It was recently observed that Elo ratings fail at preserving transitive relations among strategies and therefore cannot correctly extract the transitive component of a game. We provide a characterization of transitive games as a weak variant of ordinal potential games and show that Elo ratings actually do preserve transitivity when computed in the right space, using suitable invertible mappings. Leveraging this insight, we introduce a new game decomposition of an arbitrary game into transitive and cyclic components that is learnt using a neural network-based architecture and that prioritises capturing the sign pattern of the game, namely transitive and cyclic relations among strategies. We link our approach to the known concept of sign-rank, and evaluate our methodology using both toy examples and empirical data from real-world games.
Competitive online games use rating systems to match players with similar skills to ensure a satisfying experience for players. In this paper, we focus on the importance of addressing different aspects of playing behavior when modeling players for creating match-ups. To this end, we engineer several behavioral features from a dataset of over 75,000 battle royale matches and create player models based on the retrieved features. We then use the created models to predict ranks for different groups of players in the data. The predicted ranks are compared to those of three popular rating systems. O…
DeepMind（深度思维）；University College London（伦敦大学学院）；Google（谷歌）；Carnegie Mellon University（卡内基梅隆大学）
Rating strategies in a game is an important area of research in game theory and artificial intelligence, and can be applied to any real-world competitive or cooperative setting. Traditionally, only transitive dependencies between strategies have been used to rate strategies (e.g. Elo), however recent work has expanded ratings to utilize game theoretic solutions to better rate strategies in non-transitive games. This work generalizes these ideas and proposes novel algorithms suitable for N-player, general-sum rating of strategies in normal-form games according to the payoff rating system. This …
It is known that the generalized Nash equilibrium problem can be reformulated as a quasivariational inequality. Our aim in this work is to introduce a variational approach to study the existence of solutions for generalized ordinal Nash games, that is, generalized games where the player preferences are binary relations that do not necessarily admit a utility representation.
Rating systems play an important role in competitive sports and games. They provide a measure of player skill, which incentivizes competitive performances and enables balanced match-ups. In this paper, we present a novel Bayesian rating system for contests with many participants. It is widely applicable to competition formats with discrete ranked matches, such as online programming competitions, obstacle courses races, and some video games. The simplicity of our system allows us to prove theoretical bounds on robustness and runtime. In addition, we show that the system aligns incentives: that …
Hedonic games formalize coalition formation scenarios where players evaluate an outcome based on the coalition they are contained in. Due to a large number of possible coalitions, compact representations of these games are crucial. We complement known compact representation models by a distance-based approach: Players' preferences are encoded in a bipolar manner by ordinal preferences over a small set of known neighbouring players, coalitions are represented by adequate preference orders from a player's perspective, and preferences over coalitions are extended based on a directed form of Hausd…
The Elo rating system has been used world wide for individual sports and team sports, as exemplified by the European Go Federation (EGF), International Chess Federation (FIDE), International Federation of Association Football (FIFA), and many others. To evaluate the performance of artificial intelligence agents, it is natural to evaluate them on the same Elo scale as humans, such as the rating of 5185 attributed to AlphaGo Zero. There are several fundamental differences between humans and AI that suggest modifications to the system, which in turn require revisiting Elo's fundamental rationale.…
We introduce a new measure of the discrepancy in strategic games between the social welfare in a Nash equilibrium and in a social optimum, that we call selfishness level. It is the smallest fraction of the social welfare that needs to be offered to each player to achieve that a social optimum is realized in a pure Nash equilibrium. The selfishness level is unrelated to the price of stability and the price of anarchy and is invariant under positive linear transformations of the payoff functions. Also, it naturally applies to other solution concepts and other forms of games. We study the selfish…
We consider a game in which players are the vertices of a directed graph. Initially, Nature chooses one player according to some fixed distribution and gives her a buck, which represents the request to perform a chore. After completing the task, the player passes the buck to one of her out-neighbors in the graph. The procedure is repeated indefinitely and each player's cost is the asymptotic expected frequency of times that she receives the buck. We consider a deterministic and a stochastic version of the game depending on how players select the neighbor to pass the buck. In both cases we prov…
We introduce a stochastic learning process called the dampened gradient approximation process. While learning models have almost exclusively focused on finite games, in this paper we design a learning process for games with continuous action sets. It is payoff-based and thus requires from players no sophistication and no knowledge of the game. We show that despite such limited information, players will converge to Nash in large classes of games. In particular, convergence to a Nash equilibrium which is stable is guaranteed in all games with strategic complements as well as in concave games; co…
Whiting School of Engineering, Johns Hopkins University（约翰斯·霍普金斯大学怀廷工程学院）
Representing a board game and its positions by text-based notation enables the possibility of NLP applications. Language models, can help gain insight into a variety of interesting problems such as unsupervised learning rules of a game, detecting player behavior patterns, player attribution, and ultimately learning the game to beat state of the art. In this study, we applied BERT models, first to the simple Nim game to analyze its performance in the presence of noise in a setup of a few-shot learning architecture. We analyzed the model performance via three virtual players, namely Nim Guru, Ra…
