Sports Analytics Weekly by kubeia.io - 13/2024
- Week 46
- Week 45
- Week 44
- Week 43
- Week 42
- Week 41
- Week 40
- Week 39
- Week 38
- Week 37
- Week 36
- Week 35
- Week 34
- Week 31
- Week 30
- Week 29
- Week 28
- Week 27
- Week 26
- Week 25
- Week 24
- Week 23
- Week 22
- Week 21
- Week 20
- Week 19
- Week 18
- Week 17
- Week 16
- Week 15
- Week 14
- Week 13
- Week 12
- Week 11
- Week 10
- Week 9
- Week 8
- Week 7
- Week 6
- Week 5
- Week 4
- Week 3
- Week 2
- Week 1
⚽ Sports
NFL Draft 2024: The Quarterbacks
statsbomb.com - Will Morgan
With the 2024 NFL draft a little over a month away, the football world is filled with speculation on the who, where and when of a much heralded quarterback class. Following the lead of our preview last year, we’re going to run the rule over these leading draft candidates, with as many as 6 QBs potentially going in the 1st round.
UEFA Competitions Revenue Distribution 2024/25
substack.com - Swiss Ramble
The format of UEFA’s competitions will change next season with more teams taking part and more games being played.This has been well known for some time, but the details of the revenue distribution system have only recently been confirmed. UEFA said that this came “at the end of a process held in cooperation with the European Club Association (ECA) and positive consultation with the European leagues”.The new format takes some understanding, as indeed does the revenue distribution, so this blog will look at these in some detail, aiming to make them a little clearer to the proverbial “man on the street”.
📝 Sports Analytics
Identifying Attacking Overloads And Measuring Their Impact
statsbomb.com - Jaymes Monte
What follows is a summary of my thought process and methods for identifying Attacking Overloads, not necessarily in an effort to convince anyone of their merits, but more so in the hope that it can inspire further ideas and developments in the way that bespoke metrics can be created and KPIs can be evaluated.
👁️ Computer Vision
Creating Better Data: How To Map Homography
statsbomb.com - Iñaki Rabanillo Viloria
Last year, we published the first articles in our 'Creating Better Data' series from our AI team: Machine Learning Engineer Miguel Méndez Pérez explained what homography is and why it's important to our data collection, before Yohahn Ribeiro and Alex Palmer-Walsh explained how we use AI and Computer Vision technology to generate physical metrics. The latest article in this series comes from our Computer Vision Engineer Iñaki Rabanillo Viloria, who explains how exactly an engineer tasked with computing homography could go about calculating the homography of a given sports field or pitch - including example code so you too can calculate homography from still images.
🎙️ Podcast
🎧SBC Podcast #72 – Betting Clever: Horse Racing, Black Market Bookies, Oddschecker Unlimited & Exchanges
smartbettingclub.com - Josh P
In the latest SBC Podcast I am joined by SBC’s Josh as the Betting Clever podcast returns to tackle a whole host of important topics in the betting world.
With this chat recorded during Cheltenham, betting on horse racing takes centre stage, with some recent news about a black market bookmaker discussed in the wider context of getting money down on a sport beset by problems.
Andy Holding, Oddschecker Unlimited, missing account balances, accusatory emails, a listener question and exchange chat are also on the agenda in an action packed show
🧮 Statistics
Kalman Filter Explained Simply
thekalmanfilter.com - William Franklin
Most tutorials for the Kalman Filter are difficult to understand because they require advanced math skills to understand how the Kalman Filter is derived. If you have tried to read Rudolf E Kalman’s 1960 Kalman Filter paper, you know how confusing this concept can be. But do you need to understand how to derive the Kalman Filter in order to use it?No. If you want to design and implement a Kalman Filter, you do not need to know how to derive it, you just need to understand how it works.The truth is, anybody can understand the Kalman Filter if it is explained in small digestible chunks. This post simply explains the Kalman Filter and how it works to estimate the state of a system.
🌩 Forecasting
Moirai: A Time Series Foundation Model for Universal Forecasting
salesforceairesearch.com - Gerald Woo, Chenghao Liu, Doyen Sahoo, Caiming Xiong
Moirai is a cutting-edge time series foundation model, offering universal forecasting capabilities. It stands out as a versatile time series forecasting model capable of addressing diverse forecasting tasks across multiple domains, frequencies, and variables in a zero-shot manner. To achieve this, Moirai tackles four major challenges: (i) construction of a LOTSA, a large-scale and diverse time series dataset, comprising 27 billion observations spanning nine distinct domains, (ii) development of multiple patch size projection layers, allowing a single model to capture temporal patterns across various frequencies, (iii) implementation of an any-variate attention mechanism, empowering a single model to handle forecasts across any variable, and (iv) integration of a mixture distribution to model flexible predictive distributions. Through comprehensive evaluation in both in-distribution and out-of-distribution settings, Moirai demonstrates its prowess as a zero-shot forecaster, consistently delivering competitive or superior performance compared to full-shot models.