Sports Analytics Weekly by kubeia.io - 39/2023

Gang caught running fake Indian cricket league to dupe Russian gamblers

StatsBomb, Opta and Deltatre have been at the forefront of providing data to media organisations, player agencies and clubs in the past decade, but there’s a lingering question that should be on the mind of all data consumers: how exactly is this done? How, in less than a day, is advanced data for myriad leagues recorded and up to date? Do these numbers magically appear after the games?

We’re six games into the Premier League season, and things are beginning to settle down.Sure, it’s still early to highlight too many trends about team performance at this stage, but as The Athletic has previously shown, it is worthwhile looking at certain metrics that stabilise quicker than goals — or indeed points — which show stylistic intent from each club.Today, we’re looking at teams’ approaches to pressing to pick out the front-foot fiends from the low-block ballers.

Explore the complete Ratings and PlayStyles for the 17,000 players available in EA SPORTS FC™ 24.

It’s been about three years since goals added (g ) was first introduced with the aim of measuring how much value a player adds to their team through on-ball actions. Since then, the g model has been refined and improved, and lots has been learned.

When evaluating quarterback decision making and performance, it's important to consider the trade-offs as a play evolves. When pressured, the ability to judge the fine line between when to throw the ball or scramble or relocate in the pocket to avoid the sack is a key skill.

Moneyball is principally a baseball story of using data analytics to support player recruitmentBut the message is much more general on how to use data analytics as an evidence-based approach to managing sporting performance as part of a David strategy to compete effectively against teams with much greater economic powerThe last twenty years have seen the generalisation of Moneyball both in its transferability to other team sports and its applicability beyond player recruitment to all other aspects of the coaching function particularly tactical analysisThere are two key requirements for the effective use of data analytics to manage sporting performance: (1) there must be buy-in to the usefulness of data analytics at all levels; and (2) the analyst must be able to understand the coaching problem from the perspective of the coaches, translate that into an analytical problem, and then translate the results of the data analysis into actionable insights for the coaches

Meet SportsLabKit: The essential toolkit for advanced sports analytics. Designed for pros and amateurs alike, we convert raw game footage into actionable data.We’re kicking off with soccer and expanding to other sports soon. Need to quantify your game? Make human movement computable with SportsLabKit.

GPT-4 with vision (GPT-4V) enables users to instruct GPT-4 to analyze image inputs provided by the user, and is the latest capability we are making broadly available. Incorporating additional modalities (such as image inputs) into large language models (LLMs) is viewed by some as a key frontier in artificial intelligence research and development [1, 2, 3]. Multimodal LLMs offer the possibility of expanding the impact of language-only systems with novel interfaces and capabilities, enabling them to solve new tasks and provide novel experiences for their users.

In this system card, [4, 5] 1 we analyze the safety properties of GPT-4V.

Welcome to this comprehensive guide on object detection using the latest “KerasCV YOLOv8” model. YOLO object detection models have found their way into countless applications, from surveillance systems to autonomous vehicles. But, what happens when you pair this capability of YOLOv8 under the KerasCV framework? Recently, KerasCV has integrated the famous YOLOv8 detection models into its library. In this article, we explore how to fine-tune YOLOv8 with a custom dataset. Along the way, we will also cover the following points.

The SoccerNet 2023 challenges were the third annual video understanding challenges organized by the SoccerNet team. For this third edition, the challenges were composed of seven vision-based tasks split into three main themes. The first theme, broadcast video understanding, is composed of three high-level tasks related to describing events occurring in the video broadcasts: (1) action spotting, focusing on retrieving all timestamps related to global actions in soccer, (2) ball action spotting, focusing on retrieving all timestamps related to the soccer ball change of state, and (3) dense video captioning, focusing on describing the broadcast with natural language and anchored timestamps. The second theme, field understanding, relates to the single task of (4) camera calibration, focusing on retrieving the intrinsic and extrinsic camera parameters from images. The third and last theme, player understanding, is composed of three low-level tasks related to extracting information about the players: (5) re-identification, focusing on retrieving the same players across multiple views, (6) multiple object tracking, focusing on tracking players and the ball through unedited video streams, and (7) jersey number recognition, focusing on recognizing the jersey number of players from tracklets. Compared to the previous editions of the SoccerNet challenges, tasks (2-3-7) are novel, including new annotations and data, task (4) was enhanced with more data and annotations, and task (6) now focuses on end-to-end approaches. More information on the tasks, challenges, and leaderboards are available on this https URL. Baselines and development kits can be found on this https URL.

Abstract:Training data is a critical requirement for machine learning tasks, and labeled training data can be expensive to acquire, often requiring manual or semi-automated data collection pipelines. For tracking applications, the data collection involves drawing bounding boxes around the classes of interest on each frame, and associate detections of the same "instance" over frames. In a semi-automated data collection pipeline, this can be achieved by running a baseline detection and tracking algorithm, and relying on manual correction to add/remove/change bounding boxes on each frame, as well as resolving errors in the associations over frames (track switches). In this paper, we propose a data correction pipeline to generate ground-truth data more efficiently in this semi-automated scenario. Our method simplifies the trajectories from the tracking systems and let the annotator verify and correct the objects in the sampled keyframes. Once the objects in the keyframes are corrected, the bounding boxes in the other frames are obtained by interpolation. Our method achieves substantial reduction in the number of frames requiring manual correction. In the MOT dataset, it reduces the number of frames by 30x while maintaining a HOTA score of 89.61% . Moreover, it reduces the number of frames by a factor of 10x while achieving a HOTA score of 79.24% in the SoccerNet dataset, and 85.79% in the DanceTrack dataset. The project code and data are publicly released at this https URL.

