# Randomizing puzzles

13 views

### Robin Patenall

Dec 11, 2002, 8:56:11 AM12/11/02
to
I'm considering a puzzle for a game (I'm working on and off on one) that
will be random but always soluble without guessing if you think about
it, and I'd like some advice.

The first problem is when should the puzzle be randomized. I can do this
when the game is started so that if you save the game before the puzzle,
every session restored from this game will have the same solution. The
other position I can randomize the puzzle is when you start it so that
a game saved before the puzzle is triggered will have a different
solution in every restored session. Any opinions ?

The other problem I have is the complexity of the puzzle. It is a sort
of combination problem and ,solving it logically isn't too difficult
once you have figured out how. Using the logical solution I could quite
happily see it done with 8 items without annoying anybody too much. With
eight items there are 40320 possible solutions which basically precludes
brute forcing the problem. It only starts to become feasible to force
when we are down to four items (24 solutions), and then only by somebody
terminally bored.
So the problem is this, do I :

a) make it impossible unless you can find the logical solution.
b) make it trivial with the logical solution but fesible to force.
c) like a) but make the puzzle optional some how. Perhaps requiring the
player to do a lot of to-ing and fro-ing to avoid the puzzle.

I suspose that c) is optimal but it would be annoying if everybody just
skipped the puzzle. How is everybody else handling difficult possibly
skipable problems ?

--
Robin Patenall

Dec 11, 2002, 9:16:20 AM12/11/02
to
On 11 Dec 2002 05:56:11 -0800, R.Pat...@eim.surrey.ac.uk (Robin
Patenall) wrote:

>The first problem is when should the puzzle be randomized. I can do this
>when the game is started so that if you save the game before the puzzle,
>every session restored from this game will have the same solution. The
>other position I can randomize the puzzle is when you start it so that
>a game saved before the puzzle is triggered will have a different
>solution in every restored session. Any opinions ?

My preference would be to randomize at the beginning of the game so,
as you described: form the on the puzzle stays constant. This so a
player who has to backtrack from a previous save isn't forced to solve
the same puzzle in a different way after he had already solved it.

>So the problem is this, do I :
>
> a) make it impossible unless you can find the logical solution.
> b) make it trivial with the logical solution but fesible to force.
> c) like a) but make the puzzle optional some how. Perhaps requiring the
> player to do a lot of to-ing and fro-ing to avoid the puzzle.
>
>I suspose that c) is optimal but it would be annoying if everybody just
>skipped the puzzle. How is everybody else handling difficult possibly
>skipable problems ?

I would chose a) if it is relevant to the plot and c) if it isn't. If
it is relevant you would want people to make an effort to progress.

If it isn't you could do what they did in "the Pawn". They had a maze
in it which was just there for people who liked mazes. It had a sign
next to its entrance saying "This maze is not relevant to the rest of
your game". I've seen this in console RPG's as well. Grandia has
several dungeons which are optional and the player is informed of this
when entering.

Cheers,
Harry

-------------
It's a bird...
It's a plane...
-------------------
To send mail remove SPAMBLOCK from adress.

### Richard Bos

Dec 11, 2002, 1:03:16 AM12/11/02
to
R.Pat...@eim.surrey.ac.uk (Robin Patenall) wrote:

> The first problem is when should the puzzle be randomized. I can do this
> when the game is started so that if you save the game before the puzzle,
> every session restored from this game will have the same solution. The
> other position I can randomize the puzzle is when you start it so that
> a game saved before the puzzle is triggered will have a different
> solution in every restored session. Any opinions ?

The second, if the solution really _is_ logical. Applying the same
logical solution to a different set of data really isn't a different
puzzle.

> So the problem is this, do I :
>
> a) make it impossible unless you can find the logical solution.
> b) make it trivial with the logical solution but fesible to force.
> c) like a) but make the puzzle optional some how. Perhaps requiring the
> player to do a lot of to-ing and fro-ing to avoid the puzzle.

d) make it solvable if you can find the logical solution, but provide an
alternative solution to the same puzzle somewhere else, but harder to
reach than the logical solution - for example, a puzzle gnome who
requires a certain offering before he'll solve your puzzle, or a
solution book locked in a safe with the key well hidden.

This means that most people will find it easier to solve the puzzle the
official way, but those who can't will have an alternative available -
making the alternative hardish means it won't be found accidentally by
most players.

Richard

### Peter Seebach

Dec 11, 2002, 3:31:43 PM12/11/02
to

Robin Patenall <R.Pat...@eim.surrey.ac.uk> wrote:
>The first problem is when should the puzzle be randomized. I can do this
>when the game is started so that if you save the game before the puzzle,
>every session restored from this game will have the same solution. The
>other position I can randomize the puzzle is when you start it so that
>a game saved before the puzzle is triggered will have a different
>solution in every restored session. Any opinions ?

