Build and evaluate the machine-learning models sports teams actually use in Python, from regression to neural nets. It is the applied ML skillset employers screen for.
Taught by Patrick McFarlane. Included with AthlyticZ Membership.
Patrick McFarlane
Director of Predictive Modeling
Who this is for
Python users who want to build real machine-learning models on sports data, end to end.
Not for you if
You've never written Python
You want R, not Python
You want theory without building
You want a shallow survey, not applied depth
What you leave with
A full ML pipeline in scikit-learn
GLMs, GAMs, trees, and neural networks on real sports data
Honest model evaluation and selection
Automated pipelines with PyCaret and an intro to Bayesian ML
Patrick McFarlane introduces sports dataThe opening of Python MachineZ: what sports data is, where it comes from, and why it rewards the methods this course builds.
The full curriculum
11 sections · 43 lessons. Every section is listed with its lesson count; click any section to see its lessons.
Expand all
01Introduction to Advanced Machine Learning in Sports Analytics4 lessons
1.1Machine Learning in Sports
1.2Overview of Sports Data
1.3Overview of Machine Learning
1.4Syllabus Review
02Linear Regression and The Machine Learning Pipeline5 lessons
2.1Linear Regression and Least Squares
2.2Predicting PGA Tour Drive Distance
2.3Dataset Splits
2.4Normalizing and Regularization
2.5Model Construction and Performance
03Exploratory Data Analysis and Feature Engineering3 lessons
3.1Motivating Data Exploration
3.2Data Visualization
3.3Feature Engineering
Career payoffA clean scikit-learn pipeline is the single most-screened-for artifact in ML interviews.
04Hyperparameter Tuning and Model Selection4 lessons
4.1Introduction to Hyperparameters
4.2Cross Validation
4.3Hyperparameter Tuning
4.4Model Selection
05Model Performance Evaluation3 lessons
5.1Performance Metrics
5.2Choosing Evaluation Criteria
5.3Exploring Model Behavior
06Automating Machine Learning Pipelines3 lessons
6.1Motivating Machine Learning Pipelines
6.2Scikit-Learn Pipelines
6.3Pycaret
Career payoffKnowing when to reach for GAMs vs. trees vs. nets is the judgment employers pay for.
07Generalized Linear Models4 lessons
7.1Motivating Generalized Linear Models
7.2Outcome Distributions
7.3Link Functions
7.4NFL Field Goal Probability
08Generalized Additive Models3 lessons
8.1Motivating Generalized Additive Models
8.2Basis Functions and B-Splines
8.3NBA In-Game Win Probability
Modules 08 and 10 · NBA win probability
Down 8, six minutes left. What are your odds?
Win probability across 282,834 game states, margin against time remaining, from a spline GAM. Then the calibration duel: the GAM at 0.507 test log loss against a PyTorch net at 0.508. When the spline ties the net, ship the one you can explain.
Career payoffHonest evaluation is the part interviews probe hardest, and you practice it on real data.
09Tree-based Methods5 lessons
9.1Motivating Tree-based Methods
9.2Decision Trees
9.3Random Forest
9.4Gradient Boosting Machines
9.5WNBA Shot Probability
Modules 09 and 11 · WNBA shot probability
Two models, one shot chart. Who wins?
The same jump-shot data fitted twice: a random forest of 200 trees, then a hierarchical Bayesian model in PyMC where every player borrows strength from the league. Log loss ties, tree 0.654, gradient boosting 0.655, random forest 0.656, and the Bayesian model adds a posterior for every player.
10Neural Networks5 lessons
10.1Motivating Neural Networks
10.2Neural Network Architecture
10.3Backpropagation
10.4Hyperparameters and Best Practices
10.5Revisiting NBA In-Game Win Probability
11Introduction to Applied Bayesian Statistics4 lessons
11.1Bayes' Theorem and Sampling
11.2Introduction to PyMC
11.3Revisiting WNBA Shot Probability
11.4Wrapping Up
What's included
Lifetime accessEvery module and lesson, yours to keep and revisit, forever.
Posit cloud workspaceA provisioned enterprise IDE and compute. Nothing to install.
Project files and codeEvery notebook, dataset, and finished build, to keep and adapt.
All future updatesNew lessons and refreshes as the tools move, at no extra cost.
Self-pacedStart today, go at your own pace, no cohort to wait for.
Included with MembershipOr get this course and the full catalog with Membership.
A practice world built from this course
The lessons teach the method. The practice world is where you
use it, generated from this course's own code, data and recordings.
AthlyticZ WorldsA practice world worked through end to end.
Interactive practice worlds built from each course's own code, data and recordings. Every quantity computed live from the source data, every instructor quotation verified against the recording, and no exercise published unless its reference solution runs and passes its own tests.
Your instructor
Patrick McFarlane
Director of Predictive Modeling
Known for Director of Predictive Modeling
Creator of py_ball, an open-source Python API for NBA and WNBA data
As Director of Predictive Modeling for the Philadelphia Phillies, Patrick builds models that inform in-game strategy, player acquisition, and front-office decision-making. An engineer turned data scientist, he holds a BS in Aerospace Engineering from Notre Dame and an MS from MIT, quantitative training that translates directly into reliable models and clear communication. He has built predictive systems across aviation safety, finance, consulting, and now baseball, so his examples go well beyond sports and apply to any industry. His teaching moves from clean data pipelines to tuned models and executive-ready results, so you can deliver value your stakeholders trust.
Yes. AthlyticZ Membership includes this and the entire course catalog, plus the live Masterclass. If you plan to take more than a couple of courses, Membership is the better math. Both figures are on their order pages.
Is it self-paced?
Yes. Start today and go at your own pace, with lifetime access to every lesson and all future updates.
Do I need to install anything?
No. You work on the provisioned Posit platform in the browser, on the same enterprise tools professional teams use.
Taking more than one course? Get everything.
Membership includes this course, the full course catalog of 1,000+ lessons, and every live Masterclass. If you are taking more than a couple of courses it is the better math, and the figures are on the order pages.
Build and evaluate the machine-learning models sports teams actually use in Python, from regression to neural nets. It is the applied ML skillset employers screen for.
