| Summary: | The video and computer gaming industry has seen a significant rise in popularity over the last decades and is now the worlds biggest digital entertainment industry [market-gamingvsmovies]. Where games meant Space Invaders and Doom in the old days, gaming is now everything from ultra-realistic shooter games to farm management simulators integrated in the Facebook platform, to games on our smartphones and tablets. This popularity has brought with it the attention of hackers and exploiters, and game cheats flourish in the shady parts of the internet. This thesis has two parts. The first part presents a background study, an oveorview of today's situation in regard to cheating and anti-cheating in online games. The second part is a case study about bots and bot detection in Browser-Based Multiplayer Games (BBMG).In the background study, the main finding is that there are over twenty websites selling cheats or cheat subscriptions, and that the type of cheats available for a game is heavily dependent on the game genre. Most or all first-person shooters (FPS) have available aimbots and wallhacks, and all major massively multiplayer online games (MMO) are exposed to bot programs. The cheats are being sold either alone or through monthly subscriptions, allowing free use of all cheats for all supported games by that vendor. There have been many anti-cheat actors over the past decade, but three known services being continually developed. The two biggest are Valve Anti-Cheat, used by most games managed through Valve's online gaming platform Steam, and PunkBuster, which protects amongst others the Battlefield series. The last big anti-cheat software is the proprietary Warden, used only in Blizzard's own games like the Diablo. Warcraft and Starcraft games. Many cheats are still able to bypass the security mechanisms provided by these, and it is a continuos arms race between cheat developers and anti-cheat developers, just like in the virus and anti-virus industry.In the case study, two bots were written for a non-disclosed BBMG and their performance was tested quantitatively by playing several game accounts in parallel. It showed that bot use yields large performance gains both in the early game stages and for advanced players. As a result of the case study research on bots in online games, a Bot-Detection System (BDS) is created and presented in the last chapter. The goal of the BDS is to detect the use of bots in BBMG by identifying a set of attributes and comparing these to average human behavior. A score system ranging from 0 to 100 is introduced, where the average human behavior is defined as 0, while increasing non-human behavior is given a score >0, with max 100. The BDS is then employed on the two bots and returns scores of 64 and 52, while the six human play testers receive scores of 1 or 0.
|
|---|