In _Spellbreaker_, there is a puzzle like this, which rejects saving during
the puzzle. To make it worse, they don't merely not allow brute-forcing;
they don't allow guessing! The game actually keeps track of what information
you have, and you don't "win" unless you have successfully proven the answer.

-s
--
Copyright 2002, all wrongs reversed. Peter Seebach / se...@plethora.net
\$ chmod a+x /bin/laden Please do not feed or harbor the terrorists.
C/Unix wizard, Pro-commerce radical, Spam fighter. Boycott Spamazon!
Consulting, computers, web hosting, and shell access: http://www.plethora.net/

### Mark J. Tilford

Dec 11, 2002, 8:57:06 PM12/11/02
to
On 11 Dec 2002 05:56:11 -0800, Robin Patenall <R.Pat...@eim.surrey.ac.uk> wrote:
>
> The other problem I have is the complexity of the puzzle. It is a sort
> of combination problem and ,solving it logically isn't too difficult
> once you have figured out how. Using the logical solution I could quite
> happily see it done with 8 items without annoying anybody too much. With
> eight items there are 40320 possible solutions which basically precludes
> brute forcing the problem. It only starts to become feasible to force

40320 is too much for brute force? I was able to brute force the vault in
John's Fire Witch (36^3 == 46656) in a few hours on a 486.

--
------------------------
Mark Jeffrey Tilford
til...@ugcs.caltech.edu

### PTN

Dec 12, 2002, 1:21:54 AM12/12/02
to

> My preference would be to randomize at the beginning of the game so,
> as you described: form the on the puzzle stays constant. This so a
> player who has to backtrack from a previous save isn't forced to solve
> the same puzzle in a different way after he had already solved it.

In 1893: AWFM I have about four different randomized puzzles. (Only two I
think are that way in the text version in the archive). I chose to randomize
all of them exactly at the point when it becomes possible to solve them in
the game. So potentially, if someone does backtrack from an old save, the
puzzle could be different the second time.

The reason I did it this way is because time is an important element to
1893. Some puzzles required a lot of searching for a specific clue, then a
very quick solution once you have that clue. By randomizing at the point of
finding the clue, I prevent players from cutting down their game time by
spending lots of time finding the clue, then restoring, and going directly
to the solution without doing all the sleuthing in-between.

Another reason is that as a player, I seldom restart an adventure from
scratch. Chances are I have one game, and I play it through to the end,
restoring to earlier saves as necessary. If you randomized at the beginning
of the adventure, a player like myself would never even know you had done
it, since every time I would approach the puzzle it would be the same.
Randomizing during the game rather than at the beginning will make it a
harder puzzle, and make the player think more, both probably good things.

-- Peter

http://www.illuminatedlantern.com/1893
1893: A World's Fair Mystery

### Stephen L Breslin

Dec 12, 2002, 2:22:42 AM12/12/02
to
On Thu, 12 Dec 2002 00:21:54 -0600, "PTN"

>I chose to randomize all of [the puzzles] exactly at the point when it

>becomes possible to solve them in the game. So potentially, if
>someone does backtrack from an old save, the puzzle could be
>different the second time.

This is good, I guess, your main point being that it's nice to
randomize puzzles after the beginning of the game.

>The reason I did it this way is because time is an important element to
>1893. Some puzzles required a lot of searching for a specific clue, then a
>very quick solution once you have that clue. By randomizing at the point of
>finding the clue, I prevent players from cutting down their game time by
>spending lots of time finding the clue, then restoring, and going directly
>to the solution without doing all the sleuthing in-between.

This sounds a bit sadistic. Why make the player go through the moves
or whatever? Making the player figure it out once is all well and
good, but making the player go through the motions of finding it out
when he already has, just to get the random key -- that would
aggrivate me. Granted, most puzzles don't provide shortcuts in the
first place, but when they're conducive to shortcuts, what's the point
of sabotaging them?

### Robin Patenall

Dec 12, 2002, 5:36:49 AM12/12/02
to
"Mark J. Tilford" <til...@ralph.caltech.edu> wrote in message news:<slrnavfqtd....@ralph.earthlink.net>...

> On 11 Dec 2002 05:56:11 -0800, Robin Patenall <R.Pat...@eim.surrey.ac.uk> wrote:
> >
> > The other problem I have is the complexity of the puzzle. It is a sort
> > of combination problem and ,solving it logically isn't too difficult
> > once you have figured out how. Using the logical solution I could quite
> > happily see it done with 8 items without annoying anybody too much. With
> > eight items there are 40320 possible solutions which basically precludes
> > brute forcing the problem. It only starts to become feasible to force
>
> 40320 is too much for brute force? I was able to brute force the vault in
> John's Fire Witch (36^3 == 46656) in a few hours on a 486.

Did you automate the problem or were you just _really_ persistant ?

