Re: PBM design and formulae - a summary

From: gym@dcs.ed.ac.uk (Graham Matthews)

Date: Sat, 03 Jul 1993 16:31:08 +0000

gym@dcs.ed.ac.uk (Graham Matthews) writes:
> [GYM:]
> : > ... Obviously [knowing formulas] does [give a serious advantage].
> : >does. I just said that I think it screws the game up ...
> [RW:]
> : >You haven't answered the point that people having played long enough,
> : >will get some idea of the formulas, by getting killed off a few times
> : >if in no other way.
> :
> : That is inevitable - its called experience.
>
Hel, Queen of LogeSpawn writes
> Much of the satisfaction of playing a game for yours truly consists of looking
> at a position without having needing to have ever played the game once, and
> figuring out the odds-on play of the situation with a fair degree of assurance
> that I can do it quite a bit better than most people, even one who had played
> many times, can.  I know that when I first started in Galaxy I spent three
> entire days thinking about it.  After that despite some slip-ups in sending
> orders and in diploming, I did fairly well and often with a serious time
> constraint.   For the most part I even enjoyed the tactical and economical
> development of each game unwind in its unique way.
>
> And I had no experience.  I don't want to have to play a game in which I
> will lose to players who are no smarter but simply got wiped out more times
> earlier.  Zwagonian, now leading in RW's Game 7, was a newbie when he started.
> He played by the seat of his pants and did just fine.  That is the kind of
> game that I relish and you presumably don't.

Oh come on use some imagination! If a game has good clear relationships between
things then players can play well. For example of an obviously clear relationship
take FlatTop: in this game the weather effects take of from carriers. Now this is
pretty damn obvious don't you think. I only need to know the rules and have some
commonsense to work that out. I don't need to know any exact equations (eg. like
the probability of takeoff failure ...).

So if a game has clear relationships between things inexperienced players can
still do fine. Again you seem to confuse rules and equations.

gym@dcs.ed.ac.uk (Graham Matthews) writes:
> : even that is not entirely satisfactory. The best way to avoid having
> : experience as a problem is to rely on the players to either :-
> :
> : a) play only with players of similar experience
> : b) help newbie players out alot by sharing strategy etc
> :
> : I must be frank and say that the basic problem here seems to be
> : one of winning. People who game to win are never going to do either
> : a) or b) and more's the pity to them. People who game for the
> : enjoyment or for the competition will do both a) and b) as it makes
> : for a better game.
Hel, Queen of LogeSpawn writes
> Give me a SERIOUS Break!!
>
> Do you know what you are saying?  There are people who game to win, yet
> enjoy the competition and the enjoyment.  My code of ethics is that you
> should _ALWAYS_ strive to win in any situation by the best of your ability
> at the given conditions.  That recently lead to a heart-rending quarrel
> with a net.friend (and resulted in some measure of growing up :-) when
> our ideas of striving for victory differed.  Yet I have always looked for
> better opponents, brought back in worthy (IMNSHO) foes for what I thought
> was good replacement positions, taught new players, shared my scripts with
> people who I regard as sporting, and discussed the intricacies of strategy
> and tactics with everyone who cares to talk including people whom I may
> have been or am engaged in a war to the death at the time.  I am fairly
> convinced that desJ and several others are such players.

Sure I agree that there are people who game to win *and* enjoy the competition.
My wording was a little loose - my fault.

Graham Matthews:
> : >>b) spend hours maximising formulas, adjusting tiny details etc.
> : >>My complaint with b) is that it turns the game into a "who has the
> : >>most free time at present"
Hel, Queen of LogeSpawn writes
> First of all, I see no reason why "who has most free time at present" is
> in any way inferior to "who can play the most number of games" (and gain
> experience to use in others!) and they end up being close anyway!  Secondly
> it is not quite true that whoever is using more time tend to do better,
> since yours truly played three galaxy games spending perhaps half an hour on
> writing orders each turn during severe time pressure (i.e.  I was getting
> about very little free time, incl. sleep time) and despite mistakes such
> is inevitable in these circumstances, I did JUST FINE.  Again, David desJardins
> and Sjoerd Schreuder seem to be both busy people who at times will have even
> less time than I do.  They manage just fine.   Whoever is the better
> strategiest does better in almost all intelligent strategy games....

