This book is alternately enlightening and infurating. On the one hand, it is a seminal text for game design because it integrates games into the existing traditions of literary theory in a more thoughtful way than had previously been done. On the other hand, at the time it was written, it seems to have been intended primarily as a volley in an ongoing culture war in a tiny corner of digital humanities.

Aarseth calls bullshit on certain then-widespread claims about hypertext by pointing to actually-existing hypertext systems like StorySpace and the World Wide Web and noting that they don't provide the claimed capabilities -- and here, he is right. Then he extends the criticism to Ted Nelson, whose ideas about hypertext Aarseth understands no better than Tim Berners-Lee did: a mistake, because Ted's proposed (and prototyped) systems do, indeed, have those features. This is a pattern Aarseth repeats throughout the book: he makes perfectly valid and trenchant criticisms of an existing idea, only to either extend those criticisms beyond reasonable bounds or propose a 'solution' that has exactly the same problems.

This book is also an interesting time capsule of ergodic literature in the 90s. The MUD transcripts show that internet culture has not changed since 1996 (being, as they are, a bunch of furries role-playing that they are licking and cuddling each other). The section on Racter has a very 80s flavor but demonstrates exactly the same kinds of problems we see in modern entertainment-oriented "AI" systems.

This is a fun book to read (especially if you like to be angry about things that don't matter), and it's a historically important book. It is also an enlightening book, because it reframes now-common experiences in interesting ways that have not seen widespread use.