Listen to any pundit's post-match reaction and you will hear variations of that soundbite. But can you analyse an athlete's state of mind, based on their on-pitch body language? In an era when football is increasingly leaning on data to demonstrate physical attributes, statistics offering an accurate indication of a player's psychological qualities, such as emotional control and leadership, are harder to come by. But Premier League clubs including Brighton are using a technique intended to help in that regard with selection and recruitment. Thomas Tuchel made headlines by telling England's players to communicate more after he evaluated their interactions during the final of Euro 2024, but counting the number of times players gesture or talk to each other on the pitch tells only part of the mental battle being played out.

Canada national team officials have used drones prior to the Paris Olympics and their Tokyo 2020 women's gold medal could be tarnished, officials said on Friday. The developments emerged after Bev Priestman was removed as Olympics head coach for Canada's women's team, following the flying of a drone over New Zealand's training session on Monday. Priestman, 38, was judged as "highly likely" to have been aware of the incident, leading to her suspension by Canada Soccer. Canadian media reported that both of the country's senior teams - men's and women's - have relied on drones for years. Canada Soccer chief executive Kevin Blue confirmed he had received "anecdotal feedback" related to drone use during the men's team's run to the Copa America semi-finals this summer and that coach Jesse Marsch had only been made aware of it after the event.

The Canadian Olympic Committee has removed women's national soccer head coach Bev Priestman for the remainder of the Paris Games after staffers allegedly used a drone to spy on an opponent. Two Canadian team staffers, assistant coach Jasmine Mander and analyst Joseph Lombardi, were "sent home immediately" for allegedly using a drone to spy on a New Zealand practice. Canada beat New Zealand, 2-1, Thursday. Priestman, who has denied involvement, initially volunteered to step away from the club prior to the committee's decision. Canada Soccer CEO and General Secretary Kevin Blue said in a COC release that "additional information has come to our attention regarding previous drone use against opponents, predating the Paris 2024 Olympic Games."

To recruit good players in the new College Football 25 video game from EA Sports, you must be willing to engage in activities most adults would find odd. You will press a button to scour the social media of a high schooler. You will send a direct message to that high schooler. You will try to make that high schooler interested enough in you that they will accept when you offer to book them a visit to see you. Your significant other will think it's extremely weird that you stay up doing this until well after midnight.

Existing work on Temporal Question Answering (TQA) has predominantly focused on questions anchored to specific timestamps or events (e.g. "Who was the US president in 1970?"). Little work has studied questions whose temporal context is relative to the present time (e.g. "Who was the previous US president?"). We refer to this problem as Present-Anchored Temporal QA (PATQA). PATQA poses unique challenges: (1) large language models (LLMs) may have outdated knowledge, (2) complex temporal relationships (e.g. 'before', 'previous') are hard to reason, (3) multi-hop reasoning may be required, and (4) the gold answers of benchmarks must be continuously updated. To address these challenges, we introduce the PAT-Questions benchmark, which includes single and multi-hop temporal questions. The answers in PAT-Questions can be automatically refreshed by re-running SPARQL queries on a knowledge graph, if available. We evaluate several state-of-the-art LLMs and a SOTA temporal reasoning model (TEMPREASON-T5) on PAT-Questions through direct prompting and retrieval-augmented generation (RAG). The results highlight the limitations of existing solutions in PATQA and motivate the need for new methods to improve PATQA reasoning capabilities.

With the widespread adoption of large language models (LLMs) in numerous applications, the challenge of factuality and the propensity for hallucinations raises significant concerns. To address this issue, particularly in retrieval-augmented in-context learning, we introduce the hierarchical graph of thoughts (HGOT), a structured, multi-layered graph approach designed to enhance the retrieval of pertinent passages during in-context learning. The framework utilizes the emergent planning capabilities of LLMs, employing the divide-and-conquer strategy to break down complex queries into manageable sub-queries. It refines self-consistency majority voting for answer selection, which incorporates the recently proposed citation recall and precision metrics to assess the quality of thoughts, linking an answer's credibility intrinsically to the thought's quality. This methodology introduces a weighted system in majority voting, prioritizing answers based on the citation quality of their thoughts. Additionally, we propose a scoring mechanism for evaluating retrieved passages, considering factors such as citation frequency and quality, self-consistency confidence, and the retrieval module's ranking. Experiments reveal that HGOT outperforms other retrieval-augmented in-context learning methods, including Demonstrate-Search-Predict (DSP), ReAct, Self-Ask, and Retrieve-then-Read on different datasets by as much as $7\%$, demonstrating its efficacy in enhancing the factuality of LLMs.

Understanding time is a pivotal aspect of human cognition, crucial in the broader framework of grasping the intricacies of the world. Previous studies typically focus on specific aspects of time, lacking a comprehensive temporal reasoning benchmark. To address this issue, we propose TimeBench, a comprehensive hierarchical temporal reasoning benchmark that covers a broad spectrum of temporal reasoning phenomena, which provides a thorough evaluation for investigating the temporal reasoning capabilities of large language models. We conduct extensive experiments on popular LLMs, such as GPT-4, LLaMA2, and Mistral, incorporating chain-of-thought prompting. Our experimental results indicate a significant performance gap between the state-of-the-art LLMs and humans, highlighting that there is still a considerable distance to cover in temporal reasoning. We aspire for TimeBench to serve as a comprehensive benchmark, fostering research in temporal reasoning for LLMs. Our resource is available at https://github.com/zchuz/TimeBench

Natural language reasoning plays an increasingly important role in improving language models' ability to solve complex language understanding tasks. An interesting use case for reasoning is the resolution of context-dependent ambiguity. But no resources exist to evaluate how well Large Language Models can use explicit reasoning to resolve ambiguity in language. We propose to use ambiguous definite descriptions for this purpose and create and publish the first benchmark dataset consisting of such phrases. Our method includes all information required to resolve the ambiguity in the prompt, which means a model does not require anything but reasoning to do well. We find this to be a challenging task for recent LLMs. Code and data available at: https://github.com/sfschouten/exploiting-ambiguity

All the sessions from Transform 2021 are available on-demand now. That motivational speech a coach or business executive gives you might one day soon be generated by an AI assistant alongside other bits of timely advice and insight. SportsBettingDime, in collaboration with research lab OpenAI, has been experimenting with AI in the form of a GPT-3 text editor to emulate a coach that provides both play-calling advice and motivational speeches based on the situation a team is currently facing. GPT-3 is an AI language model developed by OpenAI that employs a Transformer model to create content in any voice, style, or tone by leveraging assets freely available on the internet. The basic idea is to aggregate play calls made by other head coaches facing similar situations alongside all the best motivational speeches ever given by head coaches.

The football frenzy is with us; people follow their favourite teams religiously, sipping all sorts of unhealthy beverages and munching truckloads of yummy snacks. TV stations bombard us with all kinds of analyses about what's happening in the various international competitions and keep us informed about the odds of seeing our favourite team raise that much-desired cup. But did you ever ask yourself if Artificial Intelligence (AI) has any role in all this? Of course, we know some of these technologies, such as the Goal Line Technology (GLT) and Video Assistant Referee (VAR). These systems help referees to make the right decisions during matches.