I'm told the golden wasp game (or guêpe d'or) is notoriously difficult to understand though, allegedly, quite easy to play. Easy to play badly, that is. In complexity it is said to fall somewhere between chess and the glass bead game (which, we should remember, is a fabricated game, deliberately and tantalisingly incompletely described by author Hermann Hesse). Perhaps go or mahjong would be better comparisons, though as I find draughts a bit of a challenge I'm not in any position to say.
I presume the orientation of the inlaid gold wasps has something to do with the rules of play, and the hexagonal playing pieces suggest a need to fit them together, perhaps along the lines of the cells of two competing wasp nests. That these pieces bear six coloured dots probably implies some sort of suit-matching is involved. I have no idea how many pieces are actually needed for a game, or what variations there are in the placement of the coloured dots, as my set is almost certainly incomplete. They are also reversible, with a single, generally different symbol on the other side. Perhaps each piece can be transformed into a more powerful piece as part of the evolving game-play, or maybe it's just that there are two, completely different games that can be played on the same board? One, like draughts, for the simple-minded, and another, like chess, for real game-players.
It is a very pretty board, though.
10 comments:
The Golden Wasp Game, a zero-sum, complete information game, was solved by Atari in 1997, a fact that was never reported at the time as everyone was still talking about Kasparov v Big Blue and anyway, very few people outside fiction had ever even heard of it.
Zouk,
"A zero-sum, complete information game" ... Mike nods wisely, but says nothing. As I say, without exaggeration, draughts is my limit.
Mike
To people inside fiction, of course, the game is not only well known but quite popular. While we always see them playing bridge or checkers, this is because fictional people think that were rather boorish and so put on a bit of a pretense for us. Behind the spine of the book, between the pages, and during station identification breaks it's all Guêpe d'Or.
True. I've heard entire paper fortunes have been made or lost on a "treble sting" gambit, waiting around between chapters.
Mike
Basically:
zero sum - one player's gain is always matched by an equivalent loss to the opponent(s?)
complete information - all information about the situation at any point is available to both (all?) players
Ironically, Atari -- whose name is taken from a go term meaning "threat of imminent death", the founder being an aficionado of that game -- never cracked go, while the games created by Atari can now be played by Deepmind (the system which did) better than any human. Furthermore, it learnt them simply by watching and trying them.
Deepmind founder Dennis Hassabis explains at the Royal Society:
http://youtube.com/watch?v=08Cl7ii6v
Zouk,
O, that zero-sum complete information game!
Didn't I hear on the news this week that a computer had finally beaten a go master?
Mike
Dunno, must have missed that one. Btw, I issued a challenge to Hassabis on the question of whether Deepmind could learn to make amusing and apposite comments on your blog (thereby releasing me to spend more time playing Guepe D'Or). He accepted immediately, inviting me to his offices the same day -- by which time Deepmind, having had your entire archive fed to it, had already learnt to identify interesting configurations of orange safety netting and could distinguish corvids from other bird species with 99.99% accuracy. I appreciate these matters may be of little interest to your typical reader, but I'm sure some of your former professional clients may appreciate them -- especially the team working on the Ultimate Intelligent Death Machine in the labs by the old allotments and those on project Idle Speculator, the purpose of which is shrouded in a dark cloak of secrecy.
Zouk,
Here you go:
http://www.nature.com/news/google-ai-algorithm-masters-ancient-game-of-go-1.19234
At this rate, we'll all be redundant, though the point of achieving this remains obscure to me. I like having things to do, especially as I'm no good at the golden wasp game.
Mike
https://storage.googleapis.com/deepmind-data/assets/papers/deepmind-mastering-go.pdf
Well, if I'd know it was that simple...
Mike
Post a Comment