The Loom of Games

Rattlesnake and Variations games in context

Inspired by the Glass Bead Game of Hermann Hesse’s novel, for some time now I’ve been designing a connecting game of my own. It’s still a work in progress, and goes through a couple of major revisions every year as a result of feedback from players, and my own changing understanding of what exactly I’m trying to do, and how to do it.

Recent versions involve players choosing a single theme which is then recontextualised and reexpressed in different contexts. There’s a simple pen-and-paper/ game-on-the-beach version called Rattlesnake in which players choose the themes and contexts quite freely, and another version called Variations where the themes and contexts come on a print-and-play pack of cards. The context cards have settled for now on four groups representing knowledge, uncertainty/belief, the sensual body, and emotions. The theme cards change from pack to pack, and so far the decks have featured the innovations of Leonardo da Vinci (2019), themes from the experimental mid-20th century Fluxus art movement (2019), and most recently, in a version called Chuang Tzu’s Art of War (2020), concepts from the philosophy of confrontation from the U.S. Marine Corps Book of Strategy (Warfighting, 1997). Of course, a future planned deck (and long term work in progress) will be based on the theme of glass beads.

But what’s it like to play Rattlesnake and Variations? People always ask me, or are ready to tell me when I start to explain, what’s my game like? Which other games is it similar to? I’m not the only one to see a similarity with many other games, especially those involving comparison and connection, so I’m going to use the Loom of Form and Meaning ( which is a tool for classifying types of connections, to explore the landscape of games which involve comparison and connection. I’ve previously done something similar with game apps involving comparison and connection (, and many of the core challenges used in all these games and apps are also found in verbal reasoning exercises (

In some A1 type games (same form, same meaning) we have to identify a correct answer by analysing its properties and characteristics. This can be done by asking questions and narrowing down options until we find the right answer, as in the popular children’s game Guess Who (Theo and Ora Coster, 1979), or choosing from a fixed set of clues to help other players to guess the correct answer, as in Concept (Gaëtan Beaujannot and Alain Rivollet, 2013). Parlour games such as I Spy, Hangman, 20 Questions, and Animal Vegetable Mineral follow a similar model. Card games based on Snap! (uncredited, 1866) are built on fast recognition of identity between two cards, and represent another type of A1 game. Yet another variety is Kim’s game (Rudyard Kipling, 1901), where players have to memorise and accurately recall several objects briefly displayed.

A2 type games (similar form, same meaning) play on small differences between forms, without changing their meaning. For example, in the self-proclaimed ridiculous game of Speak Out (Kevin Hill (II), 2016), players try to say phrases while wearing a mouthpiece that won’t let them articulate properly.

A3 type games (different form, same meaning) are primarily concerned with reexpression, and not in a limited way using properties and characteristics, but completely reexpressing an idea through different words, means and media. Examples are Taboo (Brian Hersch, 1989), in which players describe a word using neither the word itself nor several other words and phrases, Pictionary (Rob Angel, 1985), where only pictures may be used to describe an idea, Claymania (Richard A. Moore, 1993), which does the same using clay, as well as the popular parlour games Charades and Celebrities.

B1 type games (same form, similar meaning) reapply the same form in related ways. On my first pass at this categorisation, I had a few games classed as B1 which have ended up elsewhere. There are a few B1 type puzzles in Tribond (Ed Muccini et al, 1989), in which the central challenge is to identify the single correct word or idea which three other ideas have in common. For example cars, trees and elephants are linked by etymologically related (according to the OED) uses of the word trunk. However the vast majority of puzzles involve solutions which group similar items into a family group, for example Jupiter, the blue whale and Alaska all being the largest of a kind, making Tribond a B2 type game for the most part.

Other B2 type games (similar form, similar meaning) also involve identifying examples which share certain characteristics. In Knit Wit (Matt Leacock, 2016) there’s no single correct answer, though all acceptable answers will be similar in that they share the required features: something red and round might legitimately be a tomato, an apple, or anything else meeting the criteria.