Basicallly I just want to make sure that brute forcing the problem
was hard enough that the player would try to think of the logical
solution first. Really the only reason to try to force it is if the
player can not find a particular item that will obviously be useful
and will not be hidden. With this, there are easy solutions that just
take (for eight items) four times longer than optimal. I'll make sure
that this requires a few other problems to be solved that are not
requried if you have the optimal solution.

### Robin Patenall

Dec 12, 2002, 6:12:56 AM12/12/02
to
r...@hoekstra-uitgeverij.nl (Richard Bos) wrote in message news:<3df6d448....@news.nl.net>...

> d) make it solvable if you can find the logical solution, but provide an
> alternative solution to the same puzzle somewhere else, but harder to
> reach than the logical solution - for example, a puzzle gnome who
> requires a certain offering before he'll solve your puzzle, or a
> solution book locked in a safe with the key well hidden.
>
> This means that most people will find it easier to solve the puzzle the
> official way, but those who can't will have an alternative available -
> making the alternative hardish means it won't be found accidentally by
> most players.
>
> Richard

That's a good idea.

Basically, the puzzle will require the player to perform actions in a
couple of locations. The optimal solutions requires four movements
between the locations (i.e. A->B->A->B->A) other solutions require more.
My thought was to have the door jam shut after four or possibly six uses
forcing the player to take a long way round or do something repetitive
to open the door every time.

From what you've said I suspect it might be better to actually require
the player to fix the door but require a good hunt for the items to do
so (making sure that door jams with the player on the side from which
they can fix it.)

I'm even considering having the "fix" for the door be a multipurpose
thing that can only be used once. If you use it on the door you can't
shortcut another problem. Something like :

You can use the "disintegrate object" spell to remove the rock fall if
you can't figure out another way but you can't then use it to easily
save some innocents (or possibly easily save yourself at the expense
of some innocents, which shovels some more morality in to the game)

--
Robin Patenall

### PTN

Dec 12, 2002, 11:07:34 AM12/12/02
to

"Stephen L Breslin" wrote:
> This sounds a bit sadistic. Why make the player go through the moves
> which he already knows, just to receive the random combination number
> or whatever?

Yes, it probably is a bit sadistic. But conceptially speaking I don't really
like the player being able to act on knowledge the character hasn't gotten
yet. If the character, in the timeline of the game, doesn't know some
information, it seems strange to let the player act on it. This becomes a
bit more important for 1893: AWFM, because time really needs to pass in
order for the game to work as a whole. At least one, preferrably two game
days have to pass before getting to the endgame, this was one of the ways
that I ensured that that would be the case.

-- Peter

http://www.illuminatedlantern.com/1893

### Andrew Plotkin

Dec 12, 2002, 12:09:10 PM12/12/02
to

> "Stephen L Breslin" wrote:
>> This sounds a bit sadistic. Why make the player go through the moves
>> which he already knows, just to receive the random combination number
>> or whatever?

> Yes, it probably is a bit sadistic. But conceptially speaking I don't really
> like the player being able to act on knowledge the character hasn't gotten
> yet. If the character, in the timeline of the game, doesn't know some
> information, it seems strange to let the player act on it. This becomes a
> bit more important for 1893: AWFM, because time really needs to pass in
> order for the game to work as a whole.

One way to make this more palatable is to put a physical object in the
same place as the information, so that the player comes across both at
the same time. The player will need to use both to progress. The
physical object is easy to use -- the information puzzle is still the
real challenge -- but it solves your timing problem.

--Z

"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
*

Dec 12, 2002, 12:33:46 PM12/12/02
to
On Thu, 12 Dec 2002 10:07:34 -0600, "PTN"

>
>Yes, it probably is a bit sadistic. But conceptially speaking I don't really
>like the player being able to act on knowledge the character hasn't gotten
>yet. If the character, in the timeline of the game, doesn't know some
>information, it seems strange to let the player act on it.

Hmmm. I don't think I agree with you here. Indeed, in a table-top
role-playing game the player knows about things his character doesn't:
hitpoints, stats, the fact that the DM just threw his beer all over
the miniature battlefield. This makes sense (the not knowing, not the
beer spilling) since it enhances enjoyment of playing as a character
in a fictional setting.

In a table-top rpg your intelligence stats can actually prevent you
from solving a puzzle. If you, as a dumb warrior with an INT of 4 try
to work out a complex riddle, the DM is justified in telling you "Hey,
your character can't be that smart. Roll the dice to see if you CAN
solve it!"

In IF, however, the character is much more invisible to the player in
that sense. His abilities might be limited (for example, if you are a
fish you can't play a piano because you have fins) but the person
behind the keyboard can't be forced to be 'dumb'. You either 'get' a
puzzle or you don't. If you force the player to make his
player-character 'dumb' by solving the same puzzle twice, you are
taking the player OUT of the game and you start to annoy him.

What you are doing is forcing the player to adhere to some invisible
stats you set up but never told the player about.

