[haggis] Values of rules

[Chris Nash <chris_nash@prodigy.net>]

--- Eric Shultz <swift_4@yahoo.com> wrote:
> What I learned from game 92 is that when 8 out of 10
> rules are values, the game goes sour REAL fast.

The problem is establishing what a rule is actually worth. A value rule is worth almost nothing,
because one value rule per card type is too many. It seems to only take about half of the value
rules to narrow down the card values to a handful of possibilities. Any further value rules don't
help much, and offering value rules won't buy you anything.

Owning the set rule doesn't buy much, either - everyone assumes its there and assumes its good,
which makes everyone's card trades very single-minded. The set rule would be a lot more important
to know if the value of a set had a 50% chance of being negative... The presence of a set rule
devalues the value rules as well (it doesn't matter what the individual card values are, since
you'll be getting sets of them anyway).

The penalty rule was worth knowing... but once a rule is deemed valuable, you're unlikely to be
giving it away. For me in this game, my optimal strategy would have been to submit as soon as the
game started...

> I still think there's hope for this game. If anyone
> wants to break this sucker into rounds and turns, let
> me know.

I think there's hope - it's just a question of finding a sensible way of doing it. The set rule
got me thinking a lot about rummy games, where each player's turn (draw/meld/discard) does deliver
a small amount of global information. There needs to be a way of slowly improving the hand, which
isn't possible when everyone has the same goals, or you don't know your goals. I liked Lyman's
idea of blind rules, exchanging cards to see how your hand's value changed, but it seems that,
plus turn- or round-based play, means we end up with a thoroughly different game.