Recently I was wondering why Y has received so much less attention than Hex. In any discussion I've seen about connection games or the elegance of modern abstracts, the name Hex comes up -- why doesn't Y? It seems as though Hex has become a much more widely known classic than Y.
Is Hex so popular because it was discovered by well known people (Hein and Nash) or because it was discovered first (Hex in 1942, Y in 1950)? Craige Schensted often gets the credit for discovering Y, but Martin Gardner says that Clause Shannon, a figure as well known as John Nash, discovered the game some years earlier. I'm not sure who did what to popularize the game. Does anyone here know?
I've heard it argued that "Y must be better than Hex since Y generalizes Hex." This view point seems to overlook game play. A game which is more general than another game might be less fun, less interesting, and less playable than the game it generalizes. One reason to prefer Y over Hex is that it is simpler than Hex whose rules and board are more complicated. This appeal to Occam's razor also forgets game play though.
One thing I like to think about when I learn a game is how that game fits into a pattern of human thought. Each game can tell us a little bit about how humans think and what we're good at. There are some abstract games that simply seem unnatural, and that simply cannot be played seriously. Could it be that Hex is more popular than Y for the simple reason that the play of Hex is a more human one than Y?
I have a particular soft spot in my heart reserved for Hex, the game that initially got me hooked on abstracts, so my views on the issue are certainly biased.
You say, "why hex instead of Y?" I say, "why not?"
Given a group of "equivalent" games, I think it is natural that one would emerge as dominant and the rest would become needless variants - even if the selection is arbitrary.
"One reason to prefer Y over Hex is that it is simpler than Hex whose rules and board are more complicated."
It seems to me Hex has simpler rules and board. Or they're at least equally simple: Be the first to connect two sides of a rectangle, or three sides of a triangle.
I think Y has the simpler board, but not for a two-player competitive game. A rhombus isn't as nice as a square, but the color differentiation of the goals makes the game instantly grokable, to coin a word appropriate to the subject.
Nice to see that Y is being discussed. I think it is a very good game that deserves more attention.
One reason why it can be considered more elegant than Hex is that the edges do not need colours, although I guess that is not a strong argument in itself.
A few months ago I found a page on winning/losing first moves on small Y boards on hexwiki:
Over a few weeks I solved quite a few cases myself - manually. In the process I got a feeling that this was a truly superb game.
On small boards, many opening moves are easily solved with the help of corner templates, which I also developed. Corner templates add to the richness of Y strategy which already includes the edge templates from Hex.
I hope to see Y implemented here on Little Golem some day!
The argument that a generalized game is better is not quite good imo.
All 2 payer finite games with full information can be generalized to a game in a tree where the players share a token and take turns deciding which child to move the token to. Ending in a 1 is a win for p1, ending in a 2 is a win for p2, ending in a 0 is a draw.
This game however, is not very fun to play (i have never met anyone who thinks so).
I prefer Y to Hex by a very large margin; in fact, Y has been my favorite perfect-information game for over two years now.
One thing that has not been mentioned that affects physical game play is that in Y, one can either swap colors or replace the first move with the swap rule (it makes no difference). In Hex, one must either swap colors (which means rotating the board) or replace the piece with a new one in a reflected spot (which can be confusing for some), as the each player has a different goal.
Also, I think that Y is a more fair game than Hex, and here is some data that supports my hypothesis:
Well, for one thing, Hex is a game with a history. Claude Shannon made a machine to play it, Einstein knew it, Bohr and the circle around him knew it (because Piet Hein was part of that circle, and he pushed his game almost as eagerly as some modern inventors). Thousands of Danes knew it from Hein's newspaper columns, and the newspaper got pads printed and for a while it was a big fad in Denmark (during the war, when other entertainment was scarce - abstracts rarely get such breaks!).
Also, Hex has more active players. One of the exciting things about an abstract game to me is to see how far it can be taken, and Hex has currently been taken very far. It seems less interesting to "start anew" in a game that is so similar.
David Milne - Doesn't the length of a game of Hex or Y depend strongly on the size of the board? That is, for certain choices of board sizes Y is much faster than Hex, and for other choices Hex is much faster than Y. Did you have in mind some aspect of the games that is independent of board size? (If so, I would be interested to hear about it in more detail.) Or did you have particular board sizes in mind?
David, according to the statistics on iggamecenter, the average Y game lasts about 3 minutes shorter than Hex. Now, that data does include all board sizes, but from my experience (which I have a lot of on the site), most Y games there are played on size 19 boards (190 cells), and most Hex games on size 11 (121 cells) or 13 (161 cells) boards, making games of Y much shorter than Hex with similar numbers of cells.
Little Golem