Alternatives to score in text adventures

8 views
Skip to first unread message

Greg Maddog Knauss

unread,
Aug 12, 1992, 1:51:12 PM8/12/92
to
d...@wam.umd.edu (David M. Baggett) writes:

>Adventure (Colossal Cave) set the precedent that text adventures should
>be scored games. Its "x out y points -- this makes you a ____" format
>appears in almost all its successors, and I wonder if that's just
>because we've all been following a trend somewhat blindly.

I think it's been included because people expect it. The only reason
I pay any attention to a score is because it tells me how far along I
am.

>Making the game scored creates quite a few subtle problems. One is the
>question of whether or not you should be able to finish the game
>without getting the maximum possible score. Interestingly enough,
>Adventure did *not* reqiure you to get all the points to win. Did
>Zork? Did most Infocoms? I really don't remember.

As far as I can remember, all Infocoms had score and they all required
you get get all the points to finish the game.
Which, thinking about it, is pretty silly. A score that requires
you to get all its points to finish ceases to be a score and becomes a
counter. Or a percentage.

>I've been toying with the idea of replacing the score counter with a
>percentage. The percentage would give the player a rough idea of how
>much of the whole game he/she has seen (just like current page number
>in a novel). This could be based on number of locations visited vs.
>the total number of locations, number of items seen versus total number
>of items, etc.

I did a percentage on a game I wrote eons ago and the only reason I didn't
stick with it afterwards was because it seems even more artificial than
a score. As a programmer, my problem with the whole score/percentage/
page-count-until-the-end thing is that it removes the player from the
reality of the game. Not may cave explorers can look above them and tell
that they're 75% of the way to their destination.
But as a player, I _want_ a score. It tells me if I'm actually
making any progress and is vital for the psychological reasons you
mentioned.
Maybe you could compromise, especially with a game in the
UU universe. A "Cheez Prohgrez-o-matik" or a battery that runs down as
you make progress and only as you make progress. Something that's part
of the game's universe -- nothing too distracting -- but that tells the
player what he wants to know.
--
------------------------------------------------------------------------------
Greg "Maddog" Knauss "Aieee!" gr...@duke.quotron.com
------------------------------------------------------------------------------

John Francis

unread,
Aug 12, 1992, 4:04:26 PM8/12/92
to
d...@wam.umd.edu (David M. Baggett) starts us thinking:
>I've been working on _The Legend Lives!_, a science fiction text
>adventure set in the Unnkulian Universe, and I've come to believe that
>score may not be such a great thing to have in a text adventure. I'm
>looking for some input regarding how people think having a score
>affects gameplay.

>
>Making the game scored creates quite a few subtle problems. One is the
>question of whether or not you should be able to finish the game
>without getting the maximum possible score. Should information puzzles
>just not award points since they can be skipped? (This would ensure
>that you couldn't win the game without the maximum possible score.)
>
>I can see several ways of dealing with this issue:
>
> 1. Don't award points for information puzzles . .
> 2. Code in clever restrictions that force the player to solve
> information puzzles even if they already know the information.
> 3. Make sure that all information given at the ends of information
> puzzles is randomly generated.
> 4. Eliminate scoring.
> 5. Make information puzzles scored, and allow them to be skipped,
> but award points as soon as the information is *used*.


At a general level - awarding score points is a handy way of letting the
player know he has done something right, or that he has started down the
path towards solving a puzzle. If you eliminate scores then you will have
to come up with some other form of reward system.

To adress your particular concern - I think (3) is the optimum solution.
I am pretty sure some of the later Infocom games took this approach -
in fact I know HHGTTG did (geting the right tool to open the hatch).

If you feel that is too much work, then you need solution (6) - keep
scores for information puzzles, but allow the puzzles to be bypassed.
This allows completion with less than a full score - so what?
The obsessive completists among us would still do a complete run-through
so we could finish the game with maximum score. Those who don't care
about maximum points wouldn't bother. (Did you get the point for the
dwarven magazine in Adventure? How about the Don Woods stamp in Zork?)

My feelings about the other proposed solutions:
1. I like reward systems. If I do something right, I want praise.
2. As you say, hard to do in an unobtrusive fashion.
4. I like reward systems. If I do something right, I want praise.
5. Aargh!! Reward *cheating*?
--
John Francis jo...@apollo.hp.com
The world can be divided into two classes :-
those who divide people into two classes, and those who don't.

Volker Blasius

unread,
Aug 12, 1992, 4:10:40 PM8/12/92
to
In article <1992Aug12.1...@wam.umd.edu> d...@wam.umd.edu (David M. Baggett) writes:

>I've been working on _The Legend Lives!_, a science fiction text
>adventure set in the Unnkulian Universe,

Hurra!

> and I've come to believe that
>score may not be such a great thing to have in a text adventure. I'm
>looking for some input regarding how people think having a score
>affects gameplay.

>Adventure (Colossal Cave) set the precedent that text adventures should


>be scored games. Its "x out y points -- this makes you a ____" format
>appears in almost all its successors, and I wonder if that's just
>because we've all been following a trend somewhat blindly.

>Making the game scored creates quite a few subtle problems. One is the


>question of whether or not you should be able to finish the game

>without getting the maximum possible score. Interestingly enough,
>Adventure did *not* reqiure you to get all the points to win.

Yes, and this plagued us for months, until someone got hold of the source
code and analyzed it until he found that you had to leave the magazine
(Spelunker Today) at Witt's End to get this final point that made you
Adventurer Grandmaster. I always want to have seen all of a game I play,
all the nooks and crannies, and a score is a big help in this direction.
This implies that you should be able to finish without having got all the
points, this fact telling you that you missed something or/and that you
cheated (see below). And if you cheated, it serves you right not to know
whether you missed something.

A game that (in my opinion) overdid this, was Sierra's The Colonel's Bequest.
There were loads of redundant clues you didn't get points for if you didn't
examine each and every item with a magnifying glass and stuff like that, so
I finished the game with many, many points missing. I always read the hint
book afterwards (if there is one, and for my peace of mind there better is),
so I found out what I had overlooked, but it was a bit much for my taste.

>Did Zork? Did most Infocoms? I really don't remember.

As far as I remember, you had to have all the points to finish. This is OK
if this is also an indication that you didn't miss anything, i.e. the game
is layed out this way. In fact, I prefer games that are layed out this way,
but if I really missed something, I want to know. (I admit that Witt's End
was just a big teaser, but it served this purpose very well.)

>All the Unnkulian games are written such that you can't win and not have
>gotten all the points (unless you exploit a bug in the game). This
>creates tricky situations, though, particularly in "information puzzles."
>Consider a puzzle that the player has to solve to get some information
>needed later in the game (e.g., you must get past the troll to get into
>the Wizard's Alcove containing the scroll on which the secret password is
>written). If this information is the same from game to game (i.e., not
>randomly generated), then the player could conceivably play the game
>once, write down the needed information, and then play again (or, e.g.,
>make a walkthrough) without out ever solving any of the information
>puzzles again. Should this be allowed? Should information puzzles


>just not award points since they can be skipped? (This would ensure
>that you couldn't win the game without the maximum possible score.)

>I can see several ways of dealing with this issue:

> -- argument deleted to save bandwidth --

Yeah, this is difficult. My first thought was:
If the game contains puzzles of this kind, I think it should
- give points for obtaining the information the correct way,
- allow the player to cheat by skipping over the information puzzle and
finish without all the points, and
- maybe tell the player, "You finished with x out of y because you
cheated".
But then I remembered that e.g. I cheated in Spellcasting 201 in exactly
this way by saving, going to the lessons and writing the information down
(or rather let SCRIPT write it down for me), restoring, and using the
information without the tight schedule I'd otherwise have had. The game
didn't punish me for that, and this I liked, because I hate arcade games.

I really don't know. Maybe it depends on the game (attending to the lessons
wasn't actually solving a puzzle to gain some information).

