Internal Prototype Sport

Match Video Analysis
Player tracking and match reports from a single upload

Upload a match video and get back what an analyst would produce: players and ball tracked, the pitch calibrated, events detected, and a report with possession, expected goals, passing, pressing and formations. The coach confirms or corrects what the models found.

What It Is

Match Video Analysis follows every player and the ball through the match, maps the action onto the pitch and computes distance, speed, heatmaps, possession and pressing. It detects shots, goals, passes, set pieces and formations, estimates expected goals, and assembles a match report with per-player statistics, exportable as CSV or PDF.

Detected events go to a review screen where a coach confirms, rejects or reclassifies each one. Those reviews are used to improve detection, and a new version of a model is adopted only after it has been measured against the current one and a person has approved it.

Inside Match Video Analysis

Completed match analysis with possession, score, expected goals, passes, shots, pressing, formations and ball retention
The automated match report: possession, expected goals, passing, pressing and formations from one upload.
Event review screen listing detected shots, goals and set pieces on a timeline with confidence and confirm or reject controls
Coaches confirm or correct detected events; their reviews feed the next model.
Team heatmaps drawn on a football pitch
Pitch-calibrated team heatmaps, here drawn from simulated positions.
Player statistics table with distance, top speed, time on ball, sprints, work rate and pass completion
Per-player physical and technical metrics.
New analysis upload screen with the detection model selector set to the latest approved model
Starting an analysis, with the latest approved model as the default.

Every screenshot shows demo data: the organisations, people and figures are made up.

Why It Exists

Automatic match analysis is the feature every coach asks about, and the one Nothing Unlucky deliberately leaves out: dedicated camera systems already do it, and the per-match cost of doing it well does not fit a seat priced for volunteers. This platform is where that question is tested properly: what tracking and event detection achieve on real footage, how often a coach has to correct them, and what each run costs.

What It Is Worth

Evidence before a product decision

Measured results for tracking and event detection on real footage, rather than a vendor demo, which is the same discipline we bring to clients choosing whether to build AI features.

Models that improve under control

Detection improves with each round of corrections, and no new model goes live on its own: a person decides.

Coach in the loop

Every detected event can be confirmed or corrected, and those corrections are what the next model learns from.

Who It Helps

Coaches and analysts

Get a match report, player statistics and team heatmaps from a single upload, with the detection model used recorded against every analysis.

Product and data teams

See how a computer vision feature behaves on real-world footage, and what it would cost to run, before committing a product to it.

At a Glance

Status Internal: runs inside Egon Expert Ltd and is not sold on its own; Prototype
Outputs Match report, per-player metrics, heatmaps, detected events, CSV and PDF export
Built by Egon Expert Ltd

Need Something Like This?

We build platforms like Match Video Analysis for clients, and help in-house teams build their own. Start with a free 30-minute conversation about what you need.

Book a Free Consultation

See all products