This thesis aplied tools of economic analysis on situation in poker, where players choose to finish tournament or make a deal and take certain amount of money immediately. Theoretical frame consists of economic theories for decesion under risk and poker literature. Hypotheses are tested with regresion analysis on dataset which I obtain by my own observing. Estimations support hypothesis that players are risk-averze and loss-averze. In bigger tournaments are bigger prizepools and that is the reason, why players in bigger tournament make deals more often. Moreover deal is more likely, when players are approximetly equal skilled in poker.

This project aims to investigate how reinforcement learning (RL) techniques can be applied to the card game LimitTexas Hold’em. RL is a type of machine learning that can learn to optimally solve problems that can be formulated according toa Markov Decision Process.We considered two different RL algorithms, Deep Q-Learning(DQN) for its popularity within the RL community and DeepMonte-Carlo (DMC) for its success in other card games. With the goal of investigating how different parameters affect their performance and if possible achieve human performance.To achieve this, a subset of the parameters used by these methods were varied and their impact on the overall learning performance was investigated. With both DQN and DMC we were able to isolate parameters that had a significant impact on the performance.While both methods failed to reach human performance, both showed obvious signs of learning. The DQN algorithm’s biggest flaw was that it tended to fall into simplified strategies where it would stick to using only one action. The pitfall for DMC was the fact that the algorithm has a high variance and therefore needs a lot of samples to train. However, despite this fallacy,the algorithm has seemingly developed a primitive strategy. We believe that with some modifications to the methods, better results could be achieved. / Detta projekt strävar efter att undersöka hur olika Förstärkningsinlärning (RL) tekniker kan implementeras för kortspelet Limit Texas Hold’Em. RL är en typ av maskininlärning som kan lära sig att optimalt lösa problem som kan formuleras enligt en markovbeslutsprocess. Vi betraktade två olika algoritmer, Deep Q-Learning (DQN) som valdes för sin popularitet och Deep Monte-Carlo (DMC) valdes för dess tidigare framgång i andra kortspel. Med målet att undersöka hur olika parametrar påverkar inlärningsprocessen och om möjligt uppnå mänsklig prestanda. För att uppnå detta så valdes en delmängd av de parametrar som används av dessa metoder. Dessa ändrades successivt för att sedan mäta dess påverkan på den övergripande inlärningsprestandan. Med både DQN och DMC så lyckades vi isolera parametrar som hade en signifikant påverkan på prestandan. Trots att båda metoderna misslyckades med att uppnå mänsklig prestanda så visade båda tecken på upplärning. Det största problemet med DQN var att metoden tenderade att fastna i enkla strategier där den enbart valde ett drag. För DMC så låg problemet i att metoden har en hög varians vilket innebär att metoden behöver mycket tid för att tränas upp. Dock så lyckades ändå metoden utveckla en primitiv strategi. Vi tror att metoder med ett par modifikationer skulle kunna nå ett bättre resultat. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm

No description available.

Les implications sociales et personnelles du jeu pathologique étant importantes, il devient nécessaire d'évaluer l'ampleur de la problématique dans la population. Pour ce faire, l'utilisation d'un instrument qui évalue les joueurs avec justesse est essentielle. Cette étude, réalisée auprès de 200 adultes, porte sur la capacité du NORC DSM-IV Screen (NODS) à évaluer correctement les joueurs. De façon plus précise, elle vise à déterminer le pourcentage de faux positifs généré par le NODS chez une population de joueurs de loterie vidéo. Un critère de comparaison a été établi à partir d'une entrevue clinique. Les résultats démontrent que le NODS surestime le nombre de personnes rapportant avoir un problème de jeu. Ce résultat va à l'opposé des études de validation du NODS et questionne la pertinence d'utiliser cet instrument pour estimer la prévalence du problème de jeu dans une population.

Actuellement, aucune théorie psychologique unique et aucun modèle multifactoriel n’arrive à expliquer les facteurs causaux responsables du développement du jeu pathologique. Les progrès dans ce domaine sont entravés, entre autres, par des conceptualisations de la problématique qui ne font pas consensus dans la communauté scientifique. Blaszczynski et Nower (2002) font partie des pionniers à soutenir que les joueurs de jeu de hasard et d’argent (JHA) forment un groupe hétérogène et à avoir étudié les sous-groupes de joueurs pathologiques. Leurs recherches ainsi que celles de leurs successeurs ont démontré que l’hypothèse répandue, selon laquelle les joueurs pathologiques forment une population homogène, serait fausse et que les traitements dérivés de ces modèles ne peuvent pas s’appliquer uniformément à chaque joueur sans considérer, entre autres, le type de JHA, le genre, l’histoire développementale et la neurobiologie. La présente thèse, qui se décline en deux études, s’inscrit dans cet esprit. S’inspirant du Modèle de voies associatives du jeu problématique (Blaszczynski et Nower, 2002), mais également de la théorie de l’interaction psychostructurelle (« psycho-structural interaction ») de Griffiths (1993), la présente thèse vise à développer une meilleure compréhension du rôle du type de JHA dans la conceptualisation des comportements de jeu en comparant des joueurs d’appareils de loterie vidéo (ALV) et des joueurs de poker. Plus précisément, la première étude vise à (1) comparer les joueurs de poker et les joueurs d’ALV sur le plan de la gravité du problème de jeu et de caractéristiques psychologiques liées aux comportements de jeu, c’est-à-dire le narcissisme, la recherche de sensations, les symptômes dépressifs, les symptômes anxieux et les distorsions cognitives, (2) tester empiriquement l’effet médiateur des distorsions cognitives entre les caractéristiques psychologiques et la gravité du problème de jeu, et (3) tester un modèle prédictif de la gravité du problème de jeu et vérifier l’invariance de ce modèle selon le type de jeu. Les résultats indiquent que, comparativement aux joueurs d’ALV, les joueurs de poker de l’échantillon présentent des niveaux significativement supérieurs de narcissisme et de recherche de sensations, mais significativement inférieurs de dépression et de gravité du problème de jeu. Les analyses de médiation (bootstrap) révèlent que les distorsions cognitives sont médiatrices de l’effet du narcissisme sur la gravité du problème de jeu pour les deux groupes. Pour les joueurs d’ALV uniquement, les distorsions cognitives jouent le rôle de médiateur entre le score de dépression et celui de la gravité du problème de jeu. L’analyse multigroupe met en évidence une non-invariance dans la configuration du modèle prédictif du problème de jeu entre les deux groupes. En somme, le profil des joueurs de poker diffèrent de celui des joueurs d’ALV, ce qui suggère l’hypothèse d’une étiologie et d’un tableau clinique distinctifs. La seconde étude a pour objectif d’évaluer la validité « transjeux » d’une mesure autorapportée des distorsions cognitives, le Gambling Related Cognitions Scale (GRCS; Raylu et Oei, 2004, version française : Grall-Bronnec et al., 2012), en comparant les réponses aux items des joueurs de poker aux joueurs d’ALV. Près de la moitié des items au GRCS présentent un fonctionnement différentiel d’importance moyenne à grande dépendamment de la méthode utilisée. Les cinq items présentant un fonctionnement différentiel plus important favorisent l’endossement des joueurs de poker (comparativement aux joueurs d’ALV), indépendamment de leur score total, et gravitent autour des thèmes suivants : l’habileté, les connaissances, les apprentissages et les probabilités. Cet artéfact de la mesure a pour conséquence de majorer le score total des joueurs de poker, suggérant un plus haut niveau de distorsions cognitives, et un effet inverse pour les joueurs d’ALV. La discussion relève les implications cliniques et théoriques de ces résultats. / Currently, there exists no psychological theory or multifactorial model that adequately explains the causal factors that contribute to the development of gambling disorder. Progress in this domain is hindered by the fact that there is no scientific consensus about the conceptualization of this disorder. Blaszczynski and Nower (2002) were among the first to state that gamblers are a heterogenous group and to study subgroups of pathological gamblers. Their work and that of their successors has shown that the popular hypothesis according to which pathological gamblers are a homogenous group is false, and that treatments derived from these models cannot be uniformly applied to all gamblers without consideration for certain crucial aspects, such as type of gambling game, gender, developmental history and neurobiology. Comprised of two studies, the current thesis aims to further the knowledge in this vein of work. Inspired by the Pathways model of pathological gambling (Blaszczynski, & Nower, 2002), and by the psycho-structural interaction theory (Griffiths, 1993), this thesis aims to develop a better understanding of the influence of the structural characteristics of gambling games on gambling behaviours by comparing individuals who gamble on video lottery terminals (VLT) or at poker. More specifically, the first study aims (1) to compare poker gamblers and VLT gamblers with respect to the severity of the gambling problem and psychological characteristics related to gambling behaviours, such as narcissism, sensation-seeking, depressive or anxiety symptoms, and cognitive distortions, (2) to test a predictive model of gambling problem severity and verify its invariance with respect to the type of gambling game, and (3) to test a mediating effect of cognitive distortions between psychological characteristics and gambling problem severity. The results indicate that in comparison to VLT gamblers, poker players present significantly higher levels of narcissism and sensation-seeking, and significantly lower levels of depression and gambling problem severity. Bootstrap analyses reveal that cognitive distortions have a mediating effect on gambling problem severity for both groups of gamblers. In addition, the level of depression in VLT players significantly predicts gambling problem severity both directly and indirectly via the mediating effect of cognitive distortions. Multigroup invariance analyses reveal a non-invariance in the configuration of the predictive model of the gambling problem between the two groups. These profiles could be linked to the player's interest for a specific type of gambling game, but also to problem gambling. The second study aims to assess the reliability of a French adaptation of a self-report measure of cognitive distortions, the Gambling Related Cognitions Scale (GRCS; Raylu, & Oei, 2004, French version: Grall-Bronnec et al., 2012), by comparing the answers of poker players to those of VLT players. Depending on the method used, nearly half of the GRCS items were flagged as showing differential item functioning (DIF) with a medium to large effect. The five items with most significant DIF were more positively endorsed by poker players as compared to VLT gamblers, independent of their total score. These items were related to skill, knowledge, learning and probabilities. This artefact in the instrument consequently increases the total score of poker players, thus suggesting a higher level of cognitive distortions in this group, and has the reverse effect with VLT players. Clinical and theoretical implications derived from these findings are discussed.

Efficient algorithms exist for finding optimal strategies in extensive-form games. However human scale problems, such as poker, are typically so large that computation of these strategies remain infeasible with current technology. State space abstraction techniques allow us to derive a smaller abstract game, in which an optimal strategy can be computed and then used in the real game. This thesis introduces state of the art abstraction techniques. Most of these techniques do not deal with public information. We present a new automatic public state space abstraction technique. We examine the quality of this technique in the domain of poker. Our experimental results show that the new technique brings significant performance improvement. Powered by TCPDF (www.tcpdf.org)

Within the context of

Title: Motivations of females to play poker online. (Business study of Swedish women) Authors: Anastasiya Bochkareva - 870208 Engelbrektsgatan 8A 72216 Västerås, Sweden aba07001@student.mdh.se Karina Petrova - 880613 Södergatan 22 19547 Märsta, Sweden kpa07001@student.mdh.se Supervisor: Mikael Holmgren Examiner: Ole Liljetors Key Words: Online gambling; Online poker; Swedish women; Motivation; Emotional motivations; Material motivations, Atmosphere; Accessibility; Hypotheses; Regression analysis Institution: Mälardalen University Sweden, School of Sustainable Development of Society and Technology, Box 883, 721 23 Västerås Course: Master Thesis in Business Administration, 15 ECTS-points, spring semester 2011 Problem: What motivates Swedish women to play poker online for money? Purpose: The aim of this study is to investigate the main motives behind Swedish females’ propensity to play poker online by means of hypotheses testing through regression analysis. Method: Method includes collection of the information on definitions, theories and modelsabout gambling, online gambling (precisely online poker) and motivations togamble. Five hypotheses have been constructed based on the collected informationand the survey have been created and conducted among 397 Swedish female onlinepoker players. Further, based on gathered data, hypotheses have been tested bymeans of simple linear and multiple regressions.. Results: Regression analysis revealed that emotional and material motivations together withaccessibility of the game and surrounding atmosphere play a significant role in thereason why Swedish females play online poker for money. It was also found thatemotional motivation was a fundamental factor that triggers Swedish women togamble. Furthermore, research revealed the hypothetical target segment of femaleSwedish online poker players. Average Swedish woman is between 30-40 years oldsingle woman without children, who live high speed, active life (either studying orworking) and who spends around 30.3 hours per months on playing online poker.

Title: Solving Endgames in Large Imperfect-Information Games such as Poker Author: Bc. Karel Ha Department: Department of Applied Mathematics Supervisor: doc. Mgr. Milan Hladík, Ph.D., Department of Applied Mathematics Abstract: Endgames have a distinctive role for players. At the late stage of games, many aspects are finally clearly defined, deeming exhaustive analysis tractable. Specialised endgame handling is rewarding for games with perfect information (e.g., Chess databases pre-computed for entire classes of endings, or dividing Go board into separate independent subgames). An appealing idea would be to extend this approach to imperfect-information games such as the famous Poker: play the early parts of the game, and once the subgame becomes feasible, calculate an ending solution. However, the problem is much more complex for imperfect information. Subgames need to be generalized to account for information sets. Unfortunately, such a generalization cannot be solved straightaway, as it does not generally preserve optimality. As a consequence, we may end up with a far more exploitable strategy. There are currently three techniques to deal with this challenge: (a) disregard the problem entirely; (b) use a decomposition technique, which sadly retains only the same quality; (c) or formalize improvements of...

[pt] A ciência de dados precisa de uma grande quantidade de dados para testar e melhorar soluções. Jogos são largamente usados para abstrair situações da vida real. Rodadas de pôquer são um bom exemplo pois, por não saber as cartas dos oponentes, o jogador analisa um cenário de informação incompleta numa competição de agentes que envolve conhecimento probabilístico, análise de risco e brefe. Isso o diferencia de xadrez, damas e jogos de conhecimento perfeito e algoritmos de busca em forca bruta sobre o espaço de soluções. Usar o pôquer como um caso de teste possibilita a análise de diferentes abordagens usadas na vida real, porém num cenário mais controlado. Esta dissertação propõe um arcabouço de funcionalidades para criar e testar diferentes algorítimos de Deep Learning, que podem jogar pôquer entre sí, aprender com o histórico e maximizar suas recompensas. / [en] Data science research needs real examples to test and improve solutions. Games are widely used to mimic those real-world examples. Poker rounds are a good example of imperfect information state with competing agents dealing with probabilistic knowledge, risk assessment, and possible deception, unlike chess, checkers and perfect information brute-force search style of games. By using poker as a test-bed we can analyze different approaches used in real-world examples, in a more controlled environment, which should give great insights on how to tackle those real-world scenarios. We propose a framework to build and test different neural networks that can play against each other, learn from a supervised experience and maximize its rewards.