A little question about Unlur

[luigi]

Hi all,

Yesterday, when I was skimming through Cameron Browne's impressive
book "Connection Games: Variations on a Theme", I came across the
Unlur rules, which I thought I knew well, and was amazed to (re)
discover that a player loses if he makes the opponent's winning
connection with his own pieces. As Browne notes, this implies that
some moves by either side may actually damage that side's position,
which looks a bit strange in connection games. Then I wondered why the
rules doesn't avoid this by simply stating: "Black wins if he makes a
Y which touches three non-adjacent sides of the board and loses if he
can't possibly do it". Wouldn't it be more elegant, and thus
preferable? Is there any obvious fault with this that I can't see?

[Nick Bentley]

I think you are right. I've been designing games featuring the "one
player tries to build it, and one tries to block it" for a long time,
and there's no problem with this kind of game. It can be applied to
just about any pattern. here's a recent one:

http://nickbentley.posterous.com/new-game-coil

This one is just like unlur except one player is trying to make a loop
in a certain color, and the other is trying to block it.

Because this scheme can be applied to just about any pattern, it opens
up a whole universe of games that aren't really available any other
way.

I designed a meta game, called Mind Ninja, that takes advantage of the
fact that the scheme can be applied to any pattern.

It goes like this:

1. One person specifies a pattern (which in this case can be composed
of stones of either or both colors, as well as empty spaces.)
2. The same player places any number/combination of stones on the
board that he wishes (a super-generalized pie rule).
3. The other player decides whether to try to build the pattern or
block.
4. The players take turns placing a single stone of *any* color onto
the board, until either the pattern is build or the board is full.
5. If the pattern is built, the builder wins, and the blocker wins
otherwise.

There's a slightly more elaborate version where the players first bid
either stones or clock time for the right to specify the pattern in
step 1 above, to eliminate any possible knowledge asymmetry.

[camb...@googlemail.com]

Hi Luis,

In response to your original question:

> I wondered why the rules doesn't avoid this by simply stating:
> "Black wins if he makes a Y which touches three non-adjacent
> sides of the board and loses if he can't possibly do it".
> Wouldn't it be more elegant, and thus preferable?
> Is there any obvious fault with this that I can't see?

An argument against this might be that the rules as stated are more direct. The result is the same in each case, but this may not be immediately obvious to new players. The additional level of indirection caused by implying a win for White as a non-win for Black seems just that little bit less clear, and connection games are generally typified by their clarity.

Also, condensing the two opposed winning conditions into one might be more compact, but could be seen to obfuscate things a tiny bit.

Though these are minor points. Nice suggestion!

Regards,
Cameron

[Luis Bolaños Mures]

Hi, Cameron,

Thanks for your answer. There are three similar rule sets to consider here:

a) Black wins by forming a Y. White wins by preventing Black from doing so. This was my original suggestion. This is not equivalent to Unlur, since White can prevent Black from forming a Y by forming a Y himself, which is then a win for him.

b) Black wins by forming a Y. White wins by forming a line. This is a draw susceptible rule: by forming a Y, White prevents Black from winning but doesn't guarantee a win for himself, since Black can still prevent him from forming a line.

c) Black wins when either player forms a Y. White wins when either player forms a line. This is the Unlur rule.

I agree that the indirect White goal in my suggestion is slightly less clear to new players than the Unlur goal, but this is balanced out by the property that your stones can never harm you and the fact that it's a more compact goal, as you point out.

Then again, the cold aspect of some Unlur positions can also be seen as an extra level of indirection, but a positive one. The critical point in the game is the color choosing at the end of the contract phase, and knowing that your stones can harm you adds to the subtlety involved in that decision.

Finally, I have two questions:

a) Are cold wars (not coldness in general, of course) possible in Unlur? After some testing I'm almost convinced that they aren't, but you probably have a definitive answer.

b) I'm not sure what you mean by "the result is the same in each case", since you know my suggestion is not equivalent to Unlur. Could you clarify?

Regards,

Luis

[Luis Bolaños Mures]

On Monday, December 17, 2012 12:08:36 AM UTC+1, Luis Bolaños Mures wrote:
> c) Black wins when either player forms a Y. White wins when either player forms a line. This is the Unlur rule.

In case there's anyone reading this who doesn't know Unlur, I forgot to mention that when a player forms both a line and a Y, he wins.

Also, Cameron, this rule is the reason why I think cold wars are not possible. Am I right?