The main grid is always 4×4 and has the same 8 row/column headings, giving 16 choices.

So 16 possible subgrids stem from the main grid. The two examples shown above differ on how many rows/columns exist in the subgrids (i.e. the subgrid from naked/physical had 16 choices while the subgrid from machine/physical only had 12 choices).

Also from reading the text above, after the subgrid choice is made, the final 3×3 grid has its rows and columns formed by having the contestants choose variants from computer provided lists one after another.

Then one contestant selects a choice from the row and the other selects a choice from the columns. That is the “game” they compete at.

If you picked reasonably finite values for the length of the final grid lists (say 9 variants each) and required all the subgrids to only be 3×3, then you would need to intelligently select 1400 category headings and 1296 final games could be derived from combinations of them. Of course the first 8 category headings are already selected by Mr. Anthony

4+4=8 categories for the first grid (already defined by the text above), yielding 16 STAGE 1 choices
16 * (3+3) = 96 categories for all the subgrids (assume each one was capped at 3×3), yielding 16*3*3 (144) STAGE 2 choices
144*9 = 1296 categories in the final grid list, though only 144*(3+3)= (864) could actually be placed in 144 3×3 matrices as the row/column headings during STAGE 3. In STAGE 4 that choosing from that final matrix would yield 144 * 3 * 3 = 1296 different games total.

Of course, if you wanted more possible games, you could have variable row/column matrices during STAGE 2 and more than 9 variants in the STAGE 3 list used to form the final STAGE 4 3×3 matrix.

This would indeed be an interesting concept for an MMO, though the sizes of the matrices would probably have to be increased considerably.

The 1949 translation by Mervyn Savill is said by Those Who Know to be rather stiff and formal in comparison.

One interesting way that The Apprentice Adept dates itself is that while some of Anthony’s games are facilitated by “panels”, there are no video games in The Game.

Btw, Xanth is a different series; Split Infinity is book one of Apprentice Adept series.

Also, sorry folks, I made an error in my original post: while I bought virtually all my Piers Anthony novels at the mall at B. Dalton in Raleigh, I just remembered that I got Split Infinity one evening on a summer vacation to Cape May, NJ, in an outdoor plaza of shops. Shortly after the bookstore, that same night, one of my younger brothers, who was about 7 at the time, somehow drifted off into the crowd, and went missing for about 15 panicked minutes. We found him, a bit shaken, but alright, anxiously wandering the maze of shops. I remember clutching my copy of the book as we searched, and my mother’s grim expression; I hadn’t experienced that kind of creeping fear before, until that night. I haven’t been back to Cape May since.

You’re probably right that Anthony didn’t necessarily imagine The Game as generated Holodeck-style; for example, the Dust Slide and antique pistols that appear above may have been part of an elaborate game arena, that has all combinations of games “unrolled”. The antique pistols were presumably placed on a table by an assistant; the Dust Slide was pre-constructed. If some other players happened to have narrowed down their game to the Dust Slide and were currently using it, you’d have to wait your turn.

But yes, like Josh suggests, one could imagine this being truly generative. For example the specific options in the third round of choices could vary: instead of snowbanks and limestone cliffs, we could imagine mudslides and jungle trees. Even the some of the more abstract characteristics could potentially vary; this would exponentially increase the number of game combinations.

Could a generative game system create a well-balanced, well-designed game? In my blue-sky imagining of such a game generation system, of course!

What I find really exciting about this vision is how it combines generativity with competition. Most research into generative stories/games have enough trouble getting the generativity itself to work, leaving little thought to how actually apply the generativity to entertainment ends. Here the process of generating a game, the meta-game, is part of the competition itself. Great design.

(The whole Proton/Phaze universe is ripe IP for an excellent game/MMO, IMO. I wonder if anyone’s secured the game rights yet.)

Also, a generative system like this would to some extent be a collaboration between human players and an AI, to create a new game.

Noah, thanks for the link to Balsamo’s essay; it didn’t come up in my search because “Anthony” isn’t mentioned in the text (as you say, she’s informally riffing on several sci-fi ideas, not doing an analysis per se, references not strictly required.) Chris, nope I haven’t read the Perec book, will do.

I wonder, though, could a truly generative system work well in a competitive environment like the one this story describes? Usually designing a game to be both fair and challenging takes significant time to balance and refine. A generative system to create new games would probably be easier to apply in an experimental, co-operative environment where people are looking for new experiences and challenges rather than trying to compete for recognition.

Is this, do you think, a generative system, or a taxonomic system? Are games created to fill in the blanks, or do categories get created simply because games can fit into them?

This is a tangent, but: I don’t know if you’ve read Georges Perec’s “W, or the Memory of Childhood”, which has another take on a society dominated by games (and is divided into two different but interrelated worlds, hm). It would be interesting to compare the two.