Multiplayer Online Battle Arena (MOBA) and Real Time Strategies (RTS) games are two of the most popular esports genres both casually and professionally. These games, such as Heroes of the Storm (HotS) and Clash Royale (CR), have numerous characters to choose from and often provide difficulty levels as guidance. While character difficulty can serve as a tool to pick a character to play and learn, it is not necessarily a good measure of skill floor or ceiling. Additionally, fan sites offer win rates to assist players in identifying viable characters. For example, the site HOTS Logs, reports heroes’ win rates along with their popularity. However, factors such as player skill level can skew the overall win rates for characters. We collect HotS data from the HOTS Logs website, visualize hero performance across player MMR, and identify three clusters of heroes with similar relationships between win rates and MMR. We do a similar analysis for CR data obtained from the Clash Royale API.

• Heroes of the Storm
• Clash Royale

League of Legends (LoL) is a multiplayer online battle arena game where teams of five players compete against each other. Over the years, players have formed a metagaming strategy, consisting of five distinct roles, which has been widely adopted. This study uses logistic regression models to identify symbiotic relationships (such as Mutualism, Commensalism, and Parasitic) between two of the roles: Attack Damage Carries (ADC) and Supports. We use a traditional regression adjusted plus minus player structure for this study. Results from the 2015 LoL North American Ranked Solo season indicate only 5% of observed champion pairs are beneficial to both the ADC and Support and these pairs made up about 13% of matches played. Further, about 28% of pairs had at least one of the ADC of Support be negatively impacted and they made up over 40% of observed matches. Win-rates for Mutualism and Commensalism pairs were substantially higher than the other categories.

Heroes of the Storm (HotS) is a popular multiplayer online battle arena (MOBA) game released by Blizzard Entertainment in 2015. HotS differentiate itself from other MOBA games such as League of Legends and Defense of the Ancients 2 by introducing multiple competitive maps instead of one. As HotS is relatively young, the metagaming team compositions are constantly evolving. In this paper, we collected replay files of over 350,000 matches from a crowd sourced repository called HotsApi using the boto3 python module to develop a data-driven approach to identifying metagaming strategies. We use cluster analysis to identify four metagaming roles (Damage Dealers, Healers, Initiators, and Pushers) based on individual player endgame statistics. Finally, we determine the most popular team compositions for each map and assign primary and secondary roles for the 70+ heroes in the game.

League of Legends is a multiplayer online battle arena game where teams of five players compete against each other. Over the years, players (the crowd) have formed a metagaming strategy, which is widely adopted. This paper questions and answers whether the wisdom of the crowd defined the best strategy. We investigate play- ers’ choices of champions (and builds) and their team performance from matches in the North America and Western Europe regions, using the data gathered through the Riot Games official application program interface. We classify team compositions by players’ spells and attributes of items, and identify several non-meta strategies that show a consistent advantage over the meta.

League of Legends is a multiplayer online battle arena game where features are unlocked as players level up their accounts. Because it takes a significant amount of time to reach the max level, there exist accounts that are leveled automatically by illicit “bots” and then sold on the market at the max level. These bots participate in matches like human players but are incapable of either playing intelligently or cooperatively with teammates. This paper presents an investigation into the prevalence of bots in player-versus-player match types and their impact on match outcomes on the North America and Europe West servers, using the data gathered through the Riot Games official application program interface. We demonstrate that bots are present in all major match modes at various levels and that they negatively influence the balance of matches on both servers.

League of Legends is a multiplayer online battle arena game that follows a freemium model, and the available in-game transactions do little to impact a player’s performance or ability. Although champions can be purchased with actual (or in-game) money, another aspect of the game is a weekly rotation of ten free champions where players can test new champions before buying them. This project involves scraping champion usage data from online sources where we then analyze what lasting impact the free rotation feature and new and updated content (such as new and updated champions, new skins and official game updates) have on champion usage. Additionally, we have constructed a simple web application (LoLNOVA) that allows users to compare charts of usage statistics, perform simple data analyses, and download data for champions of their choice. Educators can use these data as they are relevant and interesting to many students and help increase students’ interest in quantitative fields.