I can think of an exception to this, though. When puzzles are
conversation-based, I can imagine that it is necessary to disallow the
use of topics the character hasn't encountered yet. Otherwise the
problem of story-flow would become to big to code. But when it comes
to logic-puzzles, involving numerous actions with objects or dials or
whatever, why torture your player with the invisible intelligence
stats?

### Brent VanFossen

Dec 12, 2002, 2:58:15 PM12/12/02
to
On 12 Dec 2002 03:12:56 -0800, R.Pat...@eim.surrey.ac.uk (Robin
Patenall) wrote:

>My thought was to have the door jam shut after four or possibly six uses
>forcing the player to take a long way round or do something repetitive
>to open the door every time.

I would hesitate to make the player suffer just because he solved a
puzzle in a suboptimum way. To force players into doing something
repetitive each time they need to pass a door or enter a room or
access a whatsit is a sure way to make them hate your game. They won't
say "Hey, if I had been more clever, I wouldn't have to do all this
every time..." Instead, they will say "This author makes me do stupid
repetitive things, and I don't like this game."

Give the clever player a bonus of some kind, if you want, but don't
punish the others.

---
Brent VanFossen

### PTN

Dec 12, 2002, 4:15:36 PM12/12/02
to

> In IF, however, the character is much more invisible to the player in
> that sense. His abilities might be limited (for example, if you are a
> fish you can't play a piano because you have fins) but the person
> behind the keyboard can't be forced to be 'dumb'. You either 'get' a
> puzzle or you don't.

I think we do agree, actually, as what you say makes a lot of sense.

Here's an example of the sort of puzzle I was talking about, in which I
would keep the randomization element in game rather than at the beginning.
Let's say you have something as simple as a combination lock on a locker.
You can't just brute force that locker open by guessing codes. So you
memorize the locker number, break into the school records room, and look up
who the locker belongs to. Then you track down that student, and take one of
his books when he isn't looking. He thinks its in his locker. You follow
him, and watch him enter the combination. When he leaves, you enter the
combo, and viola, you're in.

Now, as a player, I can restore from a previous save and just walk up to the
locker and put in the combo. Though some players might feel penalized for
not being able to do so, really it would be just plain lame for that to
work. Besides, the fun part of the game is in finding out the solution, not
the solution itself. So, you could stick in a little randomization at the
point when you watch the student enter the combination.

Of course, there are other ways to compensate for this kind of moment that

>TURN DIAL TO 24
You fumble at the dial for a while, but since you have no idea what the
combination is, your effort is futile.

Anyway, the randomization I have in mind doesn't limit a player characters
intelligence, it simply prevents players from applying information the
character has not yet received. Conversation-based topic suppression would
be another example of this, as you mentioned.

However, reading Robin's description of the puzzle itself, it seems that in
game randomization might not be fair, since it would be unclear to the
player that anything has changed, and they would just get frustrated. On the
other hand, if the randomization makes the puzzle LOOK different, too, then
the player, when RESTORING and encountering it again, could say, "Aha,
that's not what it looked like before," and start re-puzzling it out.
Without a visual signal, though, it would just be confusing.

-- Peter

http://www.illuminatedlantern.com/1893

### Mark J. Tilford

Dec 12, 2002, 4:31:08 PM12/12/02
to
On 12 Dec 2002 02:36:49 -0800, Robin Patenall <R.Pat...@eim.surrey.ac.uk> wrote:
> "Mark J. Tilford" <til...@ralph.caltech.edu> wrote in message news:<slrnavfqtd....@ralph.earthlink.net>...
>> 40320 is too much for brute force? I was able to brute force the vault in
>> John's Fire Witch (36^3 == 46656) in a few hours on a 486.
>
> Did you automate the problem or were you just _really_ persistant ?
>

Of course I automated it!

### Mike Roberts

Dec 12, 2002, 5:23:48 PM12/12/02
to
> Now, as a player, I can restore from a previous save and just
> walk up to the locker and put in the combo [without the player
> character having learned the combo in this "timeline"]. Though

> some players might feel penalized for not being able to do so,
> really it would be just plain lame for that to work.

I agree with the gist of your reasoning about PC vs. player knowledge, but I
think your approach here might be overly zealous. To my way of thinking,
proper attention to PC/player knowledge separation (well, everything about a
game, really) is for the player's benefit; in this case, as in many others,
I think playability trumps realism in terms of net player benefit. The
thing is, you as the author aren't forcing or even encouraging the player to
engage in the role playing violation here; players would be taking it upon
themselves to do so, either with full awareness of the narrative
implications or (more likely) utter obliviousness to them, and it just seems
unnecessary to force these rules - vital though they are for the author to
obey - on the player.

By way of (probably poor) analogy, this seems akin to a movie on DVD where
the director has decided to disable the "pause" and navigation buttons
during the whole third act, reasoning that the viewer won't properly
appreciate the climactic shoot-out if they don't see the whole third act in
one uninterrupted sitting. The director might be right, but I'd still be
pretty annoyed.