>I've been toying with the idea of replacing the score counter with a
>percentage. The percentage would give the player a rough idea of how
>much of the whole game he/she has seen (just like current page number
>in a novel). This could be based on number of locations visited vs.
>the total number of locations, number of items seen versus total number
>of items, etc.

Somehow this idea doesn't appeal to me very much. I like points that tell me
that the thing I just did or got is essential for getting onwards; it sure
increases my motivation to continue. And the number of points you got should
be enough indication as to the percentage of the way that's still ahead -
though in most games the number of points you get for doing things of
equivalent value rises exponentially the farther you get, especially in
games with large scores.

>Thoughts?

Yes, these are my momentary thoughts about that, nothing more.

Volker

--

DingDong Laboratories Ltd., Makers of Fine Eunuchs (TM)

Scott Mayo

unread,
Aug 12, 1992, 4:39:29 PM8/12/92
to
d...@wam.umd.edu (David M. Baggett) writes:

>Making the game scored creates quite a few subtle problems. One is the
>question of whether or not you should be able to finish the game
>without getting the maximum possible score. Interestingly enough,

>Adventure did *not* reqiure you to get all the points to win. Did

>Zork? Did most Infocoms? I really don't remember.

Depends on what you call *win*. Do you win when you meet the friendly
elves in Adventure? Or do you win when you get that last, blasted point,
the one that drove you to your wit's end? The latter is winning to some folk -
it shows they've *completely* plumbed the sordid depths of the game, and
the programmer's grungy soul. For that sort of player, score is everything.
--
"I forsee that you will meet a king, the father of the beautiful Princess
Plote DeVice; and he will insist that you rescue her from the keep of the
infamous Duke, Cared Bored Vilan." "Can we kill the DM now?" "No."
sm...@wang.com

John Francis

unread,
Aug 12, 1992, 5:19:22 PM8/12/92
to
In an earlier article, I write:
>d...@wam.umd.edu (David M. Baggett) starts us thinking:
>> [ . . . . ]

>> 3. Make sure that all information given at the ends of information
>> puzzles is randomly generated.
>> [ . . . . ]

>
>To adress your particular concern - I think (3) is the optimum solution.
>I am pretty sure some of the later Infocom games took this approach -
>in fact I know HHGTTG did (geting the right tool to open the hatch).

On re-reading the origininal article, it occurs to me that what you mean
by "randomly generated" is not what I mean - my interpretation is rather
more like "randomly selected from amongst several pre-generated answers".

David M. Baggett

unread,
Aug 12, 1992, 12:03:21 PM8/12/92
to
I've been working on _The Legend Lives!_, a science fiction text
adventure set in the Unnkulian Universe, and I've come to believe that

score may not be such a great thing to have in a text adventure. I'm
looking for some input regarding how people think having a score
affects gameplay.

Adventure (Colossal Cave) set the precedent that text adventures should
be scored games. Its "x out y points -- this makes you a ____" format
appears in almost all its successors, and I wonder if that's just
because we've all been following a trend somewhat blindly.

Making the game scored creates quite a few subtle problems. One is the


question of whether or not you should be able to finish the game
without getting the maximum possible score. Interestingly enough,
Adventure did *not* reqiure you to get all the points to win. Did
Zork? Did most Infocoms? I really don't remember.

All the Unnkulian games are written such that you can't win and not have


gotten all the points (unless you exploit a bug in the game). This
creates tricky situations, though, particularly in "information puzzles."
Consider a puzzle that the player has to solve to get some information
needed later in the game (e.g., you must get past the troll to get into
the Wizard's Alcove containing the scroll on which the secret password is
written). If this information is the same from game to game (i.e., not
randomly generated), then the player could conceivably play the game
once, write down the needed information, and then play again (or, e.g.,
make a walkthrough) without out ever solving any of the information
puzzles again. Should this be allowed? Should information puzzles
just not award points since they can be skipped? (This would ensure
that you couldn't win the game without the maximum possible score.)

I can see several ways of dealing with this issue:

1. Don't award points for information puzzles and allow the
player to skip them if they've played the game before and
have written down or memorized the needed information.

PROS: Easy to implement.
CONS: Somehow unsatisfying -- hacky

2. Code in clever restrictions that force the player to solve
information puzzles even if they already know the information.

E.g., (from UU2): magic word is written in Trollish. If you
don't see it written down with the "accent marks" you can't
quite remember the proper pronunciation.

PROS: If done well, hides the problem completely from the player.
CONS: It's a pain to come up with *good*, natural restrictions.

Unnatural restrictions are very obvious to the player and
seem like lame excuses.

3. Make sure that all information given at the ends of information
puzzles is randomly generated.

PROS: This is really a subset of #2, so the same PROS apply.
CONS: A bother to implement, particularly for words and phrases.

Eliminates some creativity in the game design, in that
you can't choose nifty magic words, etc. like "xyzzy."

May not be possible at all for some types of information,
particularly when that information depends on other,
necessarily fixed aspects of the game world.

There is still the chance that a player may get lucky and
guess. (Perhaps this is realistc, though!)

4. Eliminate scoring.

PROS: Solves the problem, in that it allows you to force or not
force solving of information puzzles on a puzzle by puzzle
basis.
CONS: Scoring may have a positive psyhological value. It seems
like it may increase the addictiveness of the game in the
same way that knowing you've only got 10 pages left to
read in your novel will make you sit there until you're
done even if you've got a plane to catch.

5. Make information puzzles scored, and allow them to be skipped,

but award points as soon as the information is *used*. The
game could also tell the player that it knows he/she has
skipped the puzzle. E.g., "You utter the magic word foobarx,
which you seem to have discovered through ESP."

PROS: Hides the problem fairly well from the player.

Alerts the player to the fact that ESP is *not* required
to get the information (in case he/she is following a
walkthrough).

CONS: Pretty much as tedious to implement as #2 above -- the game
still has to keep a flag for every information puzzle and
give the player a witty response when he/she skips a puzzle.

I've been toying with the idea of replacing the score counter with a
percentage. The percentage would give the player a rough idea of how
much of the whole game he/she has seen (just like current page number
in a novel). This could be based on number of locations visited vs.
the total number of locations, number of items seen versus total number
of items, etc.

Thoughts?

Dave Baggett
d...@wam.umd.edu

Scott Forbes

unread,
Aug 12, 1992, 5:53:18 PM8/12/92
to
gr...@Quotron.COM (Greg "Maddog" Knauss) writes:
>d...@wam.umd.edu (David M. Baggett) writes:
>>Interestingly enough,
>>Adventure did *not* reqiure you to get all the points to win. Did
>>Zork? Did most Infocoms? I really don't remember.
>
>As far as I can remember, all Infocoms had score and they all required
>you get get all the points to finish the game.

...Nnot quite. Planetfall, for example, had a number of alternate endings
depending on whether or not you solved certain puzzles -- but you could
"win" (reach the primary goal of saving the planet's population) without
completing some of these puzzles.

Also, it was possible to win Zork II without getting all the points,
depending on certain semi-random actions of the Wizard.

Spoilers ahead for Infocom games:

In Planetfall you could "win" without restoring the Planetary Defense
computers and/or without repairing Communications, but in either case
you would not be rescued from the planet after saving its population
(and would never get to see Ensign Twelfth Class Blather, either :-) ).

In Zork II, if the Wizard cast the "filch" spell on your character,
you would lose one or more treasures and would subsequently be unable
to give them to the demon (which prevented you from scoring all the
points).

Another interesting scoring system was that of Zork III, which had a
seven-point score; you received a point each time you had the opportunity
to obtain one of the seven items required to win the game, regardless
of whether or not you actually got the item. Your score was then a
measure of how many puzzles you'd *seen*, not how many you'd solved
(and in some cases, how many puzzles you had screwed up beyond hope --
some of Zork III's puzzles led to "you cannot win" outcomes, such as
killing the hooded figure or giving up the book or staff).


