Pseudo-Equilibria, or: How to Stop Worrying About Crypto and Just Analyze the Game
Alexandros Psomas, Athina Terzoglou, Yu Wei, Vassilis Zikas · 2025
Read Pseudo-Equilibria, or: How to Stop Worrying About Crypto and Just Analyze the Game by Alexandros Psomas, Athina Terzoglou, Yu Wei, Vassilis Zikas (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-22089.txt.
Introduction
Alexandros Psomas, Athina Terzoglou, and co-authors coin a loose rest state for games that run on code locks, in a 2025 paper. Game minds want to treat locks as pure black boxes in play maths. Real locks rest on world facts or crowd trust, and the two kinds part in key ways.
For crowd trust locks, pure box play proves out of reach in full. For world fact locks, stock rest states fail to cross from box to real. The loose state asks that a gain from a switch be hard to tell from none. Pure rest states map to loose ones when real locks fill the box. Sides can thus be read apart.
Catalog record
| CREATOR | Alexandros Psomas, Athina Terzoglou, Yu Wei, Vassilis Zikas |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 148 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-22089
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Pseudo-Equilibria, or: How to Stop Worrying About Crypto and Just Analyze the Game Alexandros Psomas, Athina Terzoglou, Yu Wei, Vassilis Zikas Abstract We consider the problem of a game theorist analyzing a game that uses cryptographic protocols. Ideally, the game theorist should be able to ignore all implementation details of the cryptographic protocols, and abstract them as ideal, implementation-independent primitives, in a way that conclusions in the “ideal world” can be faithfully transferred to the “real world,” where real protocols are implemented by cryptography. Achieving this goal is crucial, as the game theorist cannot — and should not be expected to — grapple with the full complexity of cryptographic implementations. This is particularly relevant in the era of Web3, where the widespread adoption of distributed ledgers has created a pressing need for a common language that bridges cryptography and game theory. The security of cryptographic protocols hinges on two types of assumptions: state-of-the-world assumptions (e.g., “one-way functions exist”, “factoring is hard”, etc.), and behavioral assumptions (e.g., “honest majority”). We observe that for cryptographic protocols that rely on behavioral assumptions (e.g., ledgers), our goal is unattainable in its full generality: the game theorist cannot always abstract away all implementation details. For state-of-the-world assumptions, we show that standard solution concepts, e.g., Nash equilibria and ϵ \epsilon -Nash equilibria, are not robust to the transfer from the ideal setting to the real world. In this paper, we propose a new solution concept: the pseudo-Nash equilibrium . Informally, a strategy profile s = ( s 1 , … , s n ) s=(s_{1},\dots,s_{n}) is a pseudo-Nash equilibrium if, for every player i i and every unilateral deviation s i ′ s^{\prime}_{i} with a larger expected utility, the utility of i i when playing s i s_{i} is (computationally) indistinguishable from her utility when playing s i ′ s^{\prime}_{i} . Pseudo-Nash is substantially simpler and more accessible to game theorists than any existing notion that attempted to address the mismatch in the (asymptotic) cryptographic method and game theory. We prove, in a very general sense, that Nash equilibria in games that use idealized, unbreakable cryptography correspond naturally to pseudo-Nash equilibria when idealized cryptography is instantiated with actual protocols (under state-of-the-world assumptions). Our translation is not only conceptually simpler than existing approaches, but also more general: it does not require tuning or restricting utility functions in the game with idealized cryptography to accommodate idiosyncrasies of cryptographic implementations. In other words, pseudo-Nash equilibria allow us to separately and seamlessly study game-theoretic and cryptographic aspects. 1 Introduction Cryptography and game theory both study interactions among agents. Cryptography traditionally takes an adversarial approach; protection is demanded against the worst-case (mis)behavior of the agents. Game theory takes the view that agents are selfish and aim to optimize their own utility. The widespread adoption of blockchain protocols has set a stage where both viewpoints are invaluable, as decentralized systems must be both strategically robust and cryptographically secure. However, while the interaction of the two fields has been fruitful, we believe it is fair to say that a universally adopted model that allows for the two fields to interact seamlessly has yet to emerge. Arguably, the most ambitious goal would be a model such that game theorists can work in an “ideal world” that completely ignores all implementation details of cryptographic protocols (i.e., treats cryptographic primitives as black boxes that provide a certain functionality), but such that game-theoretic solution concepts, e.g., equilibria, naturally carry over to the “real world,” where cryptographic primitives are replaced by their (imperfect) implementations. Such a model, by design, would also allow cryptographers to implement cryptographic primitives without worrying about downstream effects on game theoretic predictions. Simply put, our work attempts to answer the following question: Is this ambitious goal attainable? 1.1 Our contributions Unfortunately, the answer to the aforementioned question is, in general, “no.” This is because cryptographic protocols provide guarantees only under assumptions. There are, broadly speaking, two types of assumptions cryptographic protocols require. The first type is state-of-world assumptions; examples of this type are computational hardness, ideal obfuscation, or secure hardware, which are common assumptions in the design of cryptographic protocols like encryption, digital signatures, commitments, etc. The second type is behavioral assumptions (e.g., “honest majority”), which are assumptions about the parties that implement the protocol at hand. The reason we cannot have a model that seamlessly abstracts away all cryptographic details, but is faithful to game-theoretic solution concepts, is because of this second type. Many central solution concepts in game theory, e.g., Nash equilibria, are predictions about players’ behavior. Therefore, intuitively, it is not possible to ignore elements of a game that only work under specific player behavior. For a concrete example, consider the following. Example 1.1 (Games on a Ledger) . Consider the problem of a miner on a blockchain. The blockchain is an implementation of cryptographic primitives known as distributed ledgers, which, loosely speaking, provide the following functionality: every party can provide inputs to the ledger which is recorded on a future block; every party can receive the ledger’s current state. 1 1 1 We note that we oversimplify the functionality, in order to make our argument. Examples of ledger functionalities implemented by mainstream blockchain protocols like Bitcoin and Ouroboros/Cardano—and corresponding proofs of security—can be found in [ BMTZ17 , BGK+18 ] . The blockchain implements the ledger functionality assuming an honest majority of the participants, or of a resource held by the participants like hashing power (in proof of work) or stake (in proof of stake); the miners. Given an ideal ledger, consider a simple game between miners, where a miner gains a large reward R when doing a task quickly, and a small reward r when doing the task slowly, where “time” is measured by the timestamps on blocks. And, for the sake of exposition, assume that the miner pays a large cost when doing the task quickly, but no cost when doing the task slowly, and thus the miner’s best response under these rules is to do the task slowly. In the context of DeFi (Decentralized finance), [ YTZ22 ] prove that, in certain blockchain protocols, it is possible for miners to deviate in a way that they can manipulate the timestamps of blocks. For our simple game, this would mean that the miner could claim the large reward R , contradicting the prediction in the ideal game. We note that, starting from [ ES18 ] , numerous works [ BW24 , BCNPW19 , CKWN16 , FHWY22 , FGH+24 , FW21 , KKKT16 , FKKP19 , GS19 , SSZ17 , TE18 ] show how, under several different classes of blockchain protocols, there are several ways that miners can manipulate their behavior in the protocol to manipulate timestamps, their rewards in the underlying cryptocurrency, or even the value of the cryptocurrency itself. Thus, the existence of these counterexamples is well-known (but perhaps not articulated using the ideal vs real-world perspective we take in this work). Since our goal is unattainable in general, the next natural question is whether it is attainable under conditions. And, given our discussion so far, a natural condition would be the usage of cryptographic protocols that make state-of-the-world assumptions. Our next observation is that standard game-theoretic