User studies are a cornerstone of human-computer interaction research, including measures of user performance and quality of experience (QoE) – particularly important for games where frame rates and frame timings can impact performance. Unfortunately, commercial games have limited options for customization and do not log player performance data with sufficient detail for use in such studies. This paper introduces Lead Rush, a first-person shooter game designed for conducting user studies on the effects of frame timing and frame rate. Lead Rush is tuned to run at extremely high frame rates and includes hooks to induce frame time "spikes". Researchers can configure Lead Rush’s gameplay, trigger frame time spikes during specific actions, and log player data per game round and per study session. This paper also introduces a dataset from a user study on the effects of frame time spikes on player performance gathered from Lead Rush, and includes gameplay logs with both performance data and QoE results. Some analysis of the dataset is presented to illustrate its use.

With the increasing use of black-box Machine Learning (ML) techniques in critical applications, there is a growing demand for methods that can provide transparency and accountability for model predictions. As a result, a large number of local explainability methods for black-box models have been developed and popularized. However, machine learning explanations are still hard to evaluate and compare due to the high dimensionality, heterogeneous representations, varying scales, and stochastic nature of some of these methods. Topological Data Analysis (TDA) can be an effective method in this domain since it can be used to transform attributions into uniform graph representations, providing a common ground for comparison across different explanation methods. We present a novel topology-driven visual analytics tool, Mountaineer, that allows ML practitioners to interactively analyze and compare these representations by linking the topological graphs back to the original data distribution, model predictions, and feature attributions. Mountaineer facilitates rapid and iterative exploration of ML explanations, enabling experts to gain deeper insights into the explanation techniques, understand the underlying data distributions, and thus reach well-founded conclusions about model behavior. Furthermore, we demonstrate the utility of Mountaineer through two case studies using real-world data. In the first, we show how Mountaineer enabled us to compare black-box ML explanations and discern regions of and causes of disagreements between different explanations. In the second, we demonstrate how the tool can be used to compare and understand ML models themselves. Finally, we conducted interviews with three industry experts to help us evaluate our work.

Analyzing classification model performance is a crucial task for machine learning practitioners. While practitioners often use count-based metrics derived from confusion matrices, like accuracy, many applications, such as weather prediction, sports betting, or patient risk prediction, rely on a classifier's predicted probabilities rather than predicted labels. In these instances, practitioners are concerned with producing a calibrated model, that is, one which outputs probabilities that reflect those of the true distribution. Model calibration is often analyzed visually, through static reliability diagrams, however, the traditional calibration visualization may suffer from a variety of drawbacks due to the strong aggregations it necessitates. Furthermore, count-based approaches are unable to sufficiently analyze model calibration. We present Calibrate, an interactive reliability diagram that addresses the aforementioned issues. Calibrate constructs a reliability diagram that is resistant to drawbacks in traditional approaches, and allows for interactive subgroup analysis and instance-level inspection. We demonstrate the utility of Calibrate through use cases on both real-world and synthetic data. We further validate Calibrate by presenting the results of a think-aloud experiment with data scientists who routinely analyze model calibration.

Estimating a team’s win probability at any given point of a game is a common task for any sport, including esports, and is important for valuing player actions, assessing profitable bets, and engaging fans with interesting metrics. Past studies of win probability in esports have relied on data extracted from matches held in well-structured and organized professional tournaments. In these tournaments, players play on set teams, oftentimes where players are well acquainted with all participants. However, there has been little study of win probability modeling in casual gaming environments – those where players are randomly matched – even though these environments form the bulk of gaming hours played. Using interpretable win probability models trained on large CSGO data sets, we improve upon the current state of the art in Counter-Strike: Global Offensive (CSGO) win probability prediction. We identify important features, such as team HP and equipment value, across different skill levels. We also find a small benefit to using Elo-based player skill estimates in predicting win probability. Furthermore, we discuss how our win probability models can be used to investigate the problem of player-leaving in competitive matchmaking.

While esports organizations are increasingly adopting practices of conventional sports teams, such as dedicated analysts and data-driven decision-making, video-based game review is still the primary mode of game analysis. In conventional sports, advances in data collection have introduced systems that allow for sketch-based querying of game situations. However, due to data limitations, as well as differences in the sport itself, esports has seen a dearth of such systems. In this paper, we leverage player tracking data for Counter-Strike: Global Offensive (CSGO) to develop ggViz, a visual analytics system that allows users to query a large esports data set through game state sketches to find similar game states. Users are guided to game states of interest using win probability charts and round icons, and can summarize collections of states through heatmaps. We motivate our design through interviews with esports experts to especially address the issue of game review. We demonstrate ggViz's utility through detailed case studies and expert interviews with coaches, managers, and analysts from professional esports teams.

Predicting outcomes in sports is important for teams, leagues, bettors, media, and fans. Given the growing amount of player tracking data, sports analytics models are increasingly utilizing spatially-derived features built upon player tracking data. However, player-specific information, such as location, cannot readily be included as features themselves, since common modeling techniques rely on vector input. Accordingly, spatially-derived features are often constructed in relation to anchor objects, such as the distance to a ball or goal, through global feature aggregations, or via role-assignment schemes, where players are designated a distinct role in the game. In doing so, we sacrifice inter-player and local relationships in favor of global ones. To address this issue, we introduce a sport-agnostic, graph-based representation of game states. We use our proposed graph representation as input to graph neural networks to predict sports outcomes. Our approach preserves permutation invariance and allows for flexible player interaction weights. We improve upon state of the art for prediction tasks in both American football and Counter-Strike, a popular esport, reducing test set loss by 9% and 20%, respectively. Furthermore, we show how our approach can be used to answer "what if" questions.

Esports, despite its expanding interest, lacks fundamental sports analytics resources such as accessible data or proven and reproducible analytical frameworks. Even Counter-Strike: Global Offensive (CSGO), the second most popular esport, suffers from these problems. Thus, quantitative evaluation of CSGO players, a task important to teams, media, bettors and fans, is difficult. To address this, we introduce (1) a data model for CSGO with an open-source implementation; (2) a graph distance measure for defining distances in CSGO; and (3) a context-aware framework to value players' actions based on changes in their team's chances of winning. Using over 70 million in-game CSGO events, we demonstrate our framework's consistency and independence compared to existing valuation frameworks. We also provide use cases demonstrating high-impact play identification and uncertainty estimation.

Class imbalance problems manifest in domains such as financial fraud detection or network intrusion analysis, where the prevalence of one class is much higher than another. Typically, practitioners are more interested in predicting the minority class than the majority class, as the minority class may carry a higher misclassification cost. However, classifier performance deteriorates in the face of class imbalance as oftentimes classifiers may predict every point as the majority class. Methods for dealing with class imbalance include cost-sensitive learning or resampling techniques. In this paper, we introduce DeepBalance, an ensemble of deep belief networks trained with balanced bootstraps and random feature selection. We demonstrate that our proposed method outperforms baseline resampling methods such as SMOTE and under- and over-sampling in metrics such as AUC and sensitivity when applied to highly imbalanced financial transaction data sets. Additionally, we explore performance and training time implications of various model parameters. Furthermore, we show that our model is easily parallelizable, which can reduce training times. Finally, we present an implementation of DeepBalance in R.

The construction of data analysis infrastructures that handle continuously accumulating data is quickly becoming an essential requirement for many organizations such as the U.S. Department of Energy (DOE). While DOE supports some of the largest computing facilities in the world, new analysis infrastructures like Apache Spark are difficult to implement. In this paper, we propose an on-demand Spark service that mitigates these difficulties, allowing facility users to flexibly create Spark instances quickly and easily. We define a systematic approach for creating these Spark instances and validate that optimal performance benefits are maintained. Using a series of benchmarks for algorithms that are commonly used in scientific workflows, we compared the behavior of Spark tasks using facility resources with that of an open research cloud that has a dedicated Spark infrastructure deployed. Finally, we leveraged a scientific use case from the Center of Nanophase Materials at the Oak Ridge National Laboratory to demonstrate the utility of using Spark in the computing facility.

Data driven science, accompanied by the explosion of petabytes of data, has called into need dedicated analytics computing resources. Dedicated analytics clusters require large capital outlays due to their expensive hardware requirements. Additionally, if such resources are located far from the data they analyze, they also incur substantial data transfer, which has both cost and latency implications. In this paper, we benchmark a variety of high-performance computing (HPC) architectures for classic data science algorithms, as well as conduct a cost analysis of these architectures. Additionally, we compare algorithms across analytic frameworks, as well as explore hidden costs in the form of queuing mechanisms. We observe that node architectures with large memory and high memory bandwidth are better suited for big data analytics on HPC hardware. We also conclude that cloud computing is more cost effective for small or experimental data workloads, but HPC is more cost effective at scale. Additionally, we quantify the hidden costs of queuing and how it relates to data science workloads. Finally, we observe that software developed for the cloud, such as Spark, performs significantly worse than pbdR when run in HPC environments.