-- Scott

P.S.: The rec.games.int-fiction CFV is in David Lawrence's hands,
and voting should begin shortly....

Brendon Wyber, C.S.C.

unread,
Aug 12, 1992, 6:31:48 PM8/12/92
to
In article <1992Aug12.1...@wam.umd.edu>, d...@wam.umd.edu (David M.
Baggett) writes:
> I've been working on _The Legend Lives!_, a science fiction text
> adventure set in the Unnkulian Universe, and I've come to believe that
> score may not be such a great thing to have in a text adventure. I'm
> looking for some input regarding how people think having a score
> affects gameplay.

Most of the collect-the-treasure-type games gave points for collecting
treasures and storing them somewhere, and some bonuses for doing some deeds.
They didn't actually give points for solving the puzzle's themselves. Example:
in adventure you didn't get any points for oiling the rusty gate, just getting
the trident beyond. Zork 3 used this method. The score was out of seven, one
point for each of the items you needed to pass the Dungeon Master. You could
actually have full points and not have solved the entire game.

Infocom's Moonmist for a score stated what the player has done and yet to do.
Example:
> SCORE
You have meet all the guests and heard about the treasure hunt but
have yet to solve any of the clues or find the ghost.
That would be quite hard to implement in a true non-linear game.

Recognition of progress is important. I personally think that you should award
points for information type puzzles and let the player use that information in
other plays of the game, thus making it possible to solve the game without all
the points. You should definitely not say the player cheated in the score
response as it will alienate the player. However feel free to put a little
"tongue in cheek" comment in at the time the player uses the information.

Someone else had the idea of an object in the game giving progress "an acme
progress meter". I actually quite like that idea.

Be seeing you,

Brendon Wyber Computer Services Centre,
b.w...@csc.canterbury.ac.nz University of Canterbury, New Zealand.

"Ph-nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn."

Nathan Torkington

unread,
Aug 12, 1992, 6:58:00 PM8/12/92
to
In article <bsw1d...@wang.com> sm...@wang.com (Scott Mayo) writes:

Depends on what you call *win*. Do you win when you meet the friendly
elves in Adventure? Or do you win when you get that last, blasted point,
the one that drove you to your wit's end?

^^^^^^^^^

I love it!! :)

Nat.
(gn...@kauri.vuw.ac.nz)

Jim Giles

unread,
Aug 12, 1992, 7:38:04 PM8/12/92
to
In article <bsw1d...@wang.com>, sm...@wang.com (Scott Mayo) writes:
|> [...]

|> Depends on what you call *win*. Do you win when you meet the friendly
|> elves in Adventure? Or do you win when you get that last, blasted point,
|> the one that drove you to your wit's end? The latter is winning to some folk -
|> it shows they've *completely* plumbed the sordid depths of the game, and
|> the programmer's grungy soul. For that sort of player, score is everything.

The victory condition of the game is up to the programmer. I prefer
games where the victory condition is the achievement of some objective
- you return to town alive and in possession of "The Lost Widget" for
example. Expertise is demonstrated by achieving this objective with
the _least_ possible score. This is more natural to the real world
where the important thing is the achievement, not the irrelevant
side-information you pick up along the way (of course, you don't
know what's irrelevant until you've solved the puzzle).

I really prefer stategic games to puzzles anyway. So, on that basis,
I would argue to eliminate score entirely - or change its meaning so
it was a measure of how well you were doing (in terms of time or
resources) rather than how much you've unraveled. Games of this
latter kind can actually have value in the real world.

--
J. Giles

Nathan Glasser

unread,
Aug 13, 1992, 3:24:58 PM8/13/92
to
I played a text adventure game within the last few months which had no
score in it. It was called Enchanted Castle. I won't say what I thought of it,
but in the game, you had 3 goals: To escape from the castle, to rescue the
princess, and to destroy the castle. The first was the only real requirement
to end the game, but the other two were required to "win".

In article <1992Aug13.1...@wam.umd.edu> d...@wam.umd.edu (David M. Baggett) writes:
>But here's a more realistic thought: How about having a score but
>never telling the player what the maximum is? Then die-hard gamers
>would have the added challenge of actually determining what the maximum
>possible score *is*. (Even the game author might not know unless s/he
>bothered to check all the score increments for mutual exclusivity!) The
>more I think about this, the more I like it.

This reminds me of the Super-NES game "Super Mario World". (I played this
on a friend's system last month).

I'm not sure whether the game had a score or not, but someone can probably
supply that information. The similarity is that there is a single objective of
the game, which is to kill off a certain creature in order to rescue a
princess, but there are other things going on. There are all these different
"worlds" and "exits" from the worlds. Many of the worlds have alternate
exits which sometimes lead to new worlds (and sometimes not). The alternate
exits are sometimes difficult to find in the worlds where they are located, and
often hard to use once found.

The game keeps a count of all the different exits that are used. The number
that exist is quite a lot more than the number needed to actually complete the
rescue objective. It doesn't tell you how many the maximum is, though I think
it does indicate when you've found all the exits in some way, so the number
is known "on the street" through information passing among players, magazines,
etc. I think it is fair to say that some players, after completing the
objective, continuing playing the game for weeks afterwards just to go
through all the exits.

This seems completely pointless to me. (Please, no flames from NES fans.)
Games, in general, may not have a point from some points of view, but this
really seemed like a waste of time. If a text adventure game had such a
feature... I think it would not be desirable. After completing the objectives
of the game, I can't see any reason to keep playing it when there might not
even be anything more to do... At least a known fixed maximum score lets you
know this. Or a fixed set of objectives known to the player, even if some of
them are minor.
--
Nathan Glasser
fnord nat...@brokaw.lcs.mit.edu,mintaka!brokaw!nathan
YP-17 Nate on IRC
Beware the DDG!

Roar Foshaug

unread,
Aug 13, 1992, 10:08:44 AM8/13/92
to
In article <1992Aug12.1...@wam.umd.edu>, d...@wam.umd.edu (David M. Baggett) writes:
|>
|> Making the game scored creates quite a few subtle problems. One is the
|> question of whether or not you should be able to finish the game
|> without getting the maximum possible score....

What about games with alternative paths towards the goal state?
A good adventure game should offer different ways of doing things,
and what then about score calculations?

I don't think explicit scores are such a good idea. The motivation to
go on playing should come from the game itself as hints about how close
you are to solving a problem or eventually finishing the game. Note also
that 'score' can be viewed differently from a number in the upper right
corner of the screen. When the player finds an item that can be used for
protection, weapon, gives skills, has a value, that is also a score. For
the gold-hungry, a good motivation is to find gold. For others, the $$$
<insert favourite money-name> is the means to getting better equipment.
etc.

I also feel that the explicit score may come in the way of real problem
solving : "Oops, there I got ten points, does that mean this thing has
a usage, etc etc". The problem is of course that I don't know the strategy
the programmer uses for handing out scores. Result is that the user may end
up using more brain circuitry figuring out the score-giving strategy in
search of hints than actually applying logic to the gaming situation,
like "hmm, this is a screwdriver; what else can it be used to but
screwing? hmm, hmm...".

Screwing screws, that is... :-)


If you absolutely want scores and your game contains several paths,
you must space the scores out on the different paths, so finishing the game
along one (of several) path selections gives 100 % score. To spread the 100 %
over _all_ paths is a bad idea, as you can then finish the game with
far less, and the score looses all meaning. Pointless is a word that comes
to mind... :-)

Of course, by placing scores along the different paths so each path through
the game ends up with 100% opens the possibility for the user to get
more than 100% by moving "sideways" and covering several paths. But
so what?

The score strategy should at any rate be available in text form so the
user knows how it works.