In Codenames (Vlaada Chvátil, 2015), players give one word clues for their team members to try to correctly identify the images or words on the cards in play. The clues can apply to more than one card at a time: cards with the words ‘Egg, ‘Feather’ and ‘Roast’ may all be identified by the clue ‘Chicken.’ The rules specifically require that clues must relate to the meaning of the word: don’t use ‘B’ as a clue for ‘board’ and ‘block,’ or ‘well’ as a clue for ‘bell’ or ‘sell.’ (These would be C2 type clues.) The clue must be a semantic quality that the things being compared have in common, making them members of a family group sharing the similarity of that feature.

In Happy Families (uncredited, 1851) and Game of Authors (Anne W. Abbott, 1861) players take turns to ask for a named card, attempting to make sets of similar cards, as in Go Fish which is played with a standard card deck. Not wanting to get too sidetracked, but I can’t avoid noting at this point that in the essay Hermann Hesse’s Castalia: Republic of Scholars or Police State? (Osman Durrani, The Modern Language Review, 1982), we learn that ‘in the earliest-known version of the introduction to Das Glasperlenspiel, the Game is invented by one Oberrechnungsrat R. Klaiber, “um seiner Frau das Bridge zu ersetzen,” and it evolves as an increasingly complex form of the card game “happy families” (“Dichter-Quartett”).’ Dichter-Quartett is basically the German version of Game of Authors. G. W. Field (On the Genesis of the Glasperlenspiel, The German Quarterly, 1968) adds intriguingly, drawing from Hesse’s drafts, that Klaiber, ‘however, combined names and works of poets with those of painters, musicians, architects; and instead of “quartets” a “sextet” might occur for Goethe or J. S. Bach, while a “tercet” might do for Lessing or Gluck. A virtue of this game was the fact that one made it oneself. Each interested family developed its own “Bildungsquartett,” which could be specialized or extended and universalized.’ Basically, the idea of the Glass Bead Game started as a collectable card game version of Happy Families!

B3 type comparisons (different form, similar meaning) involve analogies. One way analogies are found in games is by asking questions for individuals to answer purely from their own personal perspective – there’s no right or wrong, just ‘what does this question mean to you?’ The Ungame (Rhea Zakich, 1973) might ask ‘Talk about the person who has been the biggest influence in your life.’ Scruples (Henry Makow, 1984) asks closed questions: ‘You are driving alone on a highway at night. A desperate looking person tries to flag you down. Do you stop?’ The analogy is implicit in each player giving their own answer.

C1 type games (same form, different meaning) involve single images and ideas being imaginatively interpreted in different ways.

In What Do You Meme? (Elie Ballas, 2016) the same photo is captioned differently by players, and a judge decides the ‘funniest.’

C2 type games (similar form, different meaning) also require creative thought to find surprisingly different meanings within certain constraints.

Scattergories (uncredited, 1988) is a commercialsation of the old parlour game of Categories, or Guggenheim, where players list words from predefined categories, all of which must begin with a given letter. Facts in Five (Richard Onanian, 1964) is similar, with more esoteric categories. There is no semantic similarity between the words, only lexical.

C3 type games (different form, different meaning) feature apparently unrelated words and images which must be combined or reconciled through juxtaposition, while leaving their difference intact.

In VisualEyes (Keith Dugald (I) and Steve Pickering, 2003), players have to creatively ‘read’ pairs of images on dice at face value to make everyday words and phrases. I originally classified this as a B1 game, but was persuaded to go for C1 because the images may be interpreted in different ways which are completely unrelated to each other, a pairing of a snowflake and a jacket may equally legitimately become ‘winter coat’ or ‘dandruff collar,’ and then C3 because the snowflake may instead be paired with a clock making a quite different pair to become ‘cold hands’.

In Rory’s Story Cubes (Rory O’Connor, 2005), the nine dice of each set have images which together must be combined into a story. I asked on a BoardGameGeek thread about how liberally these images were interpreted by players, noting: ‘A bee can be, well, a bee, or “be quiet,” or “let it be.” A question mark can be any question (“she asked”) or reversed to make a hook. A dice can represent chance, or “no dice!” as in “no way!” A parachute is “help came” or “get me out of here.”’ The designer himself replied on the thread to say that creative interpretation was ‘truly playing to the spirit of Rory’s Story Cubes. When I picked which icons should appear, those that held multiple interpretations always trumped those that had only one.’

I tentatively also include Tarot here: cards which each have established, esoteric and personal symbolism and significance are randomly dealt and interpreted in the context of a particular question, to hopefully get a coherent response. It’s not too different from Rory’s Story Cubes in essence.