Contrast this to on-the-death-training puzzles, where the player learns some
vital information only through the PC's death, forcing the player to use a
meta-game RESTORE or the like to roll back the game state to a pre-death
position. It's the same kind of situation in terms of player knowledge
being confused with PC knowledge; but there's the important difference that,
in this case, it's the author who's violating the rules, and the player is
forced into the non-sensical situation.

--Mike
mjr underscore at hotmail dot com

Dec 12, 2002, 5:54:41 PM12/12/02
to
On Thu, 12 Dec 2002 15:15:36 -0600, "PTN"

>
>
>Here's an example of the sort of puzzle I was talking about, in which I
>would keep the randomization element in game rather than at the beginning.
>Let's say you have something as simple as a combination lock on a locker.
>You can't just brute force that locker open by guessing codes. So you
>memorize the locker number, break into the school records room, and look up
>who the locker belongs to. Then you track down that student, and take one of
>his books when he isn't looking. He thinks its in his locker. You follow
>him, and watch him enter the combination. When he leaves, you enter the
>combo, and viola, you're in.

The Level 9 games Snowball and Worm in Paradise used this technique,
where solving the giant mazes depended on randomized color codes. I
found this very annoying, since every time you restarted the game, the
solutions to the mazes would be slightly different. Of course, the
randomization only occurred at the start of the game, so any saves
would keep the mazes constant for that particular game.

>Besides, the fun part of the game is in finding out the solution, not
>the solution itself.

Fun being the key ingredient here.That is why I disagreed with mr.
Plotkin. When a player is backtracking he is already breaking the
illusion of the story, since he is quickly moving past parts he
already solved. Maybe to get a clean game with as little moves as
possible? Maybe to conserve lamp fuel? Whatever the reason, it is a
meta-reason: something that by definition is player-related and not
story/plot related. In such instances, the designer should not create

### Quim K Holland

Dec 12, 2002, 6:51:35 PM12/12/02
to
>>>>> "PTN" == PTN <peternepstad@(removethis)gmx.de> writes:

PTN gives an excellent example of a cheater, who
learns secret combination lock in a previous
session and then in a restored session goes on to
open the combination lock without performing the
act of finding out the combination lock, and
explains that is why he prefers to randomize the
combination when the player actually finds out the
combination.

PTN> Of course, there are other ways to compensate
PTN> for this kind of moment that IF writers have

PTN> >> TURN DIAL TO 24
PTN> You fumble at the dial for a while, but since
PTN> you have no idea what the combination is,

I personally think the traditional way *is* the
best implementation, then PTN's randomization would
be a reasonable alternative, and "randomize at the
beginning so restored game can go directly at the
solution" is the worst.

If we replace the "knowledge of the combination"
in PTN's example with "physical key that matches
the lock", then it would become apparent why the
traditional way is the right thing to do. Even if
the player know where the key is hidden, the
protagonist cannot unlock the lock until he
actually goes there to pick up the key, and brings
the key with him to the lock and uses the key on
the lock.

The protagonist picks up objects and learns about
things during the course of the game. Knowledge
is not something "inventory" command lists, but it
nevertheless is part of the story.

Randomizing at the beginning is like letting the
protagonist unlock the lock without retrieving the
key.

Randomizing at the point where protagonist learns
the combination is not as bad as randomizing at
the beginning. The author is giving an alternate
solution of brute forcing the lock.

The traditional way of doing is to treat knowledge
the say way objects are treated. Either the
protagonist has already acquired it, or he
hasn't.

### Mike Roberts

Dec 12, 2002, 8:03:08 PM12/12/02
to
"Quim K Holland" <qkho...@yahoo.com> wrote:
> PTN gives an excellent example of a cheater, who
> learns secret combination lock in a previous
> session and then in a restored session [uses it].

Yes, but: who exactly is being cheated here? Not the author, who has
assembled the game into a zip file and sent it out into the world, and is
not involved in the process of playing the game. Not the game, which isn't
a sentient being and thus is not in my view capable of being involved in
such a transaction of ethics. Not some society or community, as there are
no stakes that would unfarily go to the "cheater" and deprive someone more
deserving.

I think we have a victimless crime here, if ever there were such a thing.
The only person I can see being cheated is the cheater himself, and the only
thing he's being cheated of is the proper experience as envisioned by the
author. Really, though, there are so many other ways for an author to fail
to convey the intended experience to the player; I just can't see this one
even making the first ten pages of such a list.

### Stephen L Breslin

Dec 12, 2002, 11:16:07 PM12/12/02
to
I agree completely with Mike on this point. I think most
counterarguments will be driven by the "realism = good gameplay"
fallacy.

However, there are some gameplay issues involved in related but more
complex scenarios. Say you need to walk around to like 50 different
rooms to find the code to the locker, and all these rooms are dark,
and the lamp has limited fuel. If the player learns the code and then
doesn't have enough fuel to complete the game, but can now go to the
locker and finish the game no sweat, this is, as PTN so eloquently put
it, "just plain lame."

