Player performance evaluation is a key aspect of sports analytics. Teams need to evaluate performance on player and team level for roster management, scouting, roster analysis and strategy. When evaluating the performance of ice hockey players, team analysts work together with coaches, scouts and general managers to extract insights from game and player data that may help build a better team and game strategies. Many metrics are used to evaluate player and team performance, including traditional metrics such as goals, assists, and points and more recent metrics such as Corsi and expected goals. One weakness of most such metrics is that they do not consider the context in which the value for the metric was assigned. In this chapter, we briefly discuss state of the art in player performance metrics for ice hockey and then show two examples of approaches for defining new performance metrics that use large amounts of game event data and can take context into account.

The Internet of Drones (IoD) enables coordinated UAV communication for critical applications such as disaster response, military operations, and smart agriculture. As IoD networks scale and integrate with next-generation wireless systems (e.g., 6G), securing communication between resource-constrained drones becomes essential, particularly against replay attacks that can compromise safety and mission integrity. We present JKA-HT, a lightweight authentication and key agreement protocol offering robust, time-independent replay protection and comprehensive formal security guarantees. Building on the Julia Key Agreement (JKA), JKA-HT introduces HASHTAG, a novel cryptographic tagging mechanism that achieves replay resilience without relying on synchronized clocks or special hardware. We formally verify JKA-HT in the TAMARIN prover under the Dolev-Yao threat model, proving mutual authentication, forward/backward secrecy, Key Compromise Impersonation (KCI) resistance, and replay protection. JKA-HT is the first IoD authentication protocol to achieve all these properties with formal proofs, making it well-suited for deployment in adversarial, resource-constrained environments.

Encrypted video streaming limits network operators’ visibility into user behavior and complicates Quality of Experience (QoE) estimation. While existing QoE prediction models largely assume smooth, linear playback, real-world videoon-demand (VoD) viewing is often interactive: users frequently jump forward or backwards to skip or replay content. Such interactions invalidate buffered data, trigger bursty segment requests, and introduce non-stationary traffic patterns that violate the assumptions underlying prior models. In this paper,we present the first systematic study of QoE prediction of encrypted VoD traffic that accounts for user jumping behavior. Using controlled experiments with time-varying bandwidth and explicit jump events, we show that a state-of-the-art QoE classifier trained on smooth playback degrades sharply once jumps occur, with accuracy collapsing under frequent interaction.

We further demonstrate that jumping behavior leaves a clear fingerprint in encrypted traffic: jump and non-jump sessions are perfectly distinguishable, and interaction intensity can be inferred reliably across bandwidth regimes. Leveraging these insights, we introduce PB-QoE, a lightweight and interpretable jump-aware QoE prediction pipeline that combines continuous regression, trained on non-jumping data, with an interactiondependent penalty. Our evaluation across bandwidth conditions,jump frequencies and directions, and video categories shows that PB-QoE consistently outperforms classification- and regressiononly baselines, enabling robust QoE estimation from encrypted traffic even under highly interactive playback.

Hyperpartisan news sources increasingly shape online discourse, raising urgent concerns about polarization and misinformation. This paper presents the first large-scale differential study of sentiment-driven engagement with hyperpartisan and non-hyperpartisan news on X (formerly Twitter). Analyzing 5.8 million news tweets that attracted 78.6 billion views (impressions) and 466 million interactions, we normalize interactions by view count to isolate user responsiveness, advancing beyond prior work that did not account for differences in exposure. Recognizing that the X algorithm tends to expose hyperpartisan users to hyperpartisan content more frequently, we provide insights into both what sentiments may make some publishers more successful and, more importantly, what effect those different sentiments have on these users’ engagement patterns at a large scale. Our findings show that hyperpartisan users exhibit greater sensitivity to negative content, a pattern robust to controls for topic, content prominence, and temporal variations, and consistent across left- and right-leaning hyperpartisan groups, despite subtle differences. Moreover, our analysis shows that the impact of sentiment on engagement is moderated by the depth of user cognition, varying across different interaction types such as likes, retweets, replies, and quotes. Our findings offer statistically grounded and actionable insights for content creators, recommender-system designers, and policymakers seeking to understand and curb the amplification of hyperpartisan content.

This paper analyzes how different rhetorical attributes in news tweets, specifically analytical, clout, perceptual, and risk language, influence user engagement across publishers with different bias and reliability ratings. Using the LIWC framework to quantify these linguistic dimensions in a 5.5million tweets dataset covering 1,553 news publishers and capturing over 480 million tweet interactions, we perform and present a category-based analysis that captures the relative impact that such features have on the user engagement rates associated with different political bias and reliability categories. While highly biased and unreliable publishers saw increased engagement for clout and risk language, confirming audience biases, the least biased ones benefited more from analytical language. Perception language, on the other hand, uniformly reduced engagement. These insights not only further our understanding of persuasion tactics but also have implications for curbing misinformation by aligning recommendations with audience veracity and impartiality preferences.

Performance and age have an intertwined relationship in soccer. However, despite the sport’s popularity and the great value of insight into this topic to decision-making personnel, this is a fairly unexplored topic, particularly for lower-ranked leagues. In this paper, we use a novel performance metric to investigate the peak age among soccer players in Sweden for goalkeepers, defenders, midfielders, and forwards. Each playing position is analyzed separately using univariate and bivariate methods and is also modeled by a hierarchical Bayesian model with player-specific age trajectories.

The security and integrity of TLS certificates are essential for ensuring secure transmission over the Internet and protecting millions of people from man-in-the-middle attacks. Certificate Authorities (CAs) play a crucial role in issuing and managing these certificates. This paper presents a longitudinal analysis of certificate chains for popular domains, examining their evolution over time and across different categories. Using publicly available certificate data, primarily from crt.sh, we created a longitudinal dataset of certificate chains for domains from the Tranco top-1M list. After categorizing the certificates based on their type and service category, we analyze a selected set of domains over time and identify the patterns and trends that emerge in their certificate chains. Our analysis reveals several noteworthy trends, including a trend towards shorter certificate chains and fewer paths from domains to root certificates. This implies that the certificate process is becoming more simplified and streamlined. Combined with our observations that there is an increasing use of new CAs and a shift in the types of certificates used that we observe, we expect part of this to be an effect of individual choices made by some popular CAs (e.g., less cross-signings). In general, the observed trends, patterns, and findings capture tradeoffs in overhead, backward compatibility, and security. The quick shifts in some of the observed metrics (e.g., chain lengths) therefore also highlight the importance of continued monitoring and analysis of certificate chains.

Although analytics is being used in, e.g., the evaluation of players and scouting, it is still challenging to quantify skills and playing styles of players. Such information is important for roster creation and scouting, where teams want to find players that have a playing style that fits within the team, as well as for game preparation to understand the playing style of opponents. In this paper we use player vectors to characterize a player’s playing style. The player vectors contain representations of skills that are computed from game event data. Further, we use fuzzy clustering on the vectors to generate five types of defender playing styles and five types of forward playing styles. For these types, we show the typical skill levels and players with similar styles.

Understanding the role of a particular player, or set of players, in a team is an important tool for players, scouts, and managers, as it can improve training, game adjustments and team construction. In this paper, we propose a probabilistic method for quantifying player roles in ice hockey that allows for a player to belong to different roles with some probability. Using data from the 2021–2022 NHL season, we analyze and group players into clusters. We show the use of the clusters by an examination of the relationship between player role and contract, as well as between role distribution in a team and team success in terms of reaching the playoffs.