A PERSONAL FIELD GUIDE · 2026
This is the path I use to learn statistics deeply: the books, courses, datasets, simulations, and projects that helped me move from formulas to actual statistical thinking.
This is not an exhaustive collection. It is a curated learning path: a small number of excellent resources, used in the right order, combined with projects.
01 / RANDOMNESS & INTUITION
Before formulas, learn to see randomness.
What does probability represent? Why do averages become stable? Why can random processes create surprisingly predictable behavior?
Essential
Interactive
Beginner
Seeing Theory
Brown University
WHY I USED IT
Visual intuition before equations.
WHAT IT TAUGHT ME
Probability, distributions, inference, regression, sampling.
✓ Completed
Open resource
PROJECT — Randomness Playground · Before trusting probability mathematically, I wanted to see randomness behave.
02 / PROBABILITY
A language for uncertainty.
Statistics becomes much easier once probability stops feeling like a collection of tricks and starts feeling like a language for uncertainty.
P(A | B) = P(B | A) P(A) / P(B)
Essential
Course
Intermediate
Harvard Stat 110
Joe Blitzstein · Harvard University
WHY I USED IT
My main probability foundation; reasoning over memorization.
WHAT IT TAUGHT ME
Conditional probability, expectation, distributions, LLN, CLT, Markov chains.
✓ Completed
Open resource
Deep Dive
Course
Intermediate
MIT 18.05
MIT OpenCourseWare
WHY I USED IT
A second explanation with more exercises and traditional treatment.
WHAT IT TAUGHT ME
Probability and statistical inference.
◐ Learning
Open resource
PROJECT — Casino Simulator · Can a betting strategy beat a game with negative expected value?
03 / STATISTICS WITH DATA
Reason backward from observations.
Probability describes uncertainty. Statistics starts when I have observations and want to reason backward about the world that produced them.
Read concept
Run notebook
Change parameters
Reproduce
Apply elsewhere
“I do not want statistics to remain something I can recognize in a textbook. I want to be able to reproduce it from data.”
Essential
Book + notebooks
Beginner–Intermediate
Think Stats
Allen B. Downey
WHY I USED IT
My practical introduction through datasets and computational experiments.
WHAT IT TAUGHT ME
EDA, distributions, estimation, testing, regression, time series.
✓ Completed
Open resource
Reference
Book
Beginner–Intermediate
OpenIntro Statistics
OpenIntro
WHY I USED IT
A check against the conventional statistics curriculum.
WHAT IT TAUGHT ME
Sampling, design, intervals, tests, regression.
◐ Learning
Open resource
Reference
Book
Intermediate
Introduction to Modern Statistics
OpenIntro
WHY I USED IT
A computational path through modern inference.
WHAT IT TAUGHT ME
Simulation, resampling, bootstrap, randomization.
○ Up next
Open resource
04 / STATISTICAL INFERENCE
What does the data actually allow me to conclude?
Observing data is easy. Deciding what counts as evidence is much harder.
Population →
Sample →
Estimate ± Uncertainty →
Evidence →
Decision
WHAT I LEARN HERE
Sampling distributions
Estimators
Bias and variance
Confidence intervals
Hypothesis testing
P-values
Statistical power
Type I / Type II errors
Bootstrap
Permutation tests
Multiple testing
STATISTICAL INFERENCE LAB
bootstrap a mean
construct confidence intervals
simulate p-values
investigate false positives
visualize statistical power
05 / STATISTICAL LEARNING
From “What can I infer?” to “What can I predict?”
1 — UNDERSTAND
2 — IMPLEMENT
3 — USE
Essential
Book + Python labs
Intermediate
An Introduction to Statistical Learning
ISLP
WHY I USED IT
Where classical statistics connects naturally with machine learning.
WHAT IT TAUGHT ME
Regression, classification, resampling, regularization, trees, deep learning.
◐ Learning
Open resource
06 / BAYESIAN STATISTICS
Update belief as evidence arrives.
Frequentist statistics is not the only way to reason about uncertainty.
prior
× likelihood
→ posterior
Essential
Book + notebooks
Intermediate
Think Bayes
Allen B. Downey
WHY I USED IT
A computational and intuitive Bayesian introduction.
WHAT IT TAUGHT ME
Priors, likelihoods, posteriors, prediction, decisions.
◐ Learning
Open resource
Deep Dive
Book + lectures
Advanced
Statistical Rethinking
Richard McElreath
WHY I USED IT
To think in generative models and explicit assumptions about reality.
WHAT IT TAUGHT ME
Generative modeling, inference, prediction, model criticism.
○ Up next
Open resource
07 / EXPERIMENTATION
When real decisions depend on the result.
A metric increased by 2%. Did the product actually improve, or did randomness fool me?
CONTROL
μ₀
TREATMENT
μ₁
Essential
Book
Intermediate
Trustworthy Online Controlled Experiments
Ron Kohavi · Diane Tang · Ya Xu
WHY I USED IT
Connects statistical theory to real technology experiments.
WHAT IT TAUGHT ME
A/B tests, metrics, power, validity, guardrails.
○ Up next
Open resource
08 / CAUSAL INFERENCE
What would happen if I changed the world?
CORRELATION
X ↔ Y
INTERVENTION
do(X) → Y
Essential
Book
Intermediate
The Effect
Nick Huntington-Klein
WHY I USED IT
A practical route into counterfactual reasoning.
WHAT IT TAUGHT ME
Confounding, DAGs, matching, IV, DiD, RDD.
○ Up next
Open resource
Deep Dive
Book
Intermediate–Advanced
Causal Inference: The Mixtape
Scott Cunningham
WHY I USED IT
Real research applications of causal methods.
WHAT IT TAUGHT ME
Applied causal inference strategies.
○ Up next
Open resource
09 / TIME SERIES
Time changes how data should be split, validated, and interpreted.
Essential
Book
Intermediate
Forecasting: Principles and Practice
Rob J Hyndman · George Athanasopoulos
WHY I USED IT
My structured resource for forecasting and time-series reasoning.
WHAT IT TAUGHT ME
Trend, seasonality, autocorrelation, evaluation, ARIMA.
○ Up next
Open resource
10 / SIMULATION & REAL PROJECTS
Resources teach concepts. Projects reveal whether I understand them.
FIELD NOTES
