Probability involves a lot of extra-mathematical concepts and funny transforms. I have always had a hard time internalizing it. When I heard a couple years ago that Lemurs are actually capable of abstract math, that gave me a bit of motivation to try to finally conceptualize probability in terms of whole numbers, counts, geometric shapes, basically easy-to-remember mathematical primitives.<p>This book might be interesting to folks interested in predictive science, research, or just anyone who may have taken probability or statistics and might like a refresher on some of the edges of these sometimes tough to nail down concepts.<p>I would love any feedback or questions that anyone may have. Thanks!<p>Patrick Delaney
I wrote a condensed primer on probabilities for a few struggling students years ago so I was curious to check this one out. The concept is interesting but I think the execution needs some extra work. In parts I felt like the lemur abstraction obfuscated things rather than elucidating them. I have a lot of criticism, I hope you won't be discouraged by it.<p>You mix concepts from probability and statistics, I feel it would gain from sticking with either or separating them into different parts/books. Crucial probability concepts are relegated to the end (conditional probabilities, Baye's theorem, ...). The plan in general has some issues, I was surprised to see the expected value as one of the last chapters, and you actually reference it dozens of time before it's introduced. For example slide 15 which would mean nothing to someone who'd just started learning about probabilities.<p>I mostly skipped around but took some time to read carefully the chapters on random variables and PDFs because they're usually the hardest topics to grasp for a beginner (mostly because of the confusing notations). The "random variable" chapter doesn't introduce what a random variable is, it's about dependent and independent variables which are more relevant to the topic of statistical modeling. The PDF chapter is alright but I felt like it was stretching the limits of the lemur abstraction and in the end gave up on it entirely when you start talking about curves and area under the curve. Also, a couple of nitpicks purely about the form, but I don't understand why you refer to probability as "chance" throughout this chapter, and your numbers use too many decimals, you have the luxury of choosing your numbers you can use nice round ones expressed as percentages.<p>So to conclude, my advice would be:
1) Rework the plan
2) Drop the statistics
3) Restrain yourself to discrete probabilities. It's enough for the student to get an intuition of all important concepts and it will be much easier to explain. For example the chapter on PDFs could have been on PMFs and it would have been much easier to explain without compromising your abstraction.
I see a ton of people have been using the free code I submitted on hackernews and one person using a Twitter code since yesterday even opted to buy a copy! Hey awesome, thanks!<p>It would be really great to get people's feedback, but you don't have to comment here - I realize it may take some time to digest since it's 200+ pages. Remember if you subscribe you can just send me an email at any point if you come across an error or something that wasn't quite clear, or just any other idea of how I could improve.<p>thanks!