The one reason I see to use scores is that it simplifies the
implementation. It is reasonably simple to attach scores to objects and
actions (saying magic words). It is far less simple to have a game responding
with subtle, and not-so-subtle, hints about advances. For the game to be
able to hand out such messages, the interface between the user and the
game must be more 'detailed' or 'fine-grained' in that the game must understand
more complex commands and produce responses. That leads us to creating a
more complex command parser (my favourite subject, btw :-), and also the
structure on the 'inside' of the game. I have not come up with all solutions
myself, but like speculating on these problems...


|>
|> Thoughts?
|>

Many... :-)

|> Dave Baggett
|> d...@wam.umd.edu

--
Roar Foshaug (ro...@stud.cs.uit.no)
Department of Computer Science
University of Tromsoe 'oe' is '\o{}' in Tex
N-9000 TROMSOE, NORWAY 'OE' is '\O{}' in Tex

Amit Jayant Patel

unread,
Aug 13, 1992, 9:58:20 AM8/13/92
to
In article <1992Aug12....@newshost.lanl.gov>, j...@cochiti.lanl.gov (Jim Giles) writes:
|>
|> The victory condition of the game is up to the programmer. I prefer
|> games where the victory condition is the achievement of some objective
|> - you return to town alive and in possession of "The Lost Widget" for
|> example. Expertise is demonstrated by achieving this objective with
|> the _least_ possible score. This is more natural to the real world
|> where the important thing is the achievement, not the irrelevant
|> side-information you pick up along the way (of course, you don't
|> know what's irrelevant until you've solved the puzzle).

In a game, though, it's nice to explore everything --- if you were trying to
encourage people _not_ to explore everything, then why spend time writing
those "side dishes"?

Just my opinions,

Amit

--
/\\ Amit J Patel, am...@owlnet.rice.edu
\\/ I think I'm at Rice University

David M. Baggett

unread,
Aug 13, 1992, 11:59:32 AM8/13/92
to
General comments about the thread:

You guys are killing me here! There have been so many good arguments
on all sides of the scoring issue that I'm now even more undecided
than before!

Most text adventures still seem very much limited to linear
exploration. Though I hadn't realized the connection between parallel
puzzles and scoring problems, one of my goals with Legend has been to
"delinearize" at least the first half of the game. If you go far with
this you can have several distinct paths that lead to the same
information being gleaned (or object being gained), as a previous
poster pointed out. So what happens if you do both paths just for the
heck of it? Or how about if the two paths are mutually exclusive
but one is significantly harder than the other?

The theory that having multiple solutions to a single problem makes the
game more realistic is interesting. I don't think it's been really
tested to date. Can anyone think of a text adventure where many
puzzles can be solved in several ways?

An interesting and related note (spoiler ahead for UU2 - skip this
paragraph to avoid reading it): I originally had two solutions to
the outhouse puzzle in UU2 -- you were able to burn it down until
the final beta-test version. (You can still burn it down in the
release version, but you can't win the game if you do. :)

I changed that bit at the last minute to specifically *avoid* having
two solutions to the same puzzle, because I thought one of the
solutions was too easy and hence didn't warrant the standard score for
solving the puzzle. So there's another aspect to all of this: having
multiple solutions to puzzles is *much* harder, because you (the
designer) then have much less control over the difficulty of the game
overall. Though this may not seem all bad, the genre is already
extremely limiting (compared to noninteractive fiction) and I'd hate to
give up even more control over things.

In article <BsxDH...@rice.edu> am...@owlnet.rice.edu writes:
>In a game, though, it's nice to explore everything --- if you were trying to
>encourage people _not_ to explore everything, then why spend time writing
>those "side dishes"?

Actually it's more a matter of implementation than anything else. When
you're actually coding the game (and beta-testing) you come up with all
sorts of bizarre scenarios that "should work." For the ones that are
easy to code (e.g., things that just produce a snappy response), it
takes so little time to add that you figure, "Ah, what the heck."

Things like making the game understand "xyzzy" are so trivial to do
that you begin to get in this mindset where you find it hard to argue
against putting them in. The same goes for "decorations" -- game
objects that you can't take but are included in locations for
completeness so the game doesn't say "I don't see any chair here" when
you say "examine chair" in a room with a chair in it.

From the comments so far it seems like keeping some kind of progress
indicator is important, but it's not clear that "points" are the best
manifestation of the concept. I really liked Greg's "Acme Prahgress-O-Metur"
idea; i.e., an object you carry around that tells you how you're doing.
But what the meter should tell you is still a little unclear.

Taken to an extreme, I suppose you should really be able to solve a
game with two completely different walkthroughs; i.e, with no puzzles
solved the same way in each game. With this kind of freedom it's clear
that score is totally useless, even for completists who feel that they
MUST see the whole game in one walkthrough -- it may not be *possible*
to see every part of the game in one walkthrough, since some puzzle
solutions might be mutually exclusive with some locations.

Maybe the best gauge is the number of words of game text you've read
vs. the total number of words in the game. (Wouldn't *that* be
fun to implement!)

But here's a more realistic thought: How about having a score but
never telling the player what the maximum is? Then die-hard gamers
would have the added challenge of actually determining what the maximum
possible score *is*. (Even the game author might not know unless s/he
bothered to check all the score increments for mutual exclusivity!) The

more I think about this, the more I like it. You could still have the
"You have earned XXX points -- this makes you a _____" schtick, though
you'd just stick in a clause that said if (score > some_high_value)
then say "veteran gamer," or perhaps have categories that might even be
impossible to attain (i.e., without worrying about the fact that
they're unattainable).

How about it, folks?

Dave Baggett
--
ADVENTIONS: interactive fiction (text adventures) for the 90's!
d...@wam.umd.edu * Compu$erve: 76440,2671 * GEnie: [coming soon]

Adam Justin Thornton

unread,
Aug 13, 1992, 11:22:56 AM8/13/92
to
I always kind of liked scores, although back when I was writing
games for my Apple II (Applesoft BASIC, btw, is _not_ a recommended
adventure development platform. 6502 assembler is even worse.),
they usually were of the "Pick up Object X for Y Points" variety.

I don't see why it would be terribly hard to have flags indicating
whether or not you've solved the puzzle for each given puzzle; you're
probably going to have less than 300 such puzzles even in a very large
game; if each flag is set as a bit in a long status string, I don't see
why this is a problem.

One of the better implementations of this I have seen is Infocom's
_Bureaucracy_. If you give the paranoid the password before legitimately
learning it, he snarls "You're just guessing", you can't see the pillar
(though you can climb it) in the airport before getting the ticket,
and so on.

Adam
--
"Man is conceived in sin and born in corruption and he passeth from the
stink of the didie to the stench of the shroud. There is _always_
something." -- Robert Penn Warren | Vote Cthulhu in '92! | ad...@rice.edu
If Rice shared my opinions I wouldn't have this disclaimer | 64,928

David M. Baggett

unread,
Aug 13, 1992, 12:43:21 PM8/13/92
to
ad...@owlnet.rice.edu (Adam Justin Thornton) writes:
>I don't see why it would be terribly hard to have flags indicating
>whether or not you've solved the puzzle for each given puzzle; you're
>probably going to have less than 300 such puzzles even in a very large
>game; if each flag is set as a bit in a long status string, I don't see
>why this is a problem.

It's not that it's hard or an implementation problem, it's just that
it's tedious and a pain. Also it seems like the time required to come
up with creative excuses (IMHO "You're just guessing the password!" is
too lame) could be better spent on other things.

Thomas Nilsson

unread,
Aug 13, 1992, 2:25:25 PM8/13/92
to
This discussion about scoring in adventure games have been most
interesting, giving many good arguments for keeping scores or at least
a progress indicator to push the player forward and given indication
to when he is on the right track.

A *really* good adventure game should be non-linear (I think we all
agree on this, and would produce that kind of games if it wasn't for
all the implementation and testing involved). In fact I think a good
game should have many parallell paths and situations where the player
makes a choice between one path and others, perhaps leading him
different amounts closer to the ultimate goal. This makes the game
more like a web or a directed graph (because you can't go back in
time) than a linear road with little sideway paths quickly leading
back onto the main track.

