Skip to content
Wordalisation: Learn how to combine words and models
2.3 Matt Pocock's Skills
Initializing search
Wordalisation: Learn how to combine words and models
Home
Agentic AI for Research
Agentic AI for Research
1 How agentic AI is changing academia
2 Difference between generative and agentic AI
3 Doing a statistical test using Claude
4 A Skill as an example of context
5 Checking the methods are correct
6 How reliably can LLMs perform statistical testing?
7 Claude Code, Obsidian and Visual Studio
8 Humans in the age of ubiquitous AI
9 How agentic AI is affecting researchers' work
10 What can you do in this new world?
11 Agentic AI can do your students' homework
Materials
Materials
Course materials
The cancer-journal-plot skill
The journal-figure skill
The slide-deck skill
Agentic AI for Statistics and Data Science
Agentic AI for Statistics and Data Science
Overview
Part 1: A worked example, end to end
Part 1: A worked example, end to end
1.1 Step by step working with AI
1.2 Importing data
1.3 Deciding on a hypothesis
1.4 Visualisation
1.5 Traditional statistics versus Agentic statistics
1.6 Statistical analysis
1.7 Turning analysis into a report
1.8 Best practices for working with agentic AI
Part 2: Adapting it to your own work
Part 2: Adapting it to your own work
2.1 How context is setup
2.2 Adapting your own framework
2.3 Matt Pocock's Skills
2.4 A football data framework
2.5 A statistical machine learning framework
2.6 Ethical aspects of Agentic AI
2.7 Fully automated research?
Materials
Materials
Course materials
Kujenga
Kujenga
1 About this course
2 Are healthier people happier people?
3 Can we predict a disease outbreak?
4 Are Ethiopian or Kenyan runners the fastest?
5 How did Google make billions of dollars?
6 Final project
Materials
Materials
Course materials
Soccermatics
Soccermatics
0 Getting started
0 Getting started
0.1 Getting Started
0.2 Using Statsbomb
0.3 Using Wyscout
0.4 Sign test
0.5 Using t-tests
1 Event data
1 Event data
1.1 Visualising football
1.2 Plotting shots
1.3 Plotting passes
1.4 Passing networks
1.5 Case study: decentralised football
1.6 Pass heat maps
1.7 Case study: attacking play
1.8 Case study: 5-second rule
1.9 Improving your visuals
2 Models
2 Models
2.1 Statistical models
2.2 Linear regression
2.3 Case study: Possession
2.4 Introducing expected goals
2.5 Fitting the xG model
2.6 Case study: the geometry of shooting
2.7 History and application of xG
3 Scouting
3 Scouting
3.1 Statistical scouting
3.2 Radar plots
3.3 Case study: spreadsheet scouting
3.4 Other approaches to scouting
4 Valuing actions
4 Valuing actions
4.1 Evaluating actions
4.2 Markov chains
4.3 Expected Threat - Position-based
4.4 Calculating xT (position-based)
4.5 Expected Threat - Action-based
4.6 Possesion Chains
4.7 Calculating xT (action-based)
4.8 Case study: value and context
5 Randomness and prediction
5 Randomness and prediction
5.1 Simulating a match
5.2 Unavoidable randomness
5.3 Simulating results
5.4 Predicting football matches
5.5 Calculating match outcomes
5.6 Statistical modelling
5.7 Team KPIs
5.8 Points Moving Average
5.9 Case study: Long term trend at Hammarby
6 Pitch control
6 Pitch control
6.1 Controlling Space
6.2 Making Voronoi Diagrams
6.3 Case study: defence in attack
6.4 Pass probability
7 Positioning and movement
7 Positioning and movement
7.1 Valuing off-ball actions
7.2 Case study: off-ball runs
7.3 Off-ball movement metrics
7.4 Expected Goals including player positions
8 Physical data
8 Physical data
8.1 Physical data
8.2 Calculate distance-based metrics for player fitness
8.3 Modelling the relationship between EPV added and distance covered
9 Clustering
9 Clustering
9.1 Clustering progressive passes
9.2 Clustering players
10 Making a web app
10 Making a web app
10.1 Installing Streamlit
10.2 Twelve Visualisations
10.3 Twelve Community
Materials
Materials
Course materials
Coming soon
2.3 Matt Pocock's Skills
¶
Available Monday 23 November 2026.
Back to top