Modern live sports broadcasts display a wide variety of graphic visualizations identifying key players in a particular play. Traditionally, these graphics are created with extensive manual annotation for post-match analysis and take a significant amount of time to be produced. To create such visualizations in near real-time, automatic on-screen player identification and localization is essential. However, it is a challenging vision problem, especially for sports such as American football where the players wear elaborate protective equipment. In this work, we propose a novel approach which uses sensor data streams captured by wearables to automatically identify and locate on-screen players with low latency and high accuracy. The approach estimates a field registration homography using on-field player positions from RFID sensors, which is then used to identify and locate individual players on-screen. Experiments using American football data show that the method outperforms a deep learning based state-of-the-art(SOTA) visiononly field registration model both in terms of accuracy of the homography and also success rate of correct homography computation. On a dataset of over 150 replay clips, the proposed method correctly estimated the homography for approximately 25% additional clips as compared to the SOTA method. We demonstrate the efficacy of our method by applying it to the problem of rendering visualizations around key players within a few minutes of the live play. The player identification accuracy for these key players was over 96% across all clips, with an end-to-end latency of less than 1 minute.

The fourth edition of Linear Algebra Done Right is on schedule for publication in two months. The hardcover version can be pre-ordered at https://t.co/zWvBFkIHh5 or at https://t.co/81wrCPaVPK. For the electronic version, see https://t.co/ii1ovMFKvH in late November 2023. https://t.co/jjSTBOl9eL" / X

AMD has refreshed its Alveo field-programmable gate arrays (FPGAs), promising a sevenfold improvement in operating latency and the ability to run more complex machine learning algorithms on the customisable silicon.FPGAs are often used by high-frequency traders, an industry in which a delay of a few fractions of a second can be the difference between profit or loss on algorithm-arranged trades. The ability to reprogram FPGAs with faster or more refined trading algorithms that speed transactions makes the machines valuable. Faster and more flexible FPGAs have obvious appeal.

Understanding broadcast videos is a challenging task in computer vision, as it requires generic reasoning capabilities to appreciate the content offered by the video editing. In this work, we propose SoccerNet-v2, a novel large-scale corpus of manual annotations for the SoccerNet video dataset, along with open challenges to encourage more research in soccer understanding and broadcast production. Specifically, we release around 300k annotations within SoccerNet's 500 untrimmed broadcast soccer videos. We extend current tasks in the realm of soccer to include action spotting, camera shot segmentation with boundary detection, and we define a novel replay grounding task. For each task, we provide and discuss benchmark results, reproducible with our open-source adapted implementations of the most relevant works in the field. SoccerNet-v2 is presented to the broader research community to help push computer vision closer to automatic solutions for more general video understanding and production purposes.

In this video, we present our paper: “SoccerNet-v2: A Dataset and Benchmarks for Holistic Understanding of Broadcast Soccer Videos” published at the CVPR 2021 workshop CVsports. We provide 300,000 temporal annotations within 500 soccer games. This allows a 17-class action spotting task, a 13-class camera boundary detection task, and a novel replay grounding task. We provide benchmarks for all these tasks to start an international challenge.

In the past two days I have been engaged in discussions regarding Andrew Gelman’s review of Book of Why.These discussions unveils some of our differences as well as some agreements. I am posting some of the discussions below, because Gelman’s blog represents the thinking of a huge segment of practicing statisticians who are, by and large, not very talkative about causation. It is interesting therefore to understand how they think, and what makes them tick.

The purpose of this book is to present the state-of-the-art as well as current research challenges in applying computer vision to problems in sports. The book was inspired by the CVPR’13 workshop titled Computer Vision in Sports. The workshop was well received and therefore led to the idea of this book. About half of the 14 chapters in this book are extended workshop papers, whereas the rest are invited papers from research groups active within this field. The book is organized into four parts focused on answering the following four research questions:

Where is the Ball?

Where are the Players?

What are they Playing?

What’s Going on?

"Precision: Statistical and Mathematical Methods in Horse Racing" thoroughly discusses the mathematical and statistical methods in handicapping and betting techniques. Differentiations, combinatorics, normal distribution, kernel smoothing and other mathematical and statistical tools are introduced. The jargons and equations are kept to a minimum so that it is easy to understand for most readers. More than 20 professional programs are freely available to download, which can allow readers to easily apply the methodology introduced in the book.

This book can be divided into three main parts: horse handicapping (Chapters 2-6), wagering (Chapters 7-9) and theories in practices (Chapters 10-11). Chapter 1 will explain why long term gains are possible in horse racing. About horse handicapping, we will start with analysing racing forms in Chapter 2. Other handicapping factors such as weight carried, jockeys, trainers and pedigrees will be discussed in Chapter 3. Some advanced statistical methods, such as chi-square test and kernel smoothing, will be introduced in Chapter 4 to further analyse those handicapping factors discussed in previous chapters. The following two chapters are about probability estimations. In Chapter 5, normal distribution and multinominal logistic regression are introduced in estimating winning probability of each race horse. In Chapter 6, we will talk about some methods in misconceptions in estimating placed probability.

Two main concepts in wagering, Kelly criterion and hedging, will be discussed in Chapters 7 and 8. To hit exotic pools, those theories in combinatorics in Chapter 9 will definitely help the readers. The author will share his experiences in betting syndicate in Chapter 10, and tell you how to be a successful professional horseplayer in the last Chapter.