Answer: This insightful question in fact anticipates several ways in which game theory -- at least as applied to political problems -- has evolved in recent decades beyond the very simple models we have focused on.

One aspect of this evolution has been to incorporate bargaining moves (e.g., messages, threats, offers, concessions, etc.) into the formal definition of the game, with the result that the number of strategies available to both players is multiplied. In fact the numbers of strategies tends to multiply so quickly that the payoff matrix form [more formally known as the normal or strategic form] that we have focused on quickly becomes unwieldy. In its place, one can use the so-called extensive form (what Dixit & Nalebuff, pp. 34ff, call a game tree) is used. If all moves (even in a long sequence of moves) are made openly (so that each player sees what the other is doing as soon as he does it, or at least before he has to make his next move), then D&N's look ahead and reason back logic [technically called backwards induction] still holds. Therefore, if we able to draw the game tree and assign payoff to all possible outcomes, we can "solve" the game, i.e., identify the best [pure] strategy for each player and thus identify the outcome that results if the players play "rationally."

Another aspect of this evolution takes account of the (very realistic) possibility that often players don't really know each other's payoffs -- that is, their goals, interests, values, and (especially important in games line Chicken) willingness to take risks. Thus a new theory of games with incomplete information has developed over past decade of two, in which players may be are incompletely informed about the characteristics of other players. Thus the theory must address the beliefs that each player has about the other, which of course may be affected by how the other player plays. Such a theory aims at an outcome in which not just strategies but also beliefs are in equilibrium.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Question:  From the list of POLI 388 question: "Why is it possible to extract advantages by holding hostages but also inherently difficult to realize these advantages?"  I would answer the first part by saying how in the hostages' dilema, if there isn't organization among the hostages, the hostage-holder can manipulate and blackmail them. But if there were to be an organized revolt (eg, a workers' union) the hostage-holder would lose his/her power.

Answer:  Your answer makes (good) observations about the game involving the hostages and the hostage-holders.  However, the point of this question is to focus on the game between the hostage holders and the authorities they are making demands on.  In this respect, the basic point is that the hostage holders' bargaining power derives from the fact that they hold hostages and can harm/kill them.  This can be used as a basis for (deterrent or compellent) threats against the authorities.  The problem is that it is never in the self-interest of the hostage holders actually to carry out this threat (because then they have no more bargaining power and have lost), so the credibility of their threat is always in doubt.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Question:  What do you mean by decomposing the execution of a threat?

Answer:  Decomposing (the execution of) a threat means carrying it out gradually or bit at a time, rather than all at once -- for example, hostage-holders executing one hostage a day (to try to escape the cedibility problem noted above.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Question:  What are tactics of decommitment?  Are they the same as what Schelling in Arms and Influence, pp. 63-66, refers to as escaping commitments?

Answer:  Yes, the same -- I shouldn't have switched terminology.  The point is that if it becomes necessary to act contrary to a previously made commitment, you want to try to accomplish this in in a way that minimizes potential damage to your reputation (as an actor who generally lives up to his commitments).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Question:  What is the distinction between tactical and  strategic surprise.  Is it is basically related to the difference between deterrence and actual conflict. For example, is a tactical surprise designed on the premise that there will actually be fighting and
the surprise is designed to have one be the victor? 

Answer:  A strategic surprise is an attack that comes "out of the blue" -- like Pearl Harbor (presumably).  A tactical surprise is when you know/expect an attack is coming but don't know exactly when or where.  D-Day was a tactical surprise, but not strategic surprise, to the Germans.  (Clearly the tactical-strategic distinction is a "more or less," rather than "either/or," distinction.)  An attack that is carrying out a (deterrent or compellent) threat cannot be a strategic surprise (unless maybe the threat had been totally discounted by the threatened party).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Question:  What do you have in mind with the "impact of nuclear weapons (plus long-range delivery systems) on international politics" item on the Review List?

Answer:  The  "impact of nuclear weapons (plus long-range delivery systems) on international politics" item on the Review List was intended primarily to remind you of Schelling's discussion in Chapter 1 of ARMS & INFLUENCE on "The Diplomacy
of Violence,"  which I also briefly recapitualted in class (probably on May 5).  He begins to characterize the "nuclear revolution" (i.e., the impact of nuclear weapons on international politics) on p. 18 but his comments on pp. 22-23 are
especially pertinent.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Question:  What is controlled response (from the Review List for the Final)?

Answer:   By being shaded in the revised review list, I meant to indicate that you are "not responsible" for this item.  Generally, controlled response  refers to the possibilty that the U.S. (or S.U.) might respond to nuclear attack attack on its homeland
(or some other extraordinary preovocation) with a nuclear response than is something less than all-out nuclear retaliation.  (Secretary McNamara's "no cities doctrine" [Schelling, A&I, pp. 154, 162-156] was a variant of controlled response.)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~