What This Book Actually Is
Paul Zeitz wrote a textbook that treats problem solving as a discipline rather than a talent. The full title is The Art And Craft Of Problem Solving By Paul Zeitz. It targets people who are already comfortable with standard math curriculum but hit a wall when they encounter competition-style problems. That covers a lot of students, coaches, and people who teach math at the high school level. The book is organized around strategies, not content areas. You get chapters on extremal principles, the pigeonhole principle, invariant arguments, recursion, and so on. Each strategy gets a theory section followed by worked examples and then problem sets. The problems range from Olympiad-level to very elementary. The difficulty gradient is intentional.
The Art And Craft Of Problem Solving By Paul Zeitz
Here is how I actually use it. I do not read it cover to cover. I pick a strategy chapter, read the theory, then work problems blind. If I am stuck past twenty minutes, I look at the first hint. If I still cannot solve it after the second hint, I read the solution, write down why I missed the key move, and move on. That process takes about forty-five minutes per problem. Reading the book straight through without solving problems gives you almost nothing. The counter-intuitive thing about this book is that the early chapters are harder to use than the later ones. Chapter 1 on general strategies sounds helpful. It is not, until you have done enough specific work that the generic advice starts mapping onto real patterns. I wasted three weeks on the first chapter before I went back and did the combinatorics problems in later chapters. Then Chapter 1 started making sense because I finally had concrete examples to attach the advice to. Another thing nobody mentions is that Zeitz assumes you know proof writing at a basic level. If you cannot write a clean direct proof or a proof by contradiction, the examples will fly over your head. You do not need much, just the ability to structure a simple argument and recognize when a case split is justified. People skip that prerequisite and then complain the book is too abstract. It is not too abstract. It is just aimed at readers who already have that skill.
Which Edition To Get
There are two main editions. The second edition has updated problem sets and a few corrected proofs. The first edition is cheaper on the used market and the core material is essentially identical. If you are buying new, get the second edition. The difference is small but the errata in the first edition are annoying if you catch them mid-solution. You can find the book through standard retailers, Amazon, or directly from Wiley. The PDF circulates online but I would not recommend it. The layout with boxed examples and the problem numbering matters more than people expect when you are working through chapters. A PDF compresses the visual structure and slows your reading.
Get the Full Details

How To Actually Learn From It
Work the problems. The theory sections are about two pages per chapter on average. The practice problems are where the learning happens. I keep a separate notebook for my attempts. If I solve a problem, I write the full solution cleanly, not just the answer. If I fail, I write down the wrong path I took and where I got stuck. That review step is what turns frustration into retention. Expect to spend roughly six to eight hours per chapter if you are doing the problems thoroughly. Some chapters take longer. The chapter on generating functions is dense and the problem set alone can eat two full days for someone who has not seen generating functions before. The pigeonhole principle chapter is faster because the core idea is simple even though the applications vary in difficulty.
Where This Approach Breaks Down
The book is narrow. It covers competition math, which means discrete math, number theory, combinatorics, and basic algebra. It does not cover calculus-based problem solving, optimization, or applied mathematics. If you are preparing for a math competition at the high school level, this is useful. If you are trying to learn general analytical problem solving for engineering or computer science, this will leave gaps. Another limitation is that the difficulty curve is steep in places. Chapter 4 on invariants has problems that assume a level of creative insight most beginners do not have yet. You will sit on those problems for a long time. I learned to skip ahead and return later. The book is not linear in the sense that every problem builds on the previous one. Some problems in a later chapter rely on ideas from earlier chapters, but not all of them do. If you want a broader foundation before tackling this book, you should work through problem-solving materials that cover proof techniques and basic set theory first. Books like How To Solve It by Polya or the Problem-Solving Strategies by Engel are good complements. Zeitz assumes you have some background. It is not a beginner text in the strict sense.
A Specific Case I Encountered
I ran into a problem in the extremal principle chapter that looked like a standard maximum principle question. The setup involved arranging numbers on a circle with a constraint on adjacent products. I spent about thirty minutes trying to construct a specific arrangement and kept hitting dead ends. The key insight the solution used was to look at the maximum element and derive a contradiction from its neighbors, not to build an arrangement directly. My workaround was to stop trying to construct and instead list all the properties the maximum element must satisfy under the constraint. That reduced the problem to checking five cases instead of an infinite search space. I wrote that shift in reasoning in my notebook. It became a reusable pattern I recognized in three later problems in the same chapter. That is the actual value of this book, not the solutions themselves but the recognition of move types.

Final Notes
This book is not fast. It will not give you quick tricks. The gains come from repeated exposure to the same strategic ideas in different contexts. Most people who finish the problem sets see a measurable improvement in their ability to approach unfamiliar problems. The improvement is slower than people hope but more durable than most shortcut methods. Pair it with a problem log and honest review. Work the chapters out of order if needed. Skip around when a topic feels too dry or too hard. The book is a reference and a training tool at the same time. It works best when you treat it like gym equipment rather than a novel.