Changelog

Release Summary

First public release of *Agentic Sport Analytics: The Evidence-Contract Field Guide* — a practitioner's field guide for building automated sport analytics from match footage to coaching decisions, measured end to end on pickleball and Australian rugby league.

• 34 chapters covering the full pipeline: capture, calibration, detection, tracking, identity, eventing, modelling, expected value, simulation, VLM grounding, reporting, the agent harness, the live coaching cockpit, practice design, sensors/hardware, deployment & rights, reproduction, and the 2026 frontier.
• Every chapter expanded with the dual-sport application-framework pattern (concrete use cases with pickleball AND rugby league applications, ecosystem tables, runnable skeletons, honest evidence labels).
• ~600 figures generated in the book's burnt-orange / black linework style, every section carrying visuals.
• Evidence-Contract discipline: measured claims carry experiment IDs (E01-E23) with disk-path artifacts; practitioner values carry [verify]; no fabricated numbers, citations, or benchmarks.
• Rugby League coverage added across all chapters (the earlier edition was pickleball-only; the field guide now treats both sports as first-class examples).
• Named *Agentic Sport Analytics* (title renamed from "Agentic Sport Analysis") with the evidence-contract framing.

Published

• Public book repo: imehr/agentic-sport-analytics (README, CHANGELOG, LICENSE, cover, HTML, paged HTML).
• GitHub Release v1.0.0 with EPUB (308 MB) and PDF (187 MB) binaries.
• Piazr website: piazr.github.io/agentic-sport-analysis (34 transformed chapters, index, about with book quote, reference materials, changelog).
• Added to the Applied AI Signals blog "My Books" collection.

Honest notes

• The book's LLM-assisted writing, review, and measurement were conducted locally; all claims are grounded in the experiment artifacts cited in each chapter.
• 2026 model/hardware specifications (DART FPS, SAM 3, M4 TOPS) are marked [verify at print] where they change quickly.