Fuzzy controller based AI for dynamic difficulty adjustment for defense of the Ancient 2 (DotA2)
28 Jul 2016-pp 95-100
TL;DR: This research offered enrichment in gaming experience for players with implementation of Dynamic Difficulty Adjustment (DDA) to avoid this misconception by applying right difficulty based on player in-game records.
Abstract: Defense of the Ancient 2 (DotA2) is a game with a huge player base achieving almost 11 millions players during the time when this paper was written. Even though this game is more focused on its multiplayer players versus players mode, it's bot match mode is still a viable feature provided by Valve for beginner players to start learning DotA2 and even for seasoned players it is still used to testing builds or sharpening their skills. However static AI implemented in this game are often mismatched between the player and the AI difficulty because player themselves don't know how far their limit is. Even for seasoned player facing AI with perfect reaction for the whole game is often still frustrating. This research offer enrichment in gaming experience for players with implementation of Dynamic Difficulty Adjustment (DDA) to avoid this misconception by applying right difficulty based on player in-game records. The result showed that the AI for Dynamic Difficulty Adjustment can be applied in the game.
Citations
More filters
24 May 2017
TL;DR: The results show that Naive Bayes classifier is a practical tool to analyze the lineups and predict the outcome based on players' choices.
Abstract: Although DOTA2 is a popular game around the world, no clear algorithm or software are designed to forecast the winning probability by analyzing the lineups. However, the author finds that Naive Bayes classifier, one of the most common classification algorithm, can analyze the lineups and predict the outcome according to the lineups and gives an improved Naive Bayes classifier. Using the DOTA2 data set published in the UCI Machine Learning Repository, we test Naive Bayes classifier's prediction of respective winning probability of both sides in the game. The results show that Naive Bayes classifier is a practical tool to analyze the lineups and predict the outcome based on players' choices.
17 citations
13 Jul 2018
TL;DR: This paper introduces MOBA-Slice, a time slice based evaluation framework of relative advantage between teams in MOBA games, a quantitative evaluation method based on learning, similar to the value network of AlphaGo.
Abstract: Multiplayer Online Battle Arena (MOBA) is currently one of the most popular genres of digital games around the world. The domain of knowledge contained in these complicated games is large. It is hard for humans and algorithms to evaluate the real-time game situation or predict the game result. In this paper, we introduce MOBA-Slice, a time slice based evaluation framework of relative advantage between teams in MOBA games. MOBA-Slice is a quantitative evaluation method based on learning, similar to the value network of AlphaGo. It establishes a foundation for further MOBA related research including AI development. In MOBA-Slice, with an analysis of the deciding factors of MOBA game results, we design a neural network model to fit our discounted evaluation function. Then we apply MOBA-Slice to Defense of the Ancients 2 (DotA2), a typical and popular MOBA game. Experiments on a large number of match replays show that our model works well on arbitrary matches. MOBA-Slice not only has an accuracy 3.7% higher than DotA Plus Assistant (A subscription service provided by DotA2) at result prediction, but also supports the prediction of the remaining time of a game, and then realizes the evaluation of relative advantage between teams.
8 citations
TL;DR: This work explores the use of intelligent robotics and dynamic difficulty adjustment mechanisms to develop a novel working memory training system based on the Nao robotic platform, which shows a significant improvement on the performance of the user in the game, which might relate to an improvement in their working memory.
Abstract: Working memory is an important function for human cognition since several day-to-day activities are related to it, such as remembering a direction or developing a mental calculation. Unfortunately, working memory deficiencies affect performance in work or education related activities, mainly due to lack of concentration, and, with the goal to improve this, many software applications have been developed. However, sometimes the user ends up bored with these games and drops out easily. To cope with this, our work explores the use of intelligent robotics and dynamic difficulty adjustment mechanisms to develop a novel working memory training system. The proposed system, based on the Nao robotic platform, is composed of three main components: First, the N-back task allows stimulating the working memory by remembering visual sequences. Second, a BDI model implements an intelligent agent for decision-making during the progress of the game. Third, a fuzzy controller, as a dynamic difficulty adjustment system, generates customized levels according to the user. The experimental results of our system, when compared to a computer-based implementation of the N-back game, show a significant improvement on the performance of the user in the game, which might relate to an improvement in their working memory. Additionally, by providing a friendly and interactive interface, the participants have reported a more immersive and better game experience when using the robotic-based system.
5 citations
Cites methods from "Fuzzy controller based AI for dynam..."
...In [24], a fuzzy logic controller was implemented for DotA2, which adjusts the different game parameters depending on the user’s abilities....
[...]
Posted Content•
TL;DR: In this article, a time slice based evaluation framework of relative advantage between teams in MOBA games is introduced, which is a quantitative evaluation method based on learning, similar to the value network of AlphaGo.
Abstract: Multiplayer Online Battle Arena (MOBA) is currently one of the most popular genres of digital games around the world. The domain of knowledge contained in these complicated games is large. It is hard for humans and algorithms to evaluate the real-time game situation or predict the game result. In this paper, we introduce MOBA-Slice, a time slice based evaluation framework of relative advantage between teams in MOBA games. MOBA-Slice is a quantitative evaluation method based on learning, similar to the value network of AlphaGo. It establishes a foundation for further MOBA related research including AI development. In MOBA-Slice, with an analysis of the deciding factors of MOBA game results, we design a neural network model to fit our discounted evaluation function. Then we apply MOBA-Slice to Defense of the Ancients 2 (DotA2), a typical and popular MOBA game. Experiments on a large number of match replays show that our model works well on arbitrary matches. MOBA-Slice not only has an accuracy 3.7% higher than DotA Plus Assistant at result prediction, but also supports the prediction of the remaining time of the game, and then realizes the evaluation of relative advantage between teams.
5 citations
13 Mar 2019
TL;DR: The result showed that the AI for Dynamic difficulty Adjustment can be applied in the game and offers enrichment in gaming experiences for players by implementing Dynamic Difficulty Adjustment (DDA).
Abstract: Starcraft 2 is a game with quite a big player base achieving around 1 million players during the time when this paper was written. Although this game's focus is at multiplayer player versus player, its bot match mode is still a viable feature provided by Blizzard for beginner players to start learning Starcraft 2. And even for seasoned players, it is still used to improve their skill. However static AI implemented in this game are often mismatched between the player's skill and the AI's difficulty because the player themselves has yet to understand the limitations of their skill. Even for the seasoned player, facing AI with better resource management and resource income for the whole game is often still frustrating. This research offers enrichment in gaming experiences for players by implementing Dynamic Difficulty Adjustment (DDA) to avoid this misconception by applying the right difficulty based on player's in-game records. The result showed that the AI for Dynamic Difficulty Adjustment can be applied in the game.
1 citations
Cites methods from "Fuzzy controller based AI for dynam..."
...DDA approach has been used in several papers like [4] which the researcher apply DDA on FPS game or at [2] which DDA was applied on MOBA game, and both give promising results....
[...]
...The AI uses the data to analyze the player's kill/death/assist ratio, experience etc to adjust the game [2]....
[...]
References
More filters
01 Oct 2014
TL;DR: Three analyses indicate that spatio-temporal behaviour of MOBA teams is highly related to team skill, and a method for obtaining accurate positional data from DotA 2 is presented.
Abstract: Multiplayer Online Battle Arena (MOBA) games are among the most played digital games in the world. In these games, teams of players fight against each other in arena environments, and the gameplay is focussed on tactical combat. In this paper, we present three data-driven measures of spatio-temporal behaviour in Defence of the Ancients 2 (DotA 2): 1) Zone changes; 2) Distribution of team members and: 3) Time series clustering via a fuzzy approach. We present a method for obtaining accurate positional data from DotA 2. We investigate how behaviour varies across these measures as a function of the skill level of teams, using four tiers from novice to professional players. Results from three analyses indicate that spatio-temporal behaviour of MOBA teams is highly related to team skill.
98 citations
18 May 2006
TL;DR: It is demonstrated that appropriate non-extreme levels of challenge and curiosity generate high values of entertainment and the extensibility of the approach to other genres of digital entertainment and edutainment is discussed.
Abstract: This paper introduces quantitative measurements/metrics of qualitative entertainment features within computer game environments and proposes artificial intelligence (AI) techniques for optimizing entertainment in such interactive systems. A human-verified metric of interest (i.e. player entertainment in real-time) for predator/prey games and a neuro-evolution on-line learning (i.e. during play) approach have already been reported in the literature to serve this purpose. In this paper, an alternative quantitative approach to entertainment modeling based on psychological studies in the field of computer games is introduced and a comparative study of the two approaches is presented. Artificial neural networks (ANNs) and fuzzy ANNs are used to model player satisfaction (interest) in real-time and investigate quantitatively how the qualitative factors of challenge and curiosity contribute to human entertainment. We demonstrate that appropriate non-extreme levels of challenge and curiosity generate high values of entertainment and we discuss the extensibility of the approach to other genres of digital entertainment and edutainment.
65 citations
15 Jul 2013
TL;DR: This paper uses the team-oriented multiplayer online game Dota 2 to study cooperation within teams and the success of teams, and chooses a statistical approach to identify factors that increase the chance of a team to win.
Abstract: Teamwork plays an important role in many areas of today's society, such as business activities. Thus, the question of how to form an effective team is of increasing interest. In this paper we use the team-oriented multiplayer online game Dota 2 to study cooperation within teams and the success of teams. Making use of game log data, we choose a statistical approach to identify factors that increase the chance of a team to win. The factors that we analyze are related to the roles that players can take within the game, the experiences of the players and friendship ties within a team. Our results show that such data can be used to infer social behavior patterns.
53 citations
01 Nov 2015
TL;DR: The presented mechanism to perform the dynamic difficulty adjustment during a game match was able to keep up with the player's abilities on 85% of all experiments, and the remaining 15% failed to suit the players' need because the adjustment did not occur on the right moment.
Abstract: Dynamic Difficulty Adjustment (DDA) consists in an alternative to the static game balancing performed in game design. DDA is done during execution, tracking the player's performance and adjusting the game to present proper challenges to the player. This approach seems appropriate to increase the player entertainment, since it provides balanced challenges, avoiding boredom or frustration during the gameplay. This paper presents a mechanism to perform the dynamic difficulty adjustment during a game match. The idea is to dynamically change the game AI, adapting it to the player skills. We implemented three different AIs to match player behaviors: beginner, regular and experienced in the game Defense of the Ancient (DotA), a modification (MOD) of the game Warcraft III. We performed a series of experiments and, after comparing all results, the presented mechanism was able to keep up with the player's abilities on 85% of all experiments. The remaining 15% failed to suit the player's need because the adjustment did not occur on the right moment.
27 citations
23 Jul 2013
TL;DR: The experimental results show that the adaptive mechanism developed in this study could dynamic balance the equilibrium of game difficulty and enhance the replayability of the game.
Abstract: This paper describes a series of experiments using the offline trained artificial neural networks (ANN). The ANN acts as an embedded game agent in a shooting game to control the nonplayer character (NPC). The training datasets of ANN are constructed by three different levels of players (expert, medium and beginner players). And then the three different levels training datasets are used to train three different level's ANN, respectively. Meanwhile, the optimal neurons of the hidden layer and the suitable period of training time is obtained by the method of three fold cross validation. In addition, a comparison between ANN and two traditional game AI — finite state machine (FSM) and computer random controlled method, is also implemented in this study. The simulated results show that ANN can get better winning rate than FSM and random method. Meanwhile, ANN obtains a pretty good human-like simulation results. Finally, a fuzzy rules-based approach is utilized to do the dynamic game difficulty adjustment. The experimental results show that the adaptive mechanism developed in this study could dynamic balance the equilibrium of game difficulty. All these, enhance the replayability of the game.
3 citations