Abstract
We consider computational games, sequences of games \(\mathcal {G}=(G_1,G_2,\ldots )\) where, for all n, \(G_n\) has the same set of players. Computational games arise in electronic money systems such as Bitcoin, in cryptographic protocols, and in the study of generative adversarial networks in machine learning. Assuming that one-way functions exist, we prove that there is 2-player zero-sum computational game \(\mathcal {G}\) such that, for all n, the size of the action space in \(G_n\) is polynomial in n and the utility function in \(G_n\) is computable in time polynomial in n, and yet there is no \(\epsilon \)-Nash equilibrium if players are restricted to using strategies computable by polynomial-time Turing machines, where we use a notion of Nash equilibrium that is tailored to computational games. We also show that an \(\epsilon \)-Nash equilibrium may not exist if players are constrained to perform at most T computational steps in each of the games in the sequence. On the other hand, we show that if players can use arbitrary Turing machines to compute their strategies, then every computational game has an \(\epsilon \)-Nash equilibrium. These results may shed light on competitive settings where the availability of more running time or faster algorithms can lead to a “computational arms race”, precluding the existence of equilibrium. They also point to inherent limitations of concepts such as “best response” and Nash equilibrium in games with resource-bounded players.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
- 2.
The games used to model protocols such as Bitcoin are actually extensive-form games, which are played over time. Our impossibility results show that there are computational Bayesian games where there is no NE when we restrict to polynomial-time players. Since Bayesian games are a special case of extensive-form games, our non-existence results carry over to extensive-form games.
- 3.
It is also possible to allow k to depend on n, but we focus on the case where k is a constant for concreteness.
- 4.
We restrict our attention to utilities and probabilities that are rational numbers.
- 5.
One question is how to deal with players who use Turing machines that fail to halt or return an action that does not belong to the action space. We deal with this issue by assigning to each player i a special action \(a_0^i\) that we take to be the action played if i’s TM does not halt or if i’s output is not an action in the action space. Any profile that includes \(a_0^i\) gives utility \(-\infty \) to all players, thus discouraging players from using TMs that fail to halt or return inappropriate actions.
- 6.
Our results also hold in a more general setting where the absolute value of a (nonzero) utility is at most polynomial and at least inversely polynomial in n.
References
Ben-Sasson, E., Tauman-Kalai, A., Kalai, E.: An approach to bounded rationality. In: Proceedings of the 19th Neural Information Processing Systems Conference, pp. 145–152 (2007)
Brassard, G., Chaum, D., Crépeau, C.: Minimum disclosure proofs of knowledge. J. Comput. Syst. Sci. 37, 156–189 (1988)
Chen, X., Deng, X., Teng, S.H.: Settling the complexity of two-player Nash equilibrium. J. ACM 53(3) (2009)
Courtois, N.T., Bahack, L.: On subversive miner strategies and block with holding attack in bitcoin digital currency (2014). arXiv preprint: http://arxiv.org/abs/1402.1718
Daskalakis, C., Goldberg, P.W., Papadimitriou, C.H.: The complexity of computing a Nash equilibrium. In: Proceedings of the 38th ACM Symposium on Theory of Computing, pp. 71–78 (2006)
Dodis, Y., Halevi, S., Rabin, T.: A cryptographic solution to a game theoretic problem. In: Bellare, M. (ed.) CRYPTO 2000. LNCS, vol. 1880, pp. 112–130. Springer, Heidelberg (2000). https://doi.org/10.1007/3-540-44598-6_7
