Thursday, February 16, 2017

Gaming It Out

You're all in playing Texas hold 'em.  Your opponent is Johnny Chan.  Your hole cards are a bust.  What are you thinking as your chances for glory seem to be disappearing in the moment?

* * * * *

Like many of my friends I am obsessing about our federal government now and its chaos and dysfunction.  Unlike them, however, I have a rigorous understanding of game theory from the days when I did formal economics.  I am going to try to bring that to bear in this post, absent any jargon that readers might find off putting.

Let me begin with E. J. Dionne's piece today, Admit it: Trump is unfit to serve.  I agree with the analysis there.  If everyone else also agreed, then the puzzle would be why the Republicans in Congress seem to be soft pedaling the issue, hoping the ship might right its course.   In his column yesterday, Greg Sargent mainly argued that Trump now looks like a weak autocrat, but then articulated this alternative:

To be sure, Trump is getting a lot of his Cabinet nominees confirmed. It’s likely that Trump and Republicans will win a lot of victories before long, ones that will be very demoralizing to Democrats. It is also true that the White House has at its disposal a tremendous range of tools to take control of events and news cycles, thus turning things around. So all of this might change soon enough. A doubling-down on Trump’s worst policies, perhaps in the form of a newly implemented and then expanded “Muslim ban,” or in the form of stepped-up deportations, remain real possibilities. A terrorist attack could empower Trump and lead to far worse.

The question is whether the above scenario is realistic, where here realistic means there is a significant likelihood of it coming about.  Black Swans (highly unlikely events) do happen.  Some might say that Trump getting elected was a black swan, especially given how the pollsters didn't forecast it.  Even after experiencing a black swan, however, the laws of probability aren't changed.  A game player choosing a strategy will still make the same sort of calculations on what is the right play.   The hole cards (this is a metaphor for any information that the player knows but nobody else knows) matter.

Understanding this, an outside observer can then try to invert the process - by observing the play the observer makes some inference about what the hole cards are likely to be.  In actual poker bluffing is possible, for this very reason.  But theory says the player must be completely inscrutable, so it is the play only that determines the inference, not some affect in the person that provides an additional clue.  (The movie Rounders refers to that additional clue as the tell.)

In the case at hand, there is really no way that a bluffing approach can work with the Democrats in Congress, the mainstream press, and the various spy agencies.  Republicans in Congress surely understand that.  Given that observation, if you make inferences based on their behavior they must feel it somewhat likely that the damage can be contained.  For them it may be important to distinguish malfeasance, which is impeachable, from mere incompetence, which is not. If the President is guilty of the latter only, they may then be able to wait out the storm.

At issue, then, is what do the spy agencies know that has not yet come out in public and what do members of Congress know about that?  Of course, we don't know that.  What we might conclude, however, is that the information is murky.

The Republicans do have incentive to keep it that way rather than admitting to any smoking gun information that can be verified, because they already have their sights on the 2018 midterm elections.  However, if such information came out anyway, they may then quickly change their approach.

We should observe that Watergate offers an imperfect parallel.  The Democrats had majorities in both houses of Congress then, while the President was Republican.  Now they are each Republican and the majority in the House is fairly substantial.

It is surely impossible to query a Republican House member to ask what it would take for them to vote for impeachment.   I won't try to answer that question for them.  I will say, however, that as long as we're not over that bar, the situation will linger pretty much as is.

What should happen and what will happen are two distinct things and we shouldn't confound the one for the other.  So as much as I agree with E.J. Dionne's analysis today, I treat it as normative only, not as predictive.  On the predictive front, the national nightmare will likely be ongoing.

Having it end suddenly would be a black swan.  We can hope for that, but it's not the right way to bet.

No comments: