Context: Real-time games like Dota 2 lack the extensive mathematical modeling of turn-based games that can be used to make objective statements about how to best play them. Understanding a real-time computer game through the same kind of modeling as a turn-based game is practically impossible. Objectives: In this thesis an attempt was made to create a model using machine learning that can predict the winning team of a Dota 2 game given partial data collected as the game progressed. A couple of different classifiers were tested, out of these Random Forest was chosen to be studied more in depth. Methods: A method was devised for retrieving Dota 2 replays and parsing them into a format that can be used to train classifier models. An experiment was conducted comparing the accuracy of several machine learning algorithms with the Random Forest algorithm on predicting the outcome of Dota 2 games. A further experiment comparing the average accuracy of 25 Random Forest models using different settings for the number of trees and attributes was conducted. Results: Random Forest had the highest accuracy of the different algorithms with the best parameter setting having an average of 88.83% accuracy, with a 82.23% accuracy at the five minute point. Conclusions: Given the results, it was concluded that partial game-state data can be used to accurately predict the results of an ongoing game of Dota 2 in real-time with the application of machine learning techniques.

As live streaming of video games has become easier, electronic sports have grown quickly and they are still increasing as tournaments grow in viewers and prizes. The purpose of this paper is to examine the theory Metagame Bounds by applying it to League of Legends and Dota 2, to see if it is a valid way of looking at character balance in the Multiplayer Online Battle Arena game genre. The main mode of both games consist of matches played on a map where a team of five players is up against another team of five players. Characters in the games generally have four abilities and a number of attributes, making them complex to compare without context. We gathered character data from websites and entered the data into a file that we used with the Metagame Bounds application. We compared the graphs that the data yielded with how often a character wins and how often it is played. To examine whether the characters were balanced we also played the games and analysed the characters in depth. All statistical data gathered was retrieved over the span of a few hours on the 21st of April 2015 from the websites Champion.gg for League of Legends and Dotabuff for Dota 2. Sirlin’s (2001) definition of multiplayer game balance is “A multiplayer game is balanced if a reasonably large number of options available to the player are viable – especially, but not limited to, during high-level play by expert players.” and with the data we see that these games are balanced in terms of characters according to that definition. Overall we believe that Metagame Bounds is a valid way of looking at balance within the genre, but the data is not completely reliable on its own and should be compared with other statistics.

Multiplayer online battle arena (MOBA) games and games in general offer developers opportunities to design and test AI to further push research within the area. MOBA is a genre that has seen an increase in popularity during recent years alongside a rapidly growing esports scene.  The game's main purpose is ultimately to be enjoyed and played by human players, who often come up with and are encouraged to create their own strategies in order to beat the game or win out over their opponents.  This thesis will explore how mimicking human behavior in AI and using statistics from human players can be beneficial to AI design. The AI will be developed in a framework similar to that used by the Conference of Games for their Dota 2 5v5 AI competition. Three different strategies will be created, each of which will mimic some aspect of human behaviour, and then tested against the built-in AI in Dota 2. All the developed AI’s managed to win with varying results. The results showed some support for the notion that human gameplay and statistics can be beneficial to AI design. There are, however, difficulties with the complexity of the game that is Dota 2.

This case study investigates if the AI text-to-image generator Midjourney can generate visual game character idea images for Dota 2 that meet the game’s art guidelines. The author defined the term “visual game character ideas” as the idea images at the early stage of visual character design process to help artists get inspiration. To achieve this, an experiment was designed and conducted where three new Dota 2 heroes’ backgrounds were developed by the author, and 32 images per hero were generated with Midjourney bot. These 96 images were evaluated to examine Midjourney’s performance based on seven aspects: accurate content, readability and identifiability, value gradient, value patterning, the number of colors, areas of rest and detail, and directionality. “YES” was given to each criterion if they meet the requirement. The value of this case study is to present the strength and weakness of the text-to-image generators for visual character design ideas, which can potentially show game artists when and how to use them in visual character design process. The result suggested that Midjourney could be used to generate visual character design ideas for Dota 2 unstably, and this instability was mainly caused by its identified flaw: content accuracy. Furthermore, it performed better for non-color-related aspects, while the performance for color-related items was significantly worse than others.

When a company includes a new technology or innovation into their value proposition, customers may perceive it as an enhancement or deterioration. This phenomenon was explored in this study with a case study of a present case in the esports industry. Research have shown that AI (Artificial Intelligence) technology can be used to predict which team is going to win in a match in the esports game DotA 2. A prototype AI called Znipe Sense was developed and analyzed during this study to answer the question: How can a predictive AI affect the value proposition of live esports consumption? Znipe Sense was included into Znipe Esports’ value proposition during a tournament in February 2018. It was observed that Znipe Sense could predict outcomes of professional matches with a higher accuracy than human experts. The observations of Znipe Sense, an interview with experienced players, interviews with business professionals and internal company documents were used as empirical material for the analysis. How Znipe Sense affected the value proposition was analyzed through the factors: Performance, Ease-of-use, Reliability, Flexibility and Affectivity, also known as the PERFA framework. It was concluded that a predictive AI can enhance the value proposition of live esports consumption through the Performance and Ease-of-use factors, and it would not affect the value proposition through Reliability or Flexibility. However, in the analysis of the Affectivity factor it was identified that there is a risk related to negative effects of gambling addiction that could deteriorate the value proposition. / När ett företag inkluderar en ny teknik eller innovation i sitt värdeerbjudande kan kunderna uppleva det som en förbättring eller försämring. Detta fenomen undersöktes i denna studie med en fallstudie av ett aktuellt fall i esportsindustrin. Forskning har visat att AI (Artificiell Intelligens) teknik kan användas för att förutsäga vilket lag som kommer att vinna i en match i esports spelet DotA 2. En AI prototyp, Znipe Sense, utvecklades och analyserades under denna studie för att svara på frågan: Hur kan ett prediktivt AI påverka värdeerbjudandet av live esports konsumtion? Znipe Sense inkluderades i Znipe Esports värdeerbjudande under en turnering i februari 2018. Det observerades att Znipe Sense var bättre än mänskliga experter på att förutse det vinnande laget i professionella matcher. Observationerna av Znipe Sense, en intervju med erfarna spelare, intervjuer med affärsfolk och interna företagsdokument användes som empiriskt material. Hur Znipe Sense påverkade värdeerbjudandet analyserades genom faktorerna: ”Performance”, ”Ease-of-use”, ”Reliability”, ”Flexibility” och ”Affectivity”, även känt som PERFA-ramverket. Slutsatsen visar på att ett prediktivt AI kan öka värdeerbjudandet av live esports konsumtion genom Performance och Ease- of-use faktorerna, och att värdeerbjudandet inte skulle påverkas genom Reliability och Flexibility. I analysen av Affectivity-faktorn identifierades emellertid att det finns en risk i relaterat till negativa effekter av spelberoende som kan försämra värdeerbjudandet.

Context: Games involve people to a large extent where they relate themselves with the game characters; this is commonly known as game immersion. Generally, some players play games for enjoyment, some for stress relaxation and so on.Game immersion is usually used to describe the degree of involvement with a game. When people play games, they don’t necessarily realize that they have been dissociated with the surrounding world. Real world dissociation (RWD) can be defined as the situation where a player is less aware of the surroundings outside the game than about what is happening in the game itself. The RWD factor has been expected to measure the losing track of time, lack of awareness of surroundings and mental transportation. Objectives: In this thesis, we measure and compare the difference in game immersion between experienced and inexperienced players using RWD factor. In addition, the study involves exploring the significance of game immersion and various approaches used to measure it. Methods: In this study literature review has been carried out to explore the meaning of game immersion and further user studies in the form of an experiment has been conducted to measure game immersion between experienced and inexperienced gamers. The game immersion has been measured using the real world dissociation (RWD) factor. After the experiment has been conducted, a statistical technique has been carried out to measure the difference in game immersion among the two groups. Results:The empirical investigation on the measurement of game immersion has been done using RWD factor. The results state that the significance value is less than 0.05 and hence null hypothesis is rejected for both the games. The measurable difference has been calculated by using Cohen’s d effect size between experienced and inexperienced players. The Cohen’s d value between experienced players and inexperienced players for Dota 2 is 0.7423 and CS:GO is 0.8383. Conclusions: After analyzing the data and calculating the effect size, the overall results state that inexperienced group of players are more immersed than the experienced group of players when measured by RWD factor. Hence it can be concluded that irrespective of the game played, inexperienced players are more dissociated from the real world than the experienced players.

El objetivo de la investigación fue analizar la relación la Motivación para jugar videojuegos online y la Calidad de Vida con Respecto a la Salud en adolescentes de 18 a 21 años residentes en Lima Metropolitana. La muestra estuvo conformada por 180 jugadores de Dota 2, seleccionados por conveniencia en cabinas de internet, eventos de videojuegos y en redes sociales. Se utilizó la Escala de Motivación hacia los videojuegos de multijugadores masivos (MMO-MS) y el cuestionario de CVRS en adolescentes (Kidscreen 27). La relación entre las dimensiones que componen cada una de las variables, fue analizada a través del test de Spearman. Se halló relaciones significativas positivas entre las dimensiones de Disociación con Autonomía económica y negativamente con las dimensiones Bienestar físico, Bienestar psicológico. Asimismo, las dimensiones Relación con los padres; Apoyo social y Entorno escolar; Socialización correlacionó positivamente con autonomía económica. También se encontró que la cantidad diaria de tiempo dedicado al juego y el bienestar psicológico se relacionan de manera negativa. / The objective of the research was to analyze the relationship between the Motivation variables to play online video games and quality of life regarding Health in adolescents between 18 and 21 years old in Metropolitan Lima. The sample was made up of 180 adolescents students from Lima who play Dota 2. The sample consisted of 180 Dota 2 players selected for convenience in internets café, videogame events and on social media. The Motivation Scale for massively multiplayer video games (MMO-MS) and the instrument of Quality of Life regarding to Health in adolescents (Kidscreen 27). The relationship between the dimensions of the variables was analyzed thought the spearman test. Significant positive relationships were found between the dimensions of Dissociation with Economic autonomy and negatively with the dimensions Physical well-being, Psychological well-being, Relationship with parents and; Social support and School environment; socialization was positively correlated with Economic autonomy. Also, it was found that the daily amount of time devoted to play and Psychological well-being are negatively related. / Tesis

La industria gamer ha llevado a la posibilidad de subsistir económicamente del streaming. Dado que los streamers generan ingresos en relación con su audiencia, mientras mayor sea esta y esté más conectada con ella mayor probabilidad habrá de que genere mayores ingresos. Por ende, surge un nuevo modelo de negocio que requiere de una gestión reputacional para construir esta audiencia. Para ello, se necesita un primer análisis del contexto peruano desde la percepción que mantienen los jugadores de Dota 2 peruanos frente a los streamers peruanos de Dota 2. Por ello, este trabajo utiliza el paradigma naturalista/interpretativo que estudia al público en su contexto natural, es decir los gamers y el Dota 2. El diseño para este sería fenomenológico pues se busca entender la percepción (reputación) del streamer peruano desde el punto de vista de los jugadores de Dota 2 en general. Finalmente, se usa un enfoque cualitativo para obtener las opiniones de los usuarios que juegan Dota 2 sobre los streamers peruanos, con el uso de entrevistas semiestructuradas para alinearlas en los 3 aspectos para medir reputación: sólida dimensión axiológica, comportamiento comprometido y calidad de servicio. / The gaming industry has led to the possibility of economically subsisting on streaming. Since streamers generate income in relation to their audience, the larger their audience and the more connected they are, the greater probability of generating more income will exist. Therefore, a new business model emerges that requires reputational management to build this audience. To do this, a first analysis of the Peruvian context is needed from the perception that Peruvian Dota 2 players maintain compared to Peruvian Dota 2 streamers. That’s why, this work uses the naturalistic paradigm /interpretive that studies the public in its natural context, that is, the gamers in their gameplay and the design for this would be phenomenological because it seeks to understand the perception (reputation) of the Peruvian streamer from the point of view of Dota 2 players in general. Finally, a qualitative approach is used to obtain the opinions of the users who play Dota 2 on the Peruvian streamers, with the use of semi-structured interviews to align them in the 3 aspects to measure reputation: solid axiological dimension, committed behavior and quality of service. / Trabajo de investigación