Now I ask you how many games other than Galaxy have you played? *MY* experience
of various games is that the people who spend the longest tend to win. Your
experience may be different.

Graham Matthews:
> : >>tiny details is (in most games) very unrealistic.
>
Hel, Queen of LogeSpawn writes
> Tell me how realistic a galactic wargame involving faster-than-light travel
> is?  Or a game with half-human wizards and elven bards??  Realistic is
> just what conforms to _YOUR_ fantasy view (as opposed to mine or someone
> else's) of the setting of the game in question.

I was waiting for this to come up. Indeed realism can mean an accurate model of
some person's fantasy world. So the realism in this case is not "reality". But
there are still in all simulations a common thread of realism - things like
consistency, predicability, etc - unless clearly stated in the rules.

Hel, Queen of LogeSpawn writes
>> And formulas are part of rules, no matter what anyone says.  Not giving
> a formula is having a hidden rule, because rules by definition is what
> drives the game and if a vital formula is unknown it amounts to having
> a rule missing from the book.

Formulas are part of the rules sure. But not telling someone the precise formula
is not hiding rules. As long as relationships between things are clear then thats
all you need.

[RW:]
> : >Ah, you would forbid the players from issuing more than vague
> : >generalities for orders?  Sounds fun.
[GYM:]
> : No it sounds like realism! On practically any level of simulation
> : nothing should be specified precisely - thats called realism! Even
> : at say a squad level you cannot order your squad to do everything
> : precisely (eg: 5 footsteps left, 2 forward ..). The trick for a
> : designer is to get things precise but not too precise.
Hel, Queen of LogeSpawn writes
> WAIT, WAIT!!  The problem with your little scenario is, if the order in
> which one of my squad walk actually makes a difference, OF COURSE you
> need to be able to control it, or at least have a good enough approximation
> to its behaviour!  [Suppose I tell Sparky to run up this passage.  There will
> be a 10% chance that he will slip on the ground....]  I can even stomach
> hidden information (like Galaxy) in which you know how much of a chance
> Sparky is likely to step on slippery floor but not exactly _whether_ this
> particular stretch of floor is slippery.  But not if I had to play blind,
> because I wold then be at a disadvantage to people who had run into that
> particular scenario before.

The key sentence here is "if the order ... control it .. approximation".
You are in fact agreeing with me that a good simulation contains imprecision,
otherwise why did you use this word "approximation". Well thats all I have
been saying so you are agreeing.

Hel, Queen of LogeSpawn writes
> In Galaxy it looks very plausible to built 1-1-1-1-0's, and it is not
> the worst design in the world.  Of course, you'd be much better off with
> pure drones or shielded drones along with a machine gun.  So you lose.
> Would I feel cheated if there is NOTHING whatever that can tell me
> ahead of time that it is wrong to build 1-1-1-1-0 aside from asking
> someone who had played before or getting wiped out a couple of times??
> I know _I_ would.  As it is, if I can't figure out from the rules that
> I should not build those small fighters except in very specific cases,
> then I deserve to get wiped out.

This seems to be something of a fundamental example. Now I have no problems
with you learning by experience that these 1-1-1-1-0 fighters such. ie. you
try them in a few games and they are useless. That is after all realistic - that
is how the character you represent in the simulated reality would have to
learn too. What I object to is you reading the code/equations to find this out.
What the hell does that simulate? Does anyone understand this point - I don't
mean necessarily agree with it, but actually understand the point?

Hel, Queen of LogeSpawn writes
> I have stayed more or less silent because David and Bruno and Russell
> are presenting our side just fine, but this little post really have to
> be one of the illogical I have seen on the net, and that is saying lots.

Well interestingly enough I have recieved email from fellow gamers who read
this group to quite the opposite tune. Ah well ...

graham

References

Re: PBM design and formulae - a summary
gym@dcs.ed.ac.uk (Graham Matthews)
Thu, 01 Jul 1993 11:40:27 +0000
Re: PBM design and formulae - a summary
Hel, Queen of LogeSpawn
Fri, 02 Jul 1993 16:43:13 +0000

Referenced By

Re: PBM design and formulae - a summary
gym@dcs.ed.ac.uk (Graham Matthews)
Thu, 01 Jul 1993 11:40:27 +0000
Re: PBM design and formulae - a summary
Hel, Queen of LogeSpawn
Fri, 02 Jul 1993 16:43:13 +0000