If part of the puzzle is doing things efficiently, then allowing a
shortcut is a problem. This is posing a hypothetical situation which
we weren't dealing with yet, I think. But ya, there's some situations
where Mike's response does not apply easily.

### PTN

Dec 13, 2002, 1:57:21 AM12/13/02
to

"Stephen L Breslin" wrote:

> I agree completely with Mike on this point. I think most
> counterarguments will be driven by the "realism = good gameplay"
> fallacy.

Yeah, its a bit hard to argue with something so obviously true as, as Mike
put it,

>>playability trumps realism in terms of net player benefit.

> However, there are some gameplay issues involved in related but more
> complex scenarios.

I was thinking along these lines, too. But also, I wonder if the puzzle
could be written so that it doesn't apply in this situation, either?

Quim Holland wrote, in his response:

> If we replace the "knowledge of the combination"
> in PTN's example with "physical key that matches
> the lock", then it would become apparent why the
> traditional way is the right thing to do.

So lets take a look at these two instances, the locked door with a key and
the locked door with a combination. Lets say to open either of them (getting
the key or combination) takes a total of 25 moves, at their most efficient
solution. On a first play through, the player fiddles around a lot, and it
ends up taking 100 turns. On a restore, or restart, and second playthrough,
the player knows exactly how to solve the problem. It will take him 25 turns
to solve the lock and key puzzle, only 3 (if you consider a 3 digit
combination) to solve the combo puzzle and proceed, unless you as an author
do something to prevent it.

In prevention lies the problem, it seems. Now, the player who got the key
probably isn't feeling the least bit put out over having to get it again
after restoring, after all, that's how these games work. Getting the
combination again, however, might be annoying to the player, because even
though the character doesn't know the answer yet, the player does, and its
annoying that you can't use that knowledge to solve the puzzle. Now,
although this is slightly irrational behavior on the part of the player, no
one ever said the player is supposed to behave rationally. Andrew's solution
would end the question completely -- oh, you need the combination AND a
small jar of vaseline. Now you have to go through the moves at least to get
the jar and you're back to a similar scenario as the key and the player

So, back to the annoyed player. The traditional method of blocking him,

>TURN DIAL to 5
You fumble with the dial but it is hopeless.

Sort of continues to annoy. Which is actually why I like the randomization
idea. And the player with previous knowledge steps up to the dial:

>TURN DIAL TO 5
You turn the dial once around to the right and stop at 5.

>TURN DIAL TO 24
You turn the dial twice to the left and stop at 24.

>TURN DIAL TO 14
You turn the dial once around to the right and stop at 14.
You tug at the lock but it doesn't open. Must not be the right combination.

Now the player, still a little annoyed (probably thinking, "I know the
combo, but this stupid game won't let me open it until I get the paper
again,") goes through the moves again, and this time discovers, much to
their surprise, the combination is different.

Now we don't have a situation where the character is not allowed to act on
the player actually really doesn't know the right answer. They need to go
retrieve the answer to open the combination. The combo, randomized, has
become as vital an object to run around and retrieve as the physical key
object did from the other door. And you've got to go get it, even if your
lamp IS running out of oil.

-- Peter

http://www.illuminatedlantern.com/1893

Dec 13, 2002, 2:26:56 AM12/13/02
to
Robin Patenall wrote:
>
> I'm considering a puzzle for a game (I'm working on and off on
> one) that will be random but always soluble without guessing

I have one question to ask: why does a player go back to
playing some part of your game? Cheating? Playing the puzzle
again? Watching or reading a cool cutscene? Walking around in
an area s/he particularly enjoyed?

In my case, I anwser yes to the last two questions much more
often than to the two previous ones.

I have played Myst many times. I've seen everything there is to
see, including some nice scenes that are not so easy to obtain
and show the authors' attention to details. I hacker speak, I
0wn this game. It has been a few years since I last played it,
but I'm confident I can still finish it under two *minutes*.
The "final combination" is not randomized. With theoretically
2^48 combinations, and at least 200 practical ones, it can
hardly be guessed. You have to play through the game to know
it, and to know where to input it... or ask a friend.
Sometimes, I just want to watch the end dialogues. Sure, I
could open one of my old savegames, but it's almost faster to
just finish the game again.

Am I doing something wrong by skipping right to the end? Should
the authors have forced me to play for ten hours for every time
I wanted to watch the end again? Do you think that Final
Fantasy VII, which puts the last savepoint at least one hour of
gameplay before the end, does rightly so?

When I go back to a game, my goals are way different from what
they were when I first played through. You are concerned that I
will ruin the game somehow by not playing it as you planned?
This concern is true for a first play-through, but is
irrelevant for the other ones. By randomizing your game, you're
not going to make me play through it again, you're going to
make me reach Alt+F4.

