Re: PBM design and formulae - a summary

From: desj@ccr-p.ida.org (David desJardins)

Date: Fri, 02 Jul 1993 22:48:52 +0000

Graham Matthews <gym@dcs.ed.ac.uk> writes:
> This is quite obviously true. The phenomenon you describe is called
> experience! How many times does it have to be said that I am not
> against experience - there is nothing that can be done about that.
> What I am against is players initially learning game information by
> reading code for example.

I understand this precisely.  I am for exactly the reverse.  (Although I
believe that the information should be documented in the rules, rather
than buried in the source code.)  If the information is in the rules,
then it is equally accessible to every player.  If the information is
only acquired through experience, then certain players (those who have
been playing a long time, or friends of those who have) have a large
inherent advantage.

The issue of mathematical or strategic ability is really a red herring.
Players with skills are going to gain an advantage in any sort of game
where those skills are applicable.  If the game rewards detailed
computations, it is going to reward them regardless of whether the
formulas are published or unpublished.  Really, the only question is
whether one gives an _additional_ advantage to those who have been
playing for a long time, or not.  I vote not.

I do understand that there are players who _prefer_ games in which
experience equals success.  In my view, these are not so much strategy
games, as games of persistence and (frequently) of sucking up to the
players who have the information.  If people prefer such games, I don't
have any problem with them playing them.  But I won't.

> If you are implying that I was suggesting that people should play
> chess without the rules then once again you misrepresent my
> position. Obviously people should have the rules when they play
> chess. But what they should not have for example is a database
> of every game thats ever been played before that they can consult
> to play (this is the equivalent of knowing the *exact* formulas in a
> strategy game).

I daresay that Graham is the only one reading this group who thinks
this.  I think the equivalent of knowing the exact rules to one game is
knowing the exact rules to another game.  He thinks that the equivalent
of knowing the exact rules to one game is having a list of every move
that has ever been made in another game.  I leave it to the readers to
judge which two things are actually equivalent.

Graham on the one hand seems to think that I cannot, from observing what
happens in several turns of a game, figure out the algorithms by which
the game operates.  On the other hand, he seems to think that if he
gives me those algorithms, I will suddenly become a superhuman computer
which deduces the best move in every possible situation.

I think that he has the difficulty of these two deductions competely
reversed.  It is relatively easy, by looking at how well medium infantry
does against light infantry in various combat situations, to figure out
the relative advantages and disadvantages of the two kinds of units.  I
may not be able to figure out whether the attack strength of medium
infantry is 2.3 times that of light infantry, or only 2.2 times.  But
the reason I can't figure it out is that it doesn't matter very much.
(If it does matter, say because there is a threshold effect, then I will
be able to use that same threshold effect to figure out the numbers
involved.  However I think we all agree that games should generally be
designed without such effects.)

On the other hand, it is impossibly difficult to "solve" the problem of
always knowing what to build and where to move it, especially in a game
where your opponents are simultaneously building and moving and you
don't know where.  That's why people play games, because they are
challenging.  Games don't radically change in nature from a minute
analysis of the formulas; or, if they do, they are horribly broken in
the first place.

However, I can't refute Graham's model.  If it were really true that it
is very hard to discover formulas, but very easy to play the game with
superhuman ability once the formulas are known, then it might well be
correct to hide the formulas.  However these premises are in my opinion
completely wrong.  Since there's no way to prove or disprove these
premises---at least that I can see---I think we have to leave it at that.

David desJardins

References

Re: PBM design and formulae - a summary
gym@dcs.ed.ac.uk (Graham Matthews)
Thu, 01 Jul 1993 11:15:09 +0000
Re: PBM design and formulae - a summary
desj@ccr-p.ida.org (David desJardins)
Thu, 01 Jul 1993 17:38:04 +0000
Re: PBM design and formulae - a summary
gym@dcs.ed.ac.uk (Graham Matthews)
Fri, 02 Jul 1993 11:37:19 +0000