?
Computing lexicographically safe Nash equilibria in finite two-person games with tight game forms given by oracles
In 1975 the first author proved that every finite tight two-person game form g is Nashsolvable,
that is, for every payoffs u and w of two players the obtained normal form game
(g; u,w) has a Nash equilibrium (NE) in pure strategies. Several proofs of this theorem
were obtained later. Here we strengthen the result and give a new proof, which is shorter
than previous ones. We show that game (g; u,w) has two types of NE, realized by a
lexicographically safe (lexsafe) strategy of one player and some special best response of
the other. The proof is constructive, we obtain a polynomial algorithm computing these
lexsafe NE. This is trivial when game form g is given explicitly. Yet, in applications g is
frequently realized by an oracle O such that size of g is exponential in the size |O| of
O. We assume that game form g = g(O) generated by O is tight and that an arbitrary
±1 game (g; u0,w0) (in which payoffs u0 and w0 are zero-sum and take only values ±1)
can be solved in time polynomial in |O|. These assumptions allow us to compute two
(one for each player) lexsafe NE in time polynomial in |O|. These NE may coincide. We
consider four types of oracles known in the literature and show that all four satisfy the
above assumptions.