--

### Richard Bos

Dec 13, 2002, 4:25:57 AM12/13/02
to
brentvanfossen@NOSPAM@hotmail.com (Brent VanFossen) wrote:

> On 12 Dec 2002 03:12:56 -0800, R.Pat...@eim.surrey.ac.uk (Robin
> Patenall) wrote:
>
> >My thought was to have the door jam shut after four or possibly six uses
> >forcing the player to take a long way round or do something repetitive
> >to open the door every time.
>
> I would hesitate to make the player suffer just because he solved a
> puzzle in a suboptimum way. To force players into doing something
> repetitive each time they need to pass a door or enter a room or
> access a whatsit is a sure way to make them hate your game. They won't
> say "Hey, if I had been more clever, I wouldn't have to do all this
> every time..." Instead, they will say "This author makes me do stupid
> repetitive things, and I don't like this game."

Oh, yes. The alternative solution should not be boring or repetitive -
just harder.

Richard

### Quintin Stone

Dec 13, 2002, 9:28:31 AM12/13/02
to
On Thu, 12 Dec 2002, Adrien Beau wrote:

> When I go back to a game, my goals are way different from what they were
> when I first played through. You are concerned that I will ruin the game
> somehow by not playing it as you planned? This concern is true for a
> first play-through, but is irrelevant for the other ones. By randomizing
> your game, you're not going to make me play through it again, you're
> going to make me reach Alt+F4.

And if it was a key, a physical object and not a combination, that you
needed to get before you could win... would that annoy you just as much?

Jigsaw and Curses are very long games that don't have combination locks
you can simply open at the beginning of the game to get to the end. Does
this mean that they suffer because of it?

I mean, I get the impression now that authors just should avoid
knowledge-style puzzles altogether because they are either open to abuse,
or they create irrational annoyance in some players that an identical
(though key-based) puzzle would not.

/====================================================================\
|| Quintin Stone O- > "You speak of necessary evil? One ||
|| Weapons Master & Coder < of those necessities is that if ||
|| Rebel Programmers Society > innocents must suffer, the guilty must ||
|| st...@rps.net < suffer more." -- Mackenzie Calhoun ||
|| http://www.rps.net/ > "Once Burned" by Peter David ||
\====================================================================/

Dec 13, 2002, 11:31:41 AM12/13/02
to
On Fri, 13 Dec 2002 09:28:31 -0500, Quintin Stone <st...@rps.net>
wrote:

>And if it was a key, a physical object and not a combination, that you
>needed to get before you could win... would that annoy you just as much?

Remember: we are talking about backtracking, which means the player
wants to move past a part of the game he has already seen. This means
he is typing the proper commands quickly and isn't really reading or
playing. IN that case I think the game should not randomize complex
puzzles.

>Jigsaw and Curses are very long games that don't have combination locks
>you can simply open at the beginning of the game to get to the end. Does
>this mean that they suffer because of it?

No. See previous comment.

>I mean, I get the impression now that authors just should avoid
>knowledge-style puzzles altogether because they are either open to abuse,
>or they create irrational annoyance in some players that an identical
>(though key-based) puzzle would not.

That isn't the point. You shouldn't avoid those puzzles. But you
shouldn't make your game annoying by making the player jump through
artificial hoops when he has to move pst something he already solved.

### Quintin Stone

Dec 13, 2002, 1:32:56 PM12/13/02
to
On Fri, 13 Dec 2002, Gadget wrote:

> Remember: we are talking about backtracking, which means the player
> wants to move past a part of the game he has already seen. This means he
> is typing the proper commands quickly and isn't really reading or
> playing. IN that case I think the game should not randomize complex
> puzzles.

Not in the conversations I saw. Backtracking would suggest that the
player already solved the puzzle in their current game. If you're talking
about following a walkthrough or command script, then yeah, randomized
puzzles can be an inconvenience. But the comments I've read were of the
opinion that you should be able to get past a combination lock even if you
didn't solve any kind of puzzle, as long as you read online somewhere what
the combination is. Or if you've played it before and already know the

My message was in response to the guy who said he'd be annoyed if he had
to play through the entire game of Myst just to see the ending. That
wasn't about backtracking or following a script. It was about whether or
not he'd be annoyed if the final puzzle was an object (or group of
objects) that he had to acquire as he solved all the other puzzles in the
game.

> That isn't the point. You shouldn't avoid those puzzles. But you
> shouldn't make your game annoying by making the player jump through
> artificial hoops when he has to move pst something he already solved.

So since I've played Jigsaw and Curses all the way through and I've
already "solved" all the puzzles, I should be able to enter some kind of
code to jump to the end of the game. Otherwise it's just hoops to jump
through.... right? I mean, I've solved all the puzzles before, why should
I have to solve them again?

### Robin Patenall

Dec 16, 2002, 6:47:54 AM12/16/02
to
Thanks to everybody. All the comments have been very helpful in sorting
out my thoughts on the puzzle. I'd thought you might like to know what
I've decided on (this of course depends on nothing better comming up or
other persuasive arguments).