As has been said before, in this kind of game a progress indicator
would be very difficult to implement and even more to define, but by
deciding on a (or perhaps a set of) 'correct' path through the game
graph we might be able to define some events, actions or situations
that are crucial to the 'perfect' solution to the game. These events
might be considered as progress points but more important is the use
of them in the end-game. Because by giving the player indications
that his solution might not have been the most perfect one we also
increase the re-playability of a game (much like the varying endings
in Planetfall, where you might end up saving the planet but not
yourself). How about for example:


Congratulations, you managed to throw down the dictator of
Agrovenia and the people pronounce you a hero. Your future
might hold much glory and praise from the people of Agrovenia,
but will the memory of the burnt and mutilated body of your
dearest comrade Claude ever fade so that you can honestly
enjoy it?


/Thomas

--
Little languages go a long way...
----------------------------------------------------------------------
Thomas Nilsson Phone Int.: (+46) 13 12 11 67
Stenbrotsgatan 57 Phone Nat.: 013 - 12 11 67
S-582 47 LINKOPING Email: th...@softlab.se
SWEDEN Thomas_...@augs.se
----------------------------------------------------------------------

Jim Giles

unread,
Aug 13, 1992, 2:41:37 PM8/13/92
to
In article <BsxDH...@rice.edu>, am...@owlnet.rice.edu (Amit Jayant Patel) writes:
|> [...]

|> In a game, though, it's nice to explore everything --- if you were trying to
|> encourage people _not_ to explore everything, then why spend time writing
|> those "side dishes"?

Well, I don't. As I said, I like strategic games rather than puzzles.
I find those "side dishes" irritating - in my own games or someone
else's. This is not a criticism, different people have different
tastes and mine aren't drawn by arbitrary puzzle games (I've never
been patient enough to win conventional Zork-type adventure games,
for example, their arbitrary puzzles don't interest me).

--
J. Giles

Neil K. Guy

unread,
Aug 14, 1992, 2:35:51 AM8/14/92
to
he...@bach.udel.edu (Ray) writes:

>In article <BsxDH...@rice.edu> am...@owlnet.rice.edu ( Amit ) writes:

>Wouldn't the point of "side dishes" (SD's) be to make the world seem more
>three-dimensional and realistic? In real life, you can do whatever
>you want, and not everything you do is going to get you closer to
>solving a particular problem.

I tend to agree with that one. One problem I have with Adams'
Hitch-hiker game, for instance, is that it is extremely constrained
and linear in nature. I'd love the opportunity to, say, roam around
the Heart of Gold and look at things. But you can't. The only
locations that appear are those crucial to solving the game. I'm much
more interested in just mucking around than solving puzzles.
Planetfall was kind of neat in that regard.

Hey, you know what they say about getting there... I just think that
half is underestimating.

- Neil K. (n_k...@sfu.ca)

Warwick Allison

unread,
Aug 14, 1992, 1:25:42 AM8/14/92
to
d...@wam.umd.edu (David M. Baggett) writes:

>I'm looking for some input regarding how people think having a score
>affects gameplay.

There is, IMO, one BIG drawback with scoring:

It encourages cheating.

Often, when a "puzzle" is solved, the player might not be aware
just HOW important it was... or not. eg. If I get the egg and I get a
score increase of 1000 points, I'm not likely to just eat the egg
straight away, am I! Also, if I meet a wandering Glimblewop, I'll save
the game, give the egg to the Glimblewop (who will eat it, since I read
earlier in the game "Glimblewops will eat anything and everything, and
never give you anything in return"), if I get some points, I'll be
happy, otherwise, I will think "oh well, that clue was right", and
restore the game. That's CHEATING. If I didn't heed the warning, then
serves me right - I shouldn't find out my mistake until the Great
Knoblemairn asks if I have brought the beginings of life into his
Court.

Of course, I SHOULD find out my mistake (not just wander aimlessly while
the Knoblemairn just ignores me).

--
Warwick
--
_-_|\ war...@cs.uq.oz.au /Disclaimer:
/ * <-- Computer Science Department, /
\_.-._/ University of Queensland, / void (if removed)
v Brisbane, Australia. /

Darren Austin

unread,
Aug 13, 1992, 11:37:16 AM8/13/92
to
In article <1992Aug13.1...@wam.umd.edu> d...@wam.umd.edu (David M. Baggett) writes:

} ad...@owlnet.rice.edu (Adam Justin Thornton) writes:
} >I don't see why it would be terribly hard to have flags indicating
} >whether or not you've solved the puzzle for each given puzzle; you're
} >probably going to have less than 300 such puzzles even in a very large
} >game; if each flag is set as a bit in a long status string, I don't see
} >why this is a problem.

} It's not that it's hard or an implementation problem, it's just that
} it's tedious and a pain. Also it seems like the time required to come
} up with creative excuses (IMHO "You're just guessing the password!" is
} too lame) could be better spent on other things.

I think the best solution to this problem is the one that Meretzky used in
most of his later games. That was to have a random choice out of a given
set. For example the password would be randomly picked out of a set, say
"xyzzy", "plugh", "yoho", "zot". If the player has never seen the password
in this session, then no matter what he tries for the password it will
fail. This forces the player to go through the path that reveals the
password. This falls into the category of trying to hide the "scoring"
problem from the player. Plus it adds a little variety to the game. It is
more work to code, but I think it works pretty well.

--Darren

--
Darren Austin | Enough with safe and sane,
SunTech | it's time for dumb and dangerous!
darren...@sun.com |

Ray

unread,
Aug 13, 1992, 9:57:31 PM8/13/92
to
In article <BsxDH...@rice.edu> am...@owlnet.rice.edu ( Amit ) writes:

" In a game, though, it's nice to explore everything --- if you were trying to
encourage people _not_ to explore everything, then why spend time writing
those "side dishes"? "

Wouldn't the point of "side dishes" (SD's) be to make the world seem more

three-dimensional and realistic? In real life, you can do whatever
you want, and not everything you do is going to get you closer to
solving a particular problem.

Without SD's, then everything that you are possibly allowed to do
leads you towards the solution or goal, and where's the challenge of that?
You know you're doing the right thing, simply because you're allowed
to do it.

Personally, I think many games suffer from a lack of SD's which would
give the player a more realistic feel of a complete world. More thought
would have to be given to what the right thing to do is -- you wouldn't
be so sure just because you were allowed to do it.


Ray 8-)


David M. Baggett

unread,
Aug 14, 1992, 1:34:16 PM8/14/92
to
In article <39...@svin02.info.win.tue.nl> lu...@info.win.tue.nl writes:
>Designing a program to fit this scoring mechanism, would lead to a
>'programmed instruction' type of game in case of (text)adventures. As a
>result most of the program would not be played in a single session.
>This has a certain appeal, right ?

Everyone so far seems to have accepted this without comment. But the
more I think about it, the more I become skeptical that an
"ultra-realistic" game like that would actually make a good text
adventure. From an artistic standpoint such a game would actually be a
lot more primitive, since it would (by the descriptions given) be
totally picaresque. If you (the player) can just wander around
anywhere and do anything then *yeah*, it's more realisitc, but it's
also *plotless* and, some will surely aruge, *pointless*.

I suppose optimally you'd have the best of both worlds:

1. You can wander around anywhere you want in a huge game world.
2. All around there are plotted sequences you can follow. E.g., you
can choose to take the northern road into town and explore the old
haunted mansion (for a whole game), OR you can take the eastern road
into the wilderness and mess around with the wood sprites (for a
whole game), OR you can go south an run into Urk the Megaolls' Tower
of Doom (for a whole game), etc.

The problem here is that the author really has to code N complete
adventures, where N is the number of possible paths that can be taken.
To put it more in perspective, one could fold all the Unnkulian games
into a single, monstrous adventure. (For that matter, one could
fold all the Infocoms into a single monstrous adventure.)

