The purpose of this section is to provide a mapping among different classification schemes for mechanical puzzles.

Over a century ago, Professor Hoffmann commented on the difficulty of devising and adhering to a rigid scheme for the classification of puzzles - many mechanical puzzles could be classified in multiple equally valid ways. There are two main schemes which have been promoted by major puzzle collectors: Jerry Slocum uses a scheme based on Hoffmann's, and which is now employed by the Lilly Library to catalogue his donated collection of over 30,000 puzzles; and James Dalgety uses a more detailed scheme (see in particular the illustrated summary) to catalogue his enormous Hordern-Dalgety collection (sample catalogue here). The evolution of Dalgety's collection is described in an article he wrote for the CCF newsletter, issue 54 Feb. 2001.

Dalgety's scheme encompasses fourteen main classes, with several sub-classes within each main class, while Slocum makes due with 10 main classes and few subclasses. There is considerable overlap and good correspondence between the two systems. The Wikipedia entry for mechanical puzzles also pretty much agrees with the top level classes.

The classification scheme I have employed on Rob's Puzzle Page is closer to Dalgety's since I prefer finer granularity. I have seen various other schemes around the web, for instance, based on the material from which the puzzle is made - e.g. plastic, wood, glass - but I don't think they're very useful for finding a puzzle or identifying isomorphic or related puzzles. I do think each classification scheme reveals something about the preferences, biases, experience, expertise, and focus of the classifier.

The table below compares and attempts to align the different schemes. I have also found it instructive to include the categories described by van Delft and Botermans in their 1978 classic Creative Puzzles of the World, and the ten sections in Hoffmann's 1893 book, since they help show how the universe of puzzles has evolved over time.

Where I have found expert advice in further classifying particular categories of puzzle, I have highlighted the experts like this.

Recent years have seen a surge in the popularity and visibility (via e.g. social media, videos, Discord) of mechanical puzzles and the number of people who enjoy, collect, create, make, promote, and sell them. New puzzle types and subtypes have come to the fore (e.g. Graduated Challenge, "N-ary," Turning Interlocking Cubes, Sequential Discovery, Restricted Opening Sequential Packing with Rotations, and on) that while perhaps in general encompassed by the existing schemes, are poorly distinguished within them. So it seems as if our taxonomies are destined to evolve if they are to remain relevant. Yet every puzzlist or taxonomist must strike a balance between useful specificity and cumbersome complexity.

Note that a different and perhaps more productive approach to classifying puzzles would be tag based. For example, a puzzle might have attributes of both assembly as well as pattern. Its assignment to one traditional category or the other would be based on a weighting of all of its attributes as determined by an expert authority or by mass consensus.

Anyway, below you will see how the traditional classification schemes for mechanical puzzles compare with my own. I hope you find this entertaining and useful.