Sports Analytics Weekly by kubeia.io - 2/2024

The man who made a living betting against companies turned bullish on DraftKings — and on sports betting companies in general.“The betting numbers have continued to be strong in the U.S., stronger than we thought they’d be,” he told the Financial Times. “The thing that we underestimated — that I think is going to be a benefit for all these companies for a while anyway — is what bad bettors the U.S. gamblers are.” 

Arsenal are in a funny spot right now. They are playing well, perhaps the best they have played all season, but the results are not going their way.

Goran Ivanisevic has seen it happen so many times over the past four years. His star pupil, Novak Djokovic, shows up to the practice court in a foul mood, griping that his game is a disaster, that he needs to get better… at everything. His serve, his attacking play, even his backhand — one of the great backhands tennis has ever seen — it’s all a mess. There is barely any acknowledgement of the resume, the 24 Grand Slam titles, the 74 other tour trophies, and more than 1,000 match victories. He’s got to improve, or he’s cooked.

Let’s accept the findings we’ve learned over the last decade that in general, a team’s (or player’s) ability to convert shots into goals regresses over time to the underlying pre-shot probability of the shots they attempt as modelled based on the usual xG stuff like where they’re taking the shot from, with what body part and some other context. In the long run a team is generally as good as the chances it creates relative to those it concedes. And in the short term, you just flip a coin, or roll some dice.

The history of economic thought witnessed several prominent economists who took seriously models and concepts in physics for the elucidation and prediction of economic phenomena. Econophysics is an emerging discipline at the intersection of heterodox economics and the physics of complex systems, with practitioners typically engaged in two overlapping but distinct methodological programs. The first is to export mathematical methods used in physics for the purposes of studying economic phenomena. The second is to export mechanisms in physics into economics. A conclusion is drawn that physics transfer is often justified at the level of mathematical transfer but unjustified at the level of mechanistic transfer.

Abstract:The most recent large language models(LLMs) such as ChatGPT and GPT-4 have shown exceptional capabilities of generalist models, achieving state-of-the-art performance on a wide range of NLP tasks with little or no adaptation. How effective are such models in the financial domain? Understanding this basic question would have a significant impact on many downstream financial analytical tasks. In this paper, we conduct an empirical study and provide experimental evidences of their performance on a wide variety of financial text analytical problems, using eight benchmark datasets from five categories of tasks. We report both the strengths and limitations of the current models by comparing them to the state-of-the-art fine-tuned approaches and the recently released domain-specific pretrained models. We hope our study can help understand the capability of the existing models in the financial domain and facilitate further improvements.

TwelveGPT is a tool to turn football data in to easy to understand words and visuals. It uses a customised language model of football to understand the game. What can it do? TwelveGPT can write scout reports, it can summarise matches and it can even provide a running commentary of the action on the pitch.Platform, reports or API: We provide our customers with access through an interactive platform, scouting or match reports, and/or access to our API, which returns visuals and words about your data.

As machine learning practitioners, we often encounter imbalanced datasets in which one class has considerably fewer instances than the other. Many machine learning algorithms assume an equilibrium between majority and minority classes, leading to a suboptimal performance on imbalanced data. Addressing class imbalance is crucial for significantly improving model performance. Machine Learning for Imbalanced Data begins by introducing the challenges posed by imbalanced datasets and the importance of addressing these issues. It then guides you through techniques that enhance performance on imbalanced data when using classical machine learning models, including various sampling and cost-sensitive learning methods. As you progress, the book delves into similar and more advanced techniques for deep learning models, employing PyTorch as the primary framework. Throughout the book, hands-on examples provide working, reproducible code that demonstrates the practical implementation of each technique. By the end of this book, you will be adept at identifying and addressing class imbalances, and confidently applying various techniques including sampling, cost-sensitive techniques, and threshold adjustment when using traditional machine learning or deep learning models.