I suspect that when it comes down to it people prefer to play a game
that is plotted and has its own settings, characters, and lore.
Computer generation of these things is out of the question (we've got
nowhere near the AI to do that), so were left with a human being doing
all of it; hence the dearth of "go anywhere, do anything" games.

And from a pragmatic standpoint, would you *really* rather wait 5
years for your next (incredibly huge) text adventure, or would you
rather have a new one every six to eight months?

Wim Jansen

unread,
Aug 14, 1992, 8:26:31 AM8/14/92
to
Really, this thread is getting way long. Ah, what the hell...

Scoring is not that difficult, provided we view the game as a directed
graph, ie. no going back (in time). Each arc represents a road (not) taken
or a puzzle (un)solved. The score for taking that road/solving that puzzle
is the 'weight' of the arc. The hiscore is reached by choosing the the
'optimal' path in the graph. Some choices may lead to death or disaster,
some to completing the game but not necessarily a glorious victory...

I think this can be applied to all sorts of games, though the 'length' of
an arc (ie. playtime) may vary considerably throughout different gametypes.

Designing a program to fit this scoring mechanism, would lead to a
'programmed instruction' type of game in case of (text)adventures. As a
result most of the program would not be played in a single session.
This has a certain appeal, right ?

Consequently, developing game would be more tedious, so a game migth end up
costing more... But, hey, it would be worth money !

Just a thought, Wim.
--
======================================================================
= -- Another Key-bored Genius gets Mouse-trapped -- =
= lu...@info.win.tue.nl ( in real life: W.M. Jansen ) =
======================================================================

Top Changwatchai

unread,
Aug 16, 1992, 7:05:38 AM8/16/92
to
If a game is going to be scored, then your option (5) is closest to how I
would handle information puzzles. That is, award points for when the info
is used. When the player is solving a particular puzzle for the first time,
she or he will want to have to go around collecting and assembling clues.
However, the seventeenth time around, he won't be as keen to go to the 83th
story of the skyscraper to reprogram the city's power grid to defeat the
alarm system to get into the office to pick the desk lock with a sharpened
kazoo just to find a scrap of paper with the code word "ululating." By then,
"realism" (in terms of forcing the character to "discover" information before
using it) starts to interfere with game enjoyment. You don't even really need
to come up with clever comments ("ESP") to enforce this realism (although
these comments don't hurt, either).

In fact, some adventures are better off without a scoring system at all.
Take _Monkey's_Island_, for instance. It's an animated adventure that's
packed with features that emphasize the storyline and humor, rather than game
mechanics. Not only does it not have a score, but it's also impossible (as
nearly as I can tell) to do something that you can't recover from. Even
falling off a cliff or sinking your ship isn't fatal. And once you figure out
a puzzle (say, threading your way through a complicated maze), the game then
skips over that puzzle when you return (i.e., you don't have to go back through
the maze when it's obvious that there's nothing more to be found there, and
that the route is already mapped).

Since playing MI, I've been thinking about whether a scoring system is
necessary for text adventures, which boast more elaborate puzzles and
situations. Probably a non-numeric scoring system is better, with a small
number of ranks which indicate how well you're doing.

Incidentally, the less linear the game (I feel) the more challenging and fun
it is to play, as well as design.

Top


Top Changwatchai

unread,
Aug 16, 1992, 7:54:16 AM8/16/92
to
Organization: The University of Texas at Austin, Austin TX
Keywords:

I'm not sure I understand you. What do you mean by "strategic games"? I've
always viewed adventure games ("interactive fiction" as Infocom calls it) as
being based on puzzles, as opposed to computer role-playing games (like
Ultima or Wizardry) or war games.

When I refer to "puzzles," though, I don't mean sterile puzzles like the
12-ball problem (which is very interesting in its own right); I refer to
puzzles that take advantage of the adventure game setting: you have a bunch
of stuff to pick up and use, a bunch of people to talk to, a bunch of skills
to exploit, and goals that may or may not be clearly defined. Take Zork III,
for example. While in some ways it was aesthetically pleasing, there simply
wasn't as much atmosphere as, for example, Planetfall. The puzzles were set
up in linear order (you had to solve one before moving on to the next), which
made them easy and which gave the sense that the player was solving the
puzzles, rather than the adventurer solving them (if that makes any sense).

Top


Top Changwatchai

unread,
Aug 16, 1992, 7:33:41 AM8/16/92
to
Just a couple more comments on puzzle nonlinearity and completeness:

The Infocom game _Arthur_ included a hints feature. After completing the game
(without using any hints), I looked through all the hints to see if I had
missed anything. I guess this is why some people like reading hint books
after solving a game. Alternatively, some games (like Infocom's _Planetfall_
and Sierra's _Leisure_Suit_Larry_) hint at major puzzles that haven't been
solved (which aren't crucial to finishing the game).

Sometimes this is fun, particularly if solving a non-crucial puzzle ("side
dish") leads to a witty or otherwise satisfying addition to the game. However,
I don't at all feel the need to get every last point in a game if it involves
simply looking at a room, searching a flowerpot, or something equally dull.

Top

Marc G. Frank

unread,
Aug 16, 1992, 11:29:07 PM8/16/92
to
In article <77...@ut-emx.uucp> ste...@ccwf.cc.utexas.edu (Top Changwatchai) writes:
>falling off a cliff or sinking your ship isn't fatal. And once you figure out
>a puzzle (say, threading your way through a complicated maze), the game then
>skips over that puzzle when you return (i.e., you don't have to go back through
>the maze when it's obvious that there's nothing more to be found there, and
>that the route is already mapped).

This brings up another point. Does anyone else think mazes in games are
a Bad Thing? For me, nothing detracts from my enjoyment of a game more
than a maze. Of course, nearly every (or perhaps all of them; I say
"nearly" in case I've simply forgotten one) text adventure I've played
has had a maze.

I can play games without mapping them -- except for that damned ObMaze.
Mazes aren't fun, they're just irritating.

Is there any reason you designers out there keep including mazes except
tradition? Does anyone think they add to the playability of a game?

+--------------------------------------------------------------------+
| Marc G. Frank mgf...@avernus.com |
+--------------------------------------------------------------------+
| "If the Pentagon needs, say, fruit, it will argue that it must |
| have fruit that can withstand the rigors of combat conditions, and |
| it will wind up purchasing the FX-700 Seedless Tactical Field |
| Grape, which will cost $160,000 per bunch, and which will have an |
| 83 percent failure rate." -- Dave Barry |
+--------------------------------------------------------------------+

James Hague

unread,
Aug 17, 1992, 9:13:47 AM8/17/92
to
Top Changwatchai writes:
>
>Incidentally, the less linear the game (I feel) the more challenging and fun
>it is to play, as well as design.

As do most people reading this group, I would assume. Unfortunately, linearity
seems to be rather popular these days and the buying public doesn't have many
complaints. (IMO linearity replaces fun with frustration--get to a puzzle, get
stuck, hit computer, sleep on it, continue.)

It is at least worth considering that there is somewhat of a backlash against
games which require too much thought and don't hold the players' hand every
step of the way. Witness all the people in comp.sys.ibm.pc.games and rec.games.
video who get frustrated and ask for "cheats" after playing a game for two
days. Then you also have people complaining because they solved a game in only
one...

--
James Hague
exu...@exu.ericsson.se

David M. Baggett

unread,
Aug 17, 1992, 10:00:03 AM8/17/92
to
exu...@exu.ericsson.se writes:
>>Incidentally, the less linear the game (I feel) the more challenging and fun
>>it is to play, as well as design.
>
>As do most people reading this group, I would assume. Unfortunately, linearity
>seems to be rather popular these days and the buying public doesn't have many
>complaints. (IMO linearity replaces fun with frustration--get to a puzzle, get
>stuck, hit computer, sleep on it, continue.)

Actually one of the biggest criticisms of UU2 has been that it's *not*
totally linear, and hence is much harder. If you have one puzzle right
after another your "solution space" is much smaller, and on top of that
you alway know exactly which puzzle to work on when. If you've got
10 puzzles still unsolved, how do you know which to work on next? Unless
the puzzles are totally independent of one another, there will be some
required solution order. (Of course, a completely strict ordering
makes the game linear. Linear but hard.)

