Ian Millington is a partner of IPR Ventures, a consulting company developing next-generation AI technologies for entertainment, modeling, and simulation. Previously he founded Mindlathe Ltd, the largest specialist AI middleware company in computer games, working with on a huge range of game genres and technologies. He has a long background in AI, including PhD research in complexity theory and natural computing. He has published academic and professional papers and articles on topics ranging from paleontology to hypertext.

Simulating physics helps cutting-edge games distinguish themselves by making virtual objects behave as we expect them to in the real world. Physics engines are the software programs that run these simulations. Building an engine is difficult, however. There are a large number of new developers (and hobbyists) coming into this market who need help through this complex process. Current introductory books are inadequate; they don’t bring enough real-world programming experience to the task. There is a need for an introductory book on game physics with solid coding guidance but which limits the math content. Ian Millington brings his extensive professional programming experience to this problem. He has developed games since 1987, has studied AI and mathematics at the PhD level, and founded Mindlathe Ltd., a company that designed and built commercial physics engines. Physics Engine Development carefully describes each step in the creation of a robust, usable physics engine. It introduces the mathematical concepts in a clear and simple manner, keeping to high school level topics and building a physics code library as it goes. Each new concept is explained in diagrams and code to make sure that even the most novice of game programmers understands. The companion CD-ROM includes the source code for a complete physics engine of commercial quality. This book will serve as a introduction to more mathematically advanced books on game physics, such as Dave Eberly’s Game Physics.

* Uses only high school algebra
* Shows how to build a complete system based on professional principles
* CD-ROM with C++ source code for a full commercial-quality physics engine

A really good introductory book

This book does a great job of covering all the basics required for a physics engine. With the exception of quaternions all other material is high school level Maths and Mechanics so it’s quite easy to follow. It covers both particle and rigid body dynamics. It contains a very accessible introduction to collision detection and resolution. It presents algorithms and considers their stability and optimisation aspects. It’s concise but provides a good base for further exploration.

The text does contain a lot of code though, which in my opinion is a distraction and should have been entirely omitted. This would reduce the size of the book, increase it’s information density, and make it even better. Where absolutely required algorithms could be presented using pseudocode.

only for beginner

The book only for beginner, after read the book, you can’t process stacking and joint, support stacking and joint is crucial for a modern physics engine, if you need these, i recommend sPhysics Based AnimationtCit’s must have for physics engine programmer.

Game Physics Engine Development Review

I agree with some of the other reviews. The book does have some good information in it on developing a small physics engine. However; the author seems to have a cavalier attitude toward the need for accuracy in the engine. The following is a quote from p. 95 in chapter 6 of the book.
"You can safely skipp this section: the mathematics are not explored in detail; there are restrictions on where we can use fake stiff springs, and the formulation is not always guaranteed to work. In particular, while they can fake the effect reasonably on their own, when more than one is combined, physical inaccuracies in the calculation can interact and cause serious problems."
If this is true, then why put the information in the book at all? This is an example of the author’s writing style and it occurs repeatedly throughout the book. If you are looking for a book on physics engine development, I would suggest looking to any of the other authors on the subject. Preferably an author that values accuracy in his physics engine instead of just faking accuracy.

A usefull book

is a very best practical reference to development a physic game engine. Another books, too many mathematics, other, very simple, this it is a book halfway among mathematics and practical code.

Great Book

This book is very thorough and has clear explanations with examples of source code. Very helpful in learning the systematics of a game physics engine.

Very suitable for newcomers to physics engine development

I was pleasantly surprised to find the book an enjoyable read, and easy to follow. It describes physics engine development in a straightforward manner. This book serves as an excellent introduction for those who are new to physics engine development.

It may not be suitable for those who already have experience in this field, and who are looking for more advanced knowledge on the subject. I would recommend it as a primer on the subject - the book to start with before moving on to more complex subject matter.

Tough initially, but excellent

Initially, this book is quite tough - there’s no denying that. Even with knowledge of vectors etc. I did struggle to begin with -obviously this book is based very much on mathematics.

But once you see how the maths is used in code, this book starts to shine - if you’re pro-efficient in C++, this is an excellent book, from the viewpoint of it creating a usable and quite sophisticated physics engine through the duration of the book; it also has a few extras that I was surprised (and pleased) to see in the book (although they are only lightly covered): ragdolls and fracture physics, as well as explosions.

The only place where this book is slightly weak, is that of collision detection. The book says the collision detection is fairly primitive, and covers a few methods for different shape types, but does recommend a stand alone collision engine, however this isn’t terribly convenient seeing as though this is supposed to create a complete usable physics engine. Note however, that this is collision detection in general, not contact resolution; the latter is covered fully and very competently.

Overall, an excellent book; either to build your own, or to give you an insight on how other engines work (which is useful for when you wish to modify them, for example, or understanding the application of certain values, e.g. damping)

Nice but quite tricky for beginners.

Hi,

This is a very nice book for about the first 150 pages where it starts to become quite thin on the ground with regards to adequate explanations of what is going on. It is not really a book suitable for beginners. You will need alot more knowledge than the introductory pages lead you to believe such is the way with oh so many games books on the market today.

Personally I find the book quite frustrating although the author is clearly an absolute master both in theory and practice of his subject but he struggles to get his vast intellect and knowledge across to the reader.

I didn’t want to rate it any less than three stars as it deserves respect because there is a clear attempt to help but it is over-shadowed by a far too brisk pace for beginners in the middle-late chapters. This is a shame really, perhaps if the same pace had been kept from the start it would have been a five star review.

Avoid this one if you are a novice, or like me are just very rusty at physics. If you are an experienced game developer you will probably love this book. Buy it, if your skill level is matched to it.

Good principles and gives you a feel for the workings :)

I found it a good read, and you can in most cases get a feel for what the writer is trying to achieve....he puts across the basic principle and then goes into a few ways of how you would compensate for such things, such as wobbling on the ground, multiple collisions etc. i.e. Introducing a few bodges etc to get a realistic look...rather than follow the maths to the letter.

Its not what I’d call a basic book which usually cover a lot of spheres and single point collisions, but then its not as advanced as eberlys....its a good intermediate book...definetly worth a read if your new to physics.

Happy coding =)

Awesome Book!

The book gives you step by step code on how to create a simple physics engine.

Good introduction to physics programming, but not much else

This book serves as a good introduction to many topics one would fine useful for implementing a physics engine. The writing is very clear, and even a high school student with a good background in mathematics shouldn’t have much difficulty comprehending most chapters. Professional developers should probably just stay clear of this book, and go straight to Dave Eberly’s "Game Physics." Don’t expect to find a robust physics engine in this book either. All demos are pretty rudimentary, poorly organized and even buggy. Nevertheless, true beginners should still find the code easy to read and understand.