Re: Suggestions for games?

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

Date: Sun, 20 Nov 1994 20:34:09 +0000

Greg Lindahl <gl8f@fermi.clas.Virginia.EDU> writes:
> What do other folks think about a forecaster? Is it possible for the
> games that you play, and do you make a significant number of errors
> that it would catch?

The Phoenix forecaster is exceptional.  I guess I had my doubts too
until using it for a while.  I don't think I could run big positions
(75+ worlds last game) nearly so well without it.  It just means that
you can write orders and not have to go over them with a microscope,
since the odds are high that if you write something wrong it will
generate a warning of some sort.  Of course it doesn't find errors
where you are just doing something stupid (at least, I don't use it for
that); it's just a way of tracking mechanical errors.

Of course the forecast isn't identical to what will happen when the turn
is actually run, because other players will do things, you may be using
information which isn't available to the forecaster, and so on.  That's
not really the point though.  The point is that if you miswrite an order
(not just syntactically, but, oh, I don't know, transfer from F1 to F2
instead of F2 to F1), you can catch it.

Unfortunately I think it's spoiled me and I may be making more
mechanical errors in other games which don't have forecasters as a
result.  You get out of the habit of carefully proofing your orders line
by line.  Which after all is something I'd rather have computers doing
than people.

I think a true forecaster for Olympia, for example, would be extremely
useful.  I also think that it's a lot of work, particularly if you
haven't designed the game program with forecasting in mind from the
beginning.

And I have some pretty vivid memories from the You Rule! playtest of the
value of a forecaster in that game!  (I miswrote a move order which
proved extremely damaging....)  In fact I would say that almost every
pbm game I've played, from simple to complicated, would benefit a lot
from a forecaster.  Starweb and QGJ are two more that come to mind.  In
every case, the main advantage is catching miswritten movement orders,
transfers, etc.  Things that aren't syntactically incorrect but clearly
make no sense when interpreted.

David desJardins
--
Copyright 1994 David desJardins.  Unlimited permission is granted to quote
from this posting for non-commercial use as long as attribution is given.

References

Suggestions for games?
minnotte@stat0.stat.rice.edu (Michael Minnotte)
Tue, 15 Nov 1994 21:36:39 +0000
Re: Suggestions for games?
griffith@netcom.com (Jim Griffith)
Fri, 18 Nov 1994 02:27:49 +0000
Re: Suggestions for games?
gl8f@fermi.clas.Virginia.EDU (Greg Lindahl)
Sat, 19 Nov 1994 21:08:40 +0000