The surrealist game Exquisite Corpse, based on the parlour game of Consequences, where different players add to a drawing or text with minimal knowledge of what other players have added before, preserves the difference between the unrelated elements until the final stage when the final composite work is revealed, making it a good example of a C3 type game.

Aside from all the above, which mostly exploit one type of comparison as the core of the game, Huggermugger (Diana Carlston, 1989) and Cranium (Whit Alexander and Richard Tait, 1998) are composite games of games in which one of several types of challenge are included in the game.

There are also other games explicitly inspired by the Glass Bead Game that haven’t been released commercially, as far as I know, and which play in the same broad space of connections between ideas. The very best I know about are these:

  • Charles Cameron’s Hipbone Games (1995) link individual ideas according to a pre-established pattern, without any pre-established theme, though one may often emerge through play. All types of links may be used freely.
  • Charles’ later DoubleQuote game (date unknown) is about as pure and deceptively simple as mind games can get – just compare and contrast any two ideas. Beautiful!
  • The Glass Plate Game (Dunbar Aitkins, 1984) builds chains of thought using visual and textual cues. The suggested types of connection have most types of comparison explicitly covered, and lack only dual meanings of words and dual uses of objects (B1 if common/related, C1 if uncommon/unrelated) which the ‘sample responses’ also don’t really accommodate. A landmark.

In the current version of the Rattlesnake and Variations games, any kind of link may be used to connect a theme to a context. For example the theme ‘apple’:

  • can connect to religion as a symbol of the abstract ideas of Temptation, and Knowledge of Good and Evil (A3 type link)
  • connects to science through Newton’s analogy of the apple and the moon under gravity (B3 type link)
  • connects to the emotion of surprise through Karlheinz Stockhausen’s idea of an apple found ‘on the moon’ as the epitome of absurd recontextualisation (a C1 type link)

Studying other comparison games has led to some interesting questions about my game design. For example, should it be working in reverse with players guessing the theme from others’ connections (like Concept and Codenames), rather than expanding on a known theme? I tried it, but actually, I hope, it is exactly the fecund open-endedness of the variations which is my game’s hallmark.

I’ve also raided other types of games for ideas. I wanted to give credit to players who showed more breadth in the range of contexts they linked to, so I adapted the endgame scoring mechanism used for green scientific structures cards in 7 Wonders (Antoine Bauza, 2010) and for verticals in Azul (Michael Kiesling, 2017). Wanting players to be able to connect to contexts they are comfortable with, I introduced an initial deck building stage like Sushi Go (Phil Walker-Harding, 2013), and not wanting to make it too easy I balanced it with players having the option to take cards from other players, as in Go Fish. I also owe Dr Geoff Crook a debt of thanks for suggesting that the scope of the connections be expanded beyond knowledge and intellectualism, and into the personal feelings, uncertainties and sensual physicality explored in games like The Ungame.

But why specifically play connection games at all? For the exhilaration of moving through mental space, navigating a disorienting cognitive environment, and to make your thinking more expansive, afterwards (after J.B. Jackson, Discovering the Vernacular Landscape, 1984).

Games are now used a lot in education, often in particular subject specialisms in ways that are less concerned with cross-curricular learning and types of reasoning, than superimposing different subject matter content onto them. As I hope I’ve started to demonstrate, games can also be used for the type of learning they offer inherently through their mechanism: Snap! for pattern recognition and fast reading, Charades or Pictionary for non-verbal expression, Taboo for vocabulary, Guess Who and Concept for deductive reasoning, Knit Wit for set theory. My own recent games take a single concept, and look at it through different lenses, a bit like exploring a key concept across different curriculum areas, and Rattlesnake in particular can be used really well with kids for brainstorming or consolidation. More generally, an understanding of connections is required because interactions have practical applications (Scott Ashman and Amanda Nelson, Why do I crave that cookie?, Science Teacher, 2012).

A recent game I bought (Sushi Go, published by Gamewright) helpfully observed on the back of the box: “Kids Learn: Strategy. Probability.” What do we learn from the other games we play? I believe we can analyse games much more deeply to see what specific purposes they they can serve in our learning and development.

Of course, they should just be good fun too.

Glass Bead Game designer