From: apoulter@nyx.cs.du.edu (Alan  Poulter)
Date: Sun, 17 Jul 94 23:44:47 PDT
To: cypherpunks@toad.com
Subject: Board games
Message-ID: <9407180644.AA00319@nyx.cs.du.edu>
MIME-Version: 1.0
Content-Type: text/plain




Timothy May writes:-
>An obvious problem with crypto card games is this: what does it
>provide that is worth the extra effort of doing encryption?

Quite so. What other sorts of games are there that could benefit from
crypto?

There are a large number of multiplayer board games which combine high
levels of complexity in move options with the necessity for inter-player
diplomacy. Such board games are suitable for play-by-email (PBEM), as
players need time to negotiate and work out moves. They generally work
by requiring all players to submit orders for their move by a deadline.
Problems with PBEM of such games revolve around having to persuade one
person to sit out and adjudicate player moves (the luckless 'games
master' or GM). There is a die roll server (send 'help' to
dice@danpost4.uni-c.dk for details) which can provide die rolls for game
functions but a person is still needed to request the appropriate die
rolls to resolve player orders.

The solution here is crypto. You can do away with the need for an extra
person as GM by having one of the players act as GM (the 'player-GM')
and resolve moves for each turn. Since all moves are due by one deadline
there is the risk that the player-GM can move in response to other
player's moves which have been sent to the player-GM for resolution. To
prevent this the player-GM must make their move before all the other
players. If done using plain text this puts the player-GM at a
disadvantage so their orders are encrypted before emailing to all other
players. These encrypted orders cannot then be changed in response to
other player orders.

During move resolution all player orders are published. This benefits
all players in that they can check that the player-GM has resolved their
orders correctly (when using a non-player GM order resolution is
typically hidden). These published orders would include the decrypt pass
phrase for the player-GM's orders, and the plain text of those orders.
It would be incumbent on one or more of the players to be able to
decrypt the player-GM's orders using the decrypt pass phrase, just to
check that when decrypted they match the plain text version.

Player-GM offers other advantages as well. For example, players can swap
the onerous role of player-GM to allow for holidays etc. Using a
non-player GM, if that person drops out then the game dies. Player-GM
also allows short deadlines, each successive deadline being handled by a
different player-GM, which speeds up the game and spreads the GMing
load.

Another use of a non-player GM is to hold secret agreements between
players. Using player-GM this is no longer possible, but public/private
key-based encryption can be used to allow players to sign secret
treaties with each other. There is a specific forfeit in many game rules
for breaching a signed treaty, so the need exists to make secret deals
that can later be verified by players not in on the secret deal. Once
public keys have been exchanged by players then secret treaties are no
problem.

Persuading gamers to use crypto to play multiplayer board games PBEM
does not take much doing. If no one wants to be a non-player GM (and few
do) then without crypto there is no game. I am currently playing in a
7-player PBEM board game called 'Empires of the Middle Ages' using the
player-GM method. We use PGP as it supports the encryption facilities
needed, runs on the various platforms different players use and is free.
However, player-GM and crypto have yet to make an impact in the PBEM
gamer community. This community is small and the ownership of particular
games is limited. The game I am playing player-GM, 'Empires of the
Middle Ages', although an excellent game, has been out of print nearly
15 years. I will be trying to start other board games using crypto and
player-GM in the near future. Watch rec.games.board and rec.games.pbm
for announcements.

I should mention that player-GM is not the only solution to the GM
problem. Another answer is to code a email-driven program which will
adjudicate orders. Here the problem is the inherent complexity of many
board games. The most successful adjudicator program to my knowledge is
the Diplomacy judge (send 'help' to judge@morrolan.eff.org for details).
Diplomacy is a relatively simple board game set loosely in Europe prior
to the Great War. As the game name suggests, players can only win the
game by good diplomacy, as military skills are not enough on their own.
Many thousands of people play PBEM Diplomacy and its variants (games
which change certain rules and/or map features from the standard game).
Although the Diplomacy judge allows anonymous opponents and faked email
in certain Diplomacy variants, it has no built-in facilities for use of
public keys, which seems to me a strange omission. More information on
the Diplomacy Judge can be found in the FAQ for rec.games.diplomacy.

Alan Poulter (apoulter@nyx.cs.du.edu/a.poulter@lut.ac.uk)