My conclusion is that you can't make everyone happy. (What an insight.)

Thomas Nilsson

unread,
Aug 17, 1992, 7:46:58 AM8/17/92
to
mgf...@avernus.com (Marc G. Frank) writes:

>This brings up another point. Does anyone else think mazes in games are
>a Bad Thing? For me, nothing detracts from my enjoyment of a game more
>than a maze. Of course, nearly every (or perhaps all of them; I say
>"nearly" in case I've simply forgotten one) text adventure I've played
>has had a maze.

There was a thread on this a while back and I think the general
consensus was that they are generally a Bad Think, unless done with a
lot of careful thought and inovation.

Thomas Nilsson

unread,
Aug 17, 1992, 8:14:59 AM8/17/92
to
d...@wam.umd.edu (David M. Baggett) writes:

>... If you (the player) can just wander around


>anywhere and do anything then *yeah*, it's more realisitc, but it's
>also *plotless* and, some will surely aruge, *pointless*.

[stuff deleted]

>The problem here is that the author really has to code N complete
>adventures, where N is the number of possible paths that can be taken.

This is not necessarily the case. It is very easy to confuse the two
very different types of movement that occurs during the playing of an
adventure game. One (which is the easy part) is the purely physical or
geographical movement through the map with all its details (objects).
We tend to stick to this when we talk about paths in the game.

But the more interesting movement is through the plot, which can be
viewed as the movement through time and events. This type of movement
is much more interesting because *this* is what makes up the story.

So like a book author the adventure author should create a world as a
stage for the events that makes up the story he is going to tell. Thus
the locations of the game need not be, or rather, shouldn't be mixed
up with the events that are taking place there.

This leads to a methodology that likes like:

- define your world losely, in concept, in time a.s.o
- describe your story as a sequence of events and from this find what
the main settings and vital locations are
- design and detail the geography of your world

The main idea here is of course to place your story in the locations
necessary for its progress, not the other way around.

Doing this will, I think, give you the best of two worlds, Dave, the
player may freely explore the immediate surroundings until the
conditions arise for the next scene of your plot to carry or sweep the
hero (sorry, player!) forward in the story.

Agreed, this is much more work, both defining a world, a story and
find the good triggering mechanisms. But, what good adventures we
would get....

Jonathan R. Ferro

unread,
Aug 17, 1992, 4:41:10 PM8/17/92
to
mgf...@avernus.com (Marc G. Frank) writes:
> This brings up another point. Does anyone else think mazes in games are
> a Bad Thing? For me, nothing detracts from my enjoyment of a game more
> than a maze. Of course, nearly every (or perhaps all of them; I say
> "nearly" in case I've simply forgotten one) text adventure I've played
> has had a maze.
>
> I can play games without mapping them -- except for that damned ObMaze.
> Mazes aren't fun, they're just irritating.

It's time once again to quote from the seminal anonymous document,
ADV.DOC (which you can obtain via anonymous FTP from
msdos.archive.umich.edu in the ARC archive msdos/games/textadv/tess.arc)

##########
Locations
---------
- No illogical mazes, mixed-up directions, or un-mappable locations
- A maze should be a real maze, not a set of impossible-to-map rooms.
- Even if the adventure consists of independent sets of locations,
there should still be a logical layout to the map.

- No empty or useless rooms
- A location just for the sake of completeness may add to the theme,
but from the player's point of view it is disappointing.
- This is especially true if the descriptions of the rooms are short,
since for long descriptions the location can significantly add to the
mood and feel of the adventrue.
##########

It has been discussed in this group before and decided that the true
complaint is with non-returning exits, meaning situation where you go,
say, west through a door and with no warning are now unable to go east
(the opposite direction) to return to your previous location.

Situations can still be developed that are "mazy" but much more
satisfying by such means as (a) giving accurate warnings when
connectivity becomes confusing, such as the descriptions of how the
hallways curve near the river and dam in Zork I, or (b) having a real
maze which has true returning exits and gives an intellectual challenge
rather than a grunt-work challenge, such as the glass maze in Sorcerer.

It's quite clear that deliberately unmappable mazes are no longer the
state of the art and a sign of Infocom's greatness that they realized so
early that determining the map was not the true puzzle and produced
Suspended to prove it.

--
Jon Ferro MIT Transportation Modelling Research Center (TMRC)
jf...@andrew.cmu.edu "We can hack it!"
SGNTR VRS VRSN 3.1B: Strppd nd cmprssd fr qckr trnsmssn! nfct yrs tdy!

Nathan Torkington

unread,
Aug 17, 1992, 8:45:57 PM8/17/92
to
In article <1992Aug17.1...@wam.umd.edu> d...@wam.umd.edu (David M. Baggett) writes:

Actually one of the biggest criticisms of UU2 has been that it's *not*
totally linear, and hence is much harder. If you have one puzzle right
after another your "solution space" is much smaller, and on top of that
you alway know exactly which puzzle to work on when. If you've got
10 puzzles still unsolved, how do you know which to work on next? Unless
the puzzles are totally independent of one another, there will be some
required solution order. (Of course, a completely strict ordering
makes the game linear. Linear but hard.)

Didn't Wonderland work this way? There was almost always lots and
lots of cute little puzzles to explore. It definitely didn't seem
linear from the little bit I played.

Nat
(gn...@kauri.vuw.ac.nz -- Nathan Torkington -- is the electronic text and
MS-DOS archivist for the Victoria University of Wellington, New Zealand)

Roar Foshaug

unread,
Aug 18, 1992, 2:02:41 PM8/18/92
to
In article <1992Aug13.1...@wam.umd.edu>, d...@wam.umd.edu (David M. Baggett) writes:
|> General comments about the thread:
|>
|> You guys are killing me here! There have been so many good arguments
|> on all sides of the scoring issue that I'm now even more undecided
|> than before!
|>

heh heh heh :-)

|> The theory that having multiple solutions to a single problem makes the
|> game more realistic is interesting. I don't think it's been really
|> tested to date. Can anyone think of a text adventure where many
|> puzzles can be solved in several ways?

Well, simply by allowing different tools to be used for the same things, the
player will can choose between different routes though the game collecting
different tools. Mostly, such multi-solutions to puzzles will only be spotted
by the player afterhand, when the game has been played over again with
different choices (ignoring the fact that most players will work on a
number of saved versions of the game, trying new solutions to earlier saved
ones and so on)...

And of course, multi-solutions may be so simple as the player going up
or down at some (one-way ?) point in the 'landscape'.

|> Dave Baggett
|> --
|> ADVENTIONS: interactive fiction (text adventures) for the 90's!
|> d...@wam.umd.edu * Compu$erve: 76440,2671 * GEnie: [coming soon]

Roar

Roar Foshaug

unread,
Aug 18, 1992, 2:11:49 PM8/18/92
to

Agreed. That is what I think too. An adventure game can in fact be viewed as
a role-playing game, and to make the world three-dimensional as Ray calls
it there should be "side dishes". That means that objects should have
different uses, there should be hints and clues that do not nescessarily
lead you forward towards the completion of the game. If such things are
well implemented you will want to play the game over and over again just to
be 'in that world'. The "side dishes" then represent undiscovered land
more than means for getting to the end.

Such simple things as being able to use the space shuttle key to open the
bottle of water when dying of thirst in the desert will bring 'depth' to
a game.

Jacob S. Weinstein

unread,
Aug 18, 1992, 7:09:57 PM8/18/92
to
In article <1992Aug13.1...@wam.umd.edu> d...@wam.umd.edu (David M. Baggett) writes:
>General comments about the thread:
>
>
>
>The theory that having multiple solutions to a single problem makes the
>game more realistic is interesting. I don't think it's been really
>tested to date. Can anyone think of a text adventure where many
>puzzles can be solved in several ways?
Yup- as I remember, _Wishbringer_ fit the bill.POSSIBLE SPOILERS AHEAD!
As part of the game, one acquired a magic stone. THe stone enabled you
to make a variety of wishes, each of which required a certain object.
For example, given the stone and an umrella, one could wish for rain. I
believe that, for each and every puzzle in the game, there were at least
two possible solutions: one involving making a wish, and one not. For
example, at one point you stumbled upon a seal stuck in a ditch. You
could extend a stick into the ditch, and the seal would climb out; or
you could wish for rain, filling up the ditch and letting the seal swim
out. The catch was that you didn't get points if you solved the puzzle
with magic-- the explanation was something like, "One ought to depend on
one's own wisdom, not on magic." I enjoyed the game quite a bit, and
generally tried to figure out both ways of solving each puzzle.

>From the comments so far it seems like keeping some kind of progress
>indicator is important, but it's not clear that "points" are the best
>manifestation of the concept. I really liked Greg's "Acme Prahgress-O-Metur"
>idea; i.e., an object you carry around that tells you how you're doing.
>But what the meter should tell you is still a little unclear.
>

Agreed. I'm currently working on a text-adventure game that uses a
similar idea. But you'll just have to wait until it comes out.


>But here's a more realistic thought: How about having a score but
>never telling the player what the maximum is? Then die-hard gamers
>would have the added challenge of actually determining what the maximum
>possible score *is*. (Even the game author might not know unless s/he
>bothered to check all the score increments for mutual exclusivity!) The
>more I think about this, the more I like it. You could still have the
>"You have earned XXX points -- this makes you a _____" schtick, though
>you'd just stick in a clause that said if (score > some_high_value)
>then say "veteran gamer," or perhaps have categories that might even be
>impossible to attain (i.e., without worrying about the fact that
>they're unattainable).

I dunno. This would raise a problem with red herrings. (SPOILER FOR UU2)
If you had a system like this in place, I would probably still be
playing UU2, convinced that if I could only figure out how to use that
damn teddy bear, I'd be able to increase my score. Now, granted, one
wants the players to play one's game as long as possible, nbut one wants
them to enjoy it in the process. After a certain point, I'd give up
trying to use the teddy for every conceivable purpose, and I'd quit,
thinking to myself, "What a stupid game! See if I send in my shareware
fee ever again!"

Top Changwatchai

unread,
Aug 21, 1992, 3:32:02 AM8/21/92
to
In article <1992Aug18.1...@news.uit.no> ro...@stud.cs.uit.no (Roar Foshaug) writes:
>In article <1992Aug13.1...@wam.umd.edu>, d...@wam.umd.edu (David M. Baggett) writes:
>|> The theory that having multiple solutions to a single problem makes the
>|> game more realistic is interesting. I don't think it's been really
>|> tested to date. Can anyone think of a text adventure where many
>|> puzzles can be solved in several ways?
>
>Well, simply by allowing different tools to be used for the same things, the
>player will can choose between different routes though the game collecting
>different tools. Mostly, such multi-solutions to puzzles will only be spotted
>by the player afterhand, when the game has been played over again with
>different choices (ignoring the fact that most players will work on a
>number of saved versions of the game, trying new solutions to earlier saved
>ones and so on)...
>
>Roar
>
>--
>Roar Foshaug (ro...@stud.cs.uit.no)
>Department of Computer Science
>University of Tromsoe 'oe' is '\o{}' in Tex
>N-9000 TROMSOE, NORWAY 'OE' is '\O{}' in Tex

Well, Infocom always claimed to have multiple solutions to puzzles, although
their adventures weren't loaded with this kind of thing. One example that
springs to mind is from Zork I: at one point, you need to cross a dammed [sic]
lake, which you can do by touching a mirror on one side. This zaps you over to
another mirror on the other side, where you can continue exploring. However,
there is also a treasure buried beneath the lake, so you actually have to
drain the lake to get the treasure (which, incidentally, also allows you to
cross the lake without using the mirrors at all).

This is a fair example of a puzzle that has more than one solution, although
only one is ultimately correct. I sort of like this idea, at least from a
puzzle-solver's perspective: not only do you have to find a solution, but
you must find the best one.

This kinda ties in with the idea of giving hints to the player to indicate
progress (as opposed to a point system). Finding a non-optimal solution to
a puzzle can allow the player to discover clues to find the best solution.
In the above puzzle, for instance, a book found across the lake may hint that
a treasure is buried beneath the lake (Zork I didn't do this, of course).
Thus using the mirrors can provide clues about draining the lake.

Top


--
ste...@ccwf.cc.utexas.edu
"I'm gonna live forever or die trying."

Top Changwatchai

unread,
Aug 21, 1992, 4:03:04 AM8/21/92
to
In article <1992Aug17.0...@avernus.com> mgf...@avernus.com (Marc G. Frank) writes:
>
>This brings up another point. Does anyone else think mazes in games are
>a Bad Thing? For me, nothing detracts from my enjoyment of a game more
>than a maze. Of course, nearly every (or perhaps all of them; I say
>"nearly" in case I've simply forgotten one) text adventure I've played
>has had a maze.
>
>I can play games without mapping them -- except for that damned ObMaze.
>Mazes aren't fun, they're just irritating.
>
>Is there any reason you designers out there keep including mazes except
>tradition? Does anyone think they add to the playability of a game?
>
>+--------------------------------------------------------------------+
>| Marc G. Frank mgf...@avernus.com |
>+--------------------------------------------------------------------+

I think there was a maze discussion here a while back (though I missed it).
My view (and probably that of lots of people) is that the traditional maze
has been done to death. A maze should either fit with the theme of the game
or offer a new twist (or preferably both). A prevalent strategy for
solving the standard all-rooms-alike maze (the method I use, at least) is to
drop objects in the rooms to distinguish between them; this is pretty
mechanical, and thus not very fun. (In fact, I solve large arbitrary mazes
with a chart of rooms and directions instead of directed graphs.) A standard
twist on this traditional maze is to not allow players to drop items; for
instance, the maze could be in a swamp where dropped items sink in the mud.

I tend to like the mazes that aren't really mazes, like the one in Zork III
where you move the walls around, or the one in Enchanter where you can change
the passageways. Right now I'm designing a "maze" where the architecture of
the place is the key to solving it. Does anybody else have neat maze
variants, or further comments?

Corey D. Stinson

unread,
Aug 26, 1992, 11:32:30 AM8/26/92
to
Actually, that isn't necessarily the case. I solved Zork I w/o
getting all the points...

Carl Muckenhoupt

unread,
Aug 31, 1992, 4:35:04 PM8/31/92
to
Actually, Infocom did a lot of multiple-solution puzzles. Few people realize
it because everyone assumes that the solution they found was the only one.
However, the end result of the different solutions is ultimately the same
in nearly all cases. For example, in the end of Enchanter, the Evil Wizard
summons a warrior to kill you. You can turn him into a newt, or you can
cast a spell on him that makes him friendly. In either case, he leaves you
alone and that is that. The only real difference is that one option is
open to people who have found the turn-into-newt spell and the other isn't.
(You have to have the other spell to get that far.) Another example: in
Hitchhiker's Guide, you must retrieve a flowerpot from the belly of a whale
and escape in a limited time. You can use your "electonic thumb" to get the
spaceship to rescue you, or you can simply put the flowerpot in the
"thing that your aunt gave you which you don't know what it is," an item that
keeps reappearring in your inventory no matter how many times you drop it.
In the latter case, you die and are resurrected on shipboard, but the "thing"
brings the flowerpot back to you. (Yes, this solution was intended by the
authors.) In either case, however, you wind up back on the ship with the
flowerpot in your hand.
--

Carl Muckenhoupt bme...@occs.cs.oberlin.edu
HAVE YOU FORGOTTEN SOMETHING?

Reply all
Reply to author
Forward
0 new messages