Feigenbaum, J., Koller, D., Shor, P.W.: A game-theoretic classification of interactive complexity classes. In: Proceedings of the Structure in Complexity Theory Conference, pp. 227–237 (1995)
Fortnow, L., Impagliazzo, R., Kabanets, V., Umans, C.: On the complexity of succinct zero-sum games. Comput. Complex. 17(3), 353–376 (2008)
Fortnow, L., Santhanam, R.: Bounding rationality by discounting time. In: Proceedings of Innovations in Computer Science Conference, pp. 143–155 (2010)
Halpern, J.Y., Pass, R.: Algorithmic rationality: game theory with costly computation. J. Econ. Theory 156, 246–268 (2015). https://doi.org/10.1016/j.jet.2014.04.007
Halpern, J.Y., Pass, R., Reichman, D.: On the nonexistence of equilibrium in computational games (2019)
Halpern, J.Y., Pass, R., Seeman, L.: Computational extensive-form games. In: Proceedings of 17th ACM Conference on Electronic Commerce (EC 2016), pp. 681–698 (2016)
Holenstein, T.: Pseudorandom generators from one-way functions: a simple construction for any hardness. In: Halevi, S., Rabin, T. (eds.) TCC 2006. LNCS, vol. 3876, pp. 443–461. Springer, Heidelberg (2006). https://doi.org/10.1007/11681878_23
Lipton, R.J., Markakis, E.: Nash equilibria via polynomial equations. In: Farach-Colton, M. (ed.) LATIN 2004. LNCS, vol. 2976, pp. 413–422. Springer, Heidelberg (2004). https://doi.org/10.1007/978-3-540-24698-5_45
Maschler, M., Solan, E., Zamir, S.: Game Theory. Cambridge University Press, Cambridge (2013)
Megiddo, N., Wigderson, A.: On play by means of computing machines. In: Theoretical Aspects of Reasoning About Knowledge: Proceedings of 1986 Conference, pp. 259–274 (1986)
Nakamoto, S.: Bitcoin: a peer-to-peer electronic cash system (2008). http://www.bitcoin.org/bitcoin.pdf
Nash, J.: Non-cooperative games. Ann. Math. 54, 286–295 (1951)
Neyman, A.: Bounded complexity justifies cooperation in finitely repeated prisoner’s dilemma. Econ. Lett. 19, 227–229 (1985)
Neyman, A.: The positive value of information. Games Econ. Behav. 3, 350–355 (1991)
Neyman, A.: Finitely repeated games with finite automata. Math. Oper. Res. 23, 513–552 (1998)
Oliehook, F., Savani, R., Gallego, J., van der Poel, E., Gross, R.: Beyond local Nash equilibria for adversarial networks (2018). http://arxiv.org/abs/1806.07268
Papadimitriou, C.H., Yannakakis, M.: On complexity as bounded rationality. In: Proceedings of 26th ACM Symposium on Theory of Computing, pp. 726–733 (1994)
Rubinstein, A.: Finite automata play the repeated prisoner’s dilemma. J. Econ. Theory 39, 83–96 (1986)
Schoenebeck, G., Vadhan, S.: The computational complexity of nash equilibria in concisely represented games. Theory Comput. 4, 270–279 (2006)
Simon, H.A.: A behavioral model of rational choice. Quart. J. Econ. 49, 99–118 (1955). https://doi.org/10.2307/1884852
Wee, H.: On obfuscating point functions. In: Proceedings of the 37th ACM Annual Symposium on Theory of Computing, pp. 523–532 (2005)
Acknowledgments
Halpern was supported in part by NSF grants IIS-178108 and IIS-1703846, a grant from the Open Philanthropy Foundation, ARO grant W911NF-17-1-0592, and MURI grant W911NF-19-1-0217. Pass was supported in part by NSF grant IIS-1703846. Most of the work was done while Reichman was a postdoc at Cornell University.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Halpern, J.Y., Pass, R., Reichman, D. (2019). On the Existence of Nash Equilibrium in Games with Resource-Bounded Players. In: Fotakis, D., Markakis, E. (eds) Algorithmic Game Theory. SAGT 2019. Lecture Notes in Computer Science(), vol 11801. Springer, Cham. https://doi.org/10.1007/978-3-030-30473-7_10
Download citation
DOI: https://doi.org/10.1007/978-3-030-30473-7_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-30472-0
Online ISBN: 978-3-030-30473-7
eBook Packages: Computer ScienceComputer Science (R0)