In another tab, I’m writing the announcement blog post for the No Bullshit Guide to Statistics. I spent the past seven years teaching myself statistics at the graduate level so that I can explain undergraduate statistics topics in a concise manner, with the focus on the important topics. Now you can benefit from this experience.

I understand some of you might not be too hot about the idea of reading 1100 pages of dense text, math formulas, and code examples, so I prepared a “sales pitch” that explains why statistics is important, and why reading the No Bullshit Guide to Statistics is the best way to learn statistics.

Why do you need to know stats?

Knowing statistics is of strategic importance in the modern world. Scientific research and modern business intelligence are both domains where you need to learn how to tell apart the “signal” from the “noise” in the datasets. If you want to squeeze out insights from your data, you need to know statistics.

Statistics is powerful

The ability to detect “patterns” in data is of great importance in the modern world. Whether you’re a scientist looking to publish papers, or working in industry and need stats to make informed business decisions, your problem is the same: you want to extract knowledge from your data using a rigorous approach.

Statistics is important

I spent seven years on this book because I think it is of strategic importance for the 21st century. I believe “alert and educated citizenry” is needed to keep the System in check. Knowledge is power, and statistical knowledge is some of the most powerful stuff out there.

Can’t someone (or something) else do the stats for me?

Using statistics requires knowing which procedure is appropriate to use in a given situation. Applying the procedures is easy—you just turn the crack and get an answer, but how do you know if this answer is valid? Understanding when you can use statistical analysis requires expertise. You can’t outsource this stuff, because it’s very tricky, with lots of edge cases and pitfalls you must watch out for.

What is so special about this book?

Here are some selling points that I plan to use in the future landing page for the book. You should get this book because:

• It is complete: the topics covered in Part 2 include all the standard STATS101 topics, as well as important stats topics from more advanced courses like linear models and Bayesian statistics.
• It comes with prerequisites included: to understand statistics, you need to know about data management (practical skills) and probability models (theoretical skills), which are both covered in Part 1 of the book. The Appendix in Part 1 includes tutorials on Python (general computational skills), Pandas (data management skills), and Seaborn (data visualizations and plotting).
• It is practical: all statistical analysis techniques are presented with lots of details and examples so that you will learn how to apply them to your own datasets.
• It is modern: statistics is a complicated topic with lots of “historical baggage” and most learning resources out there (textbooks, websites, video lessons) follow the old school approach to teaching statistic. In contrast, this book is based on the modern, computation-first statistics curriculum that makes concepts much easier to understand.
• It is computational: the book prioritizes statistical thinking and hands-on computations, instead of asking readers to memorize a bunch of formulas and follow statistical recipes “blindly.”
• It’s condensed: The book is 1100 pages in total (Part 1 = 433pp, Part 2 = 656pp). This might seem like a lot, but it’s actually only the essentials. The book is long because it presents the same concept from multiple different perspectives, using analytical methods (math) and computational methods (code).
• It comes with lots of auxiliary resources designed to support the learning experience: computational notebooks, datasets, video lessons, concept maps, helper functions, etc.
• It explains things well: I’ve optimized the structure of the text to make sure all concepts are presented as part of a coherent story, starting from prerequisites and leading through all the essential topics, progressively ramping up the complexity.

But what if learning stats is too difficult for me?

I’ve made an effort to make the book accessible to a wide audience. I don’t expect you to have any prior experience. You just need to be willing to learn.

What if I’m not good at math?

Sorry that excuse doesn’t work around here. Yes, the book contains lots of math formulas that you will have to understand, but each formula is explained in plain English, illustrated by graphs, and additionally presented as a code example. We’re talking about quadruple redundancy of explanations, so even if you don’t like math, you can still understand concepts through one of the other modalities. This quadruply-redundant narrative is why the book is so long and took seven years to write.

You have to trust me on this one. Reading a bunch of math formulas is not that bad… No matter what math trauma you might have, reading this book will not add to it. In fact, I hope it might help you see math in a new light. If you understand WHY you need some formula, and you learn what you can DO with it, then you’ll naturally want to learn about the formula to experience some of that math power.

If you’re still not sure about this, check out the first few pages of Appendix B Notation. This will expose you to all the math symbols used throughout the book. If this level of math symbols doesn’t scare you, then you totally have the required math skills.

What if I have no coding experience?

The book doesn’t require any prior programming experience. The Python code examples in the book are not programming, but more like using Python as a scientific calculator. No matter what your previous coding experience is, you can totally handle this stuff.

The book includes a detailed Python tutorial (see also bit.ly/pytut3) for people who want to level-up their Python skill and learn to “play” with the code in the JupyterLab interactive computational environment (in the JupyterLab Desktop app or on a JupyterLab cloud servers). The Python tutorial has been user-tested with absolute beginners, and it proved to be effective as an introduction to Python.

What if I don’t have the time?

A part of you might be thinking that you will never have the time to learn all these topics. Yes, you’ll need to dedicate some hours of caffeinated morning focus time to get through all the material, but it is totally doable.

The book is split into sections that introduce topics one at a time. Each section is of manageable length that fits in one day of learning. I recommend that you split your time equally between reading, playing interactively with the code examples in the notebooks, and solving exercises. Back to school, yo! But don’t worry: as an adult, you’re much better at learning school stuff now. It’s a lot of material, but it’s not going to be hard.

Do I really need to read 1100 pages?

If you’re on this mailing list, it’s because you have expressed interest in seeing the No Bullshit Guide to Statistics become a reality. There are consequences to your actions: when you ask me to write a book on statistics, you have to expect you’re going to get a detailed account of the subject matter. It’s not my fault: y’all made me do this!

If you want to learn stats, the click here gum.co/noBSstats and get started today!