisoner's dilemma terms, which might make one optimistic that the two sides will sort things out.

But surprisingly, perhaps, even in repeated games, however, fairly vicious strategies can outperform. Famously, a strategy called "tit-for-tat" did very well in many competitions, where it followed the simple strategy of doing whatever its counterpart last did. This, of course, could lead to catastrophic spirals if two such strategies played one, a kind of ultimatum game where both sides "defect" into oblivion. (This has led to more nuanced versions, like where a strategy periodically tries to cooperate, but then goes back to defecting if the other side doesn't switch strategies.)

Where does this leave us? Both sides seem stuck on the idea of convincing the other of its respective commitment to the idea that the other doesn't want a no-deal Brexit. But it also seems clear both think a no-deal Brexit wouldn't be that bad, both politically and economically. Further, both see some political capital in acrimonious discussion, as weird as that might seem. It leaves me, at least, with the nagging sense that both sides will defect -- a no-deal Brexit -- and things will get very messy indeed.

<><><><><>><><><><>><><><><>><><><><>><><><><>

A few papers worth reading:
Advances in qMRI (quantitate MRI) are fascinating, but they have been held back somewhat by MRI's sensitivity to water content, which, while regulated fairly closely, changes enough to make comparative tissue characterization difficult. A new approach does away with some of these problems, allowing, for the first time, allowing doctors to track molecular changes in the brain over time, thus permitting a deeper understanding of aging-related changes. Important work.
While colorectal cancer incidence is no longer rising among older cohorts, it is, according to a new study, rising sharply among the young, which is a surprise. Some of the increase may have to do with more sensitive detection methods, but this is seemingly more going on than that, with an increase in young adults twice as rapid as that in older adults. Researchers speculate about a new source of large intestine carcinogenesis, but it's cautionary and early work.
Collapse remains, historically speaking, an important source of flux in the global environment. A new study shows that the collapse of the Soviet Union resulted in a net emissions reduction of 7.61 gigatons of carbon dioxide equivalents from 1992 to 2011. By way of comparison, this is about one-quarter of the CO2 emissions from deforestation in Latin America over the same period.
Most of the attention paid to poor performance among active fund managers has been directed to how damn hard it is, not how the industry is structured. A new study takes a structural view, arguing, somewhat paradoxically, that the industry, despite declining in terms of managers and assets, still isn't concentrated enough, that having too many fund managers reduces the incentive for outperformance.
Decades from now we will look back on this period and likely realize we were kids playing with information matches when it comes to the onslaught of often siloed info-flow overwhelming people, especially in politics. A new paper argues that this has created both tacit and explicit opportunities for what it calls "information gerrymandering", essentially a way, based on connections and zealots, to prevent people from seeing information that might provide them with a different perspective. This is intriguing work with considerable explanatory power.

Finally, some quick hits: