The aim of this thesis was to increase understanding of King company success in the social mobile game industry by asking the question: How does a company manage to organize the innovation work in successful casual games within social mobile gaming industry? In order to answer it, we conducted a case study research with secondary data in which we examined the company to discover the elements that contribute to this success, despite a lack of research in how these kind of companies build their business model and strategies, highlighting the players' behaviour. Our findings conclude it is possible to success in social mobile game industry using incremental innovation in different aspects: games design, implementation of the games, and in the business model.  By applying this innovation, with a good viral strategy and giving the player the decision to play by free or purchasing virtual goods, King has been able to become the largest developer game company on Facebook.

Dark game design patterns are features used by game creators to manipulate the player to make certain choices. These patterns can lead to unintentional player actions causing negative experiences. In this descriptive user experience study, focused on the mobile game Candy Crush Saga, the dark patterns’ effect on desirability (whether something is fun and engaging) and the players’ decision to quit or continue to be a player due to the patterns were investigated. An online survey, where the participants were asked about their feelings towards five different dark patterns identified within Candy Crush Saga, was conducted and distributed by using Google Forms and Facebook. The survey received 60 responses from current and previous players of the game. The sample was not controlled and rather homogeneous, therefore it was not necessarily representative of the entire Candy Crush Saga audience. The analysis of the gathered data indicated correlation between the use of certain types of dark patterns and decreased desirability. Some variation could be detected between the effects of different dark patterns. Similarly, certain patterns had a stronger impact on the players’ decision to quit being a player and other patterns on the choice to continue. Four out of the five patterns studied were indicated as both a reason to quit for some and a reason to continue for others.  The results offer insight to game developers and businesses about the previously insufficiently studied connection between the dark game design patterns and their direct effects to the users’ perceived experience. It is apparent that a decrease in the use of certain dark patterns would contribute to  a more positive playing experience.

Many mobile games use a user interface designed to get their players to spend more time, money or social engagement within their application. This is done by deliberately misleading or otherwise confusing the user by, for example, making the player lose track of time when playing or by giving rewards to players who spend money or invite their friends into the game. This can be taken to the extent that it strongly affects players negatively both financially and socially. This type of game elements is also known as dark patterns, which are not only used in games but also on websites, social media and other digital applications. Dark patterns have been discussed more in recent years since consumers and users have become more aware of how deceptive digital design affects behavior. This study examines the game design of Candy Crush Saga with a focus on dark patterns, and how these affect the game as a whole as well as the users who play it. The study focuses on three categories that characterize dark patterns, namely temporal, monetary and social. Based on these categories, they have subcategories that describe specific design elements with examples. The study that has been done consists of two parts: a heuristic evaluation and a survey where a number of active players have been asked to answer questions about their gaming habits in the game Candy Crush Saga. The study finds that the game contains several dark patterns, which in varying degrees affects the players’ experience in the game. The study also discusses how cumulative dark patterns enhance each other and can therefore be perceived as particularly problematic. / Många mobilspel använder sig av ett användargränssnitt utformat för att få sina spelare att spendera mer tid, pengar eller socialt engagemang inom deras applikation. Detta genom att medvetet vilseleda eller på annat sätt förvirra användaren genom att till exempel få spelaren att förlora uppfattningen av hur mycket tid de har spenderat i spelet eller genom att ge belöningar till spelare som spenderar pengar eller bjuder in sina vänner i spelet. Detta kan tas till den grad att det starkt påverkar spelarna negativt både ekonomiskt och socialt. Denna typ av spelelement kallas även dark patterns, som förutom i spel även används på webbsidor, sociala medier och andra digitala applikationer. Dark patterns har diskuterats mer under senare år då konsumenter och användare har blivit mer medvetna om hur vilseledande digital design påverkar beteenden. I denna studie har spelet Candy Crush Sagas speldesign undersökts med fokus på dark patterns, och hur dessa påverkar spelet i sin helhet och användarna som spelar. Studien har fokuserat på de tre kategorier som kännetecknar dark patterns, nämligen de temporala, monetära och sociala. Utifrån dessa kategorier har de underkategorier som beskriver specifika designelement med exempel. Studien som har gjorts består av två delar där det dels är en heuristisk utvärdering som har gjorts, samt en enkätstudie där en andel aktiva spelare har fått svara på frågor angående deras spelvanor i spelet Candy Crush Saga. Studien konstaterar att spelet innehåller flera dark patterns, som i varierande grad påverkar spelarnas upplevelse i spelet. Studien diskuterar även hur kumulativa dark patterns förhöjer varandra och därav kan upplevas som särskilt problematiska.

Given the recent advances within a subfield of machine learning called reinforcement learning, several papers have shown that it is possible to create self-learning digital agents, agents that take actions and pursue strategies in complex environments without any prior knowledge. This thesis investigates the performance of the state-of-the-art reinforcement learning algorithm proximal policy optimization, when trained on a task with nondeterministic state transitions. The agent’s policy was constructed using a convolutional neural network and the game Candy Crush Friends Saga, a single-player match-three tile game, was used as the environment. The purpose of this research was to evaluate if the described agent could achieve a higher win rate than average human performance when playing the game of Candy Crush Friends Saga. The research also analyzed the algorithm's generalization capabilities on this task. The results showed that all trained models perform better than a random policy baseline, thus showing it is possible to use the proximal policy optimization algorithm to learn tasks in an environment with nondeterministic state transitions. It also showed that, given the hyperparameters chosen, it was not able to perform better than average human performance.