Point of randomization:

Puzzles like getting a locker combination can be safely randomized at
the point of getting the information without the game seeming unfair.
Unfortunately, the puzzle I have been think about does not have a
solution within the game, there is no slip of paper to find nor guard
to bribe. As it is not feasible to check whether the player is using
information from previous lives (it's not impossible just _very_
difficult, especially as I spend several hours over the weekend trying
to work out why I couldn't get an inform door working), it seems a
better, and fairer, idea to randomize the puzzle at the beginning of the
game. This means that any backtracking via SAVE/RESTORE will have the
same solution.

Difficultly :

The puzzle is basically a logic puzzle who diffculty is given by a
number N and involves moving between two locations. Ignoring the brute
force method where all !N possibly solutions are tried and requires only
2 movements. The most naive solution requires 2N movements, a better one
requires N and the optimal requires 4. My solution is to have N set to 8
(i.e. requiring 16, 8 and 4 movements respectivally) and prevent
movement between the locations after 6 movements. This will allow
optimal and slightly less optimal solutions to be performed unimpeded.
If the player requires more than 6 movements then they will also need
to solve the problem of reopening the door between the two location.

This makes it harder to solve the puzzle by non-optimal means by
imposing an additional problem, but makes sure that the puzzle is always
solvable. I might even make it so that reopening the door between the
locations will reduce your choices later on in the game by disallowing
a particular solution to another problem.

Again, thanks to everbody. All I have to do is write the thing now ...

--
Robin Patenall

Dec 16, 2002, 7:51:46 AM12/16/02
to
On 16 Dec 2002 03:47:54 -0800, R.Pat...@eim.surrey.ac.uk (Robin
Patenall) wrote:

>
>
>Again, thanks to everbody. All I have to do is write the thing now ...

This sounds like a really fair and interesting solution to your
dilemma. Good luck!

Cheers,
Harry

### Mary K. Kuhner

Dec 17, 2002, 3:23:28 PM12/17/02
to
In article <3df9a788...@news.nl.net>,
Richard Bos <r...@hoekstra-uitgeverij.nl> wrote:
>brentvanfossen@NOSPAM@hotmail.com (Brent VanFossen) wrote:

>> I would hesitate to make the player suffer just because he solved a
>> puzzle in a suboptimum way. To force players into doing something
>> repetitive each time they need to pass a door or enter a room or
>> access a whatsit is a sure way to make them hate your game. They won't
>> say "Hey, if I had been more clever, I wouldn't have to do all this
>> every time..." Instead, they will say "This author makes me do stupid
>> repetitive things, and I don't like this game."

>Oh, yes. The alternative solution should not be boring or repetitive -
>just harder.

When I played "Hunter, in Darkness" I found a quick way through the
'maze' the first time, but then did something fatal later and had
to restart. The second time, whatever I'd done the first time
didn't immediately recur, and I decided to just map the maze instead
so as to avoid further problems with it.

About 40 minutes later I was well on my way to deciding I hated
the game. The maze was consistent in a way that strongly suggested
it was mappable, but it was *huge*. (Zarf said, later, that I'd
underestimated how huge it was by two orders of magnitude.) Finally
I came to my senses and thought, hey, I was enjoying this game, now
I'm not, I must be doing the wrong thing. I went back and looked
for the elegant solution and found it, and enjoyed the rest. But
if this had happened earlier in the game, before I'd had that
experience of enjoying it, I would just have decided it was bad and
left it at that. And if I'd been feeling one hair more stubborn,
having wasted 40 minutes mapping, I might have kept on mapping, to
no avail, and never seen the endgame before giving up in disgust.

I don't think you want to trick or lure people into non-fun modes of
play--neither on purpose, nor accidentally. If you find that players
are interacting with your game in a way that bores or angers them,
that's something to fix. I'm not sure there's a flaw in "Hunter,
in Darkness"--I think it was unusually imperceptive of me to try
the brute-force solution there. But I sure wouldn't want to set up
a situation like that on purpose.

If you want to reward a clever solution, how about putting in something
nifty but nonessential that is reached only via cleverness, like an
Easter egg or a cutscene or a nifty object? _Zork II_ had one of
these, the solution to a completely gratuitous puzzle, and I liked that
a lot. I'd not have thought any worse of the game if I'd never found
it, because I had no idea it was there until I had actually solved
the puzzle.

Mary Kuhner mkku...@eskimo.com

### Mark J. Tilford

Dec 17, 2002, 5:54:43 PM12/17/02
to
On 17 Dec 2002 20:23:28 GMT, Mary K. Kuhner <mkku...@kingman.gs.washington.edu> wrote:
>
> If you want to reward a clever solution, how about putting in something
> nifty but nonessential that is reached only via cleverness, like an
> Easter egg or a cutscene or a nifty object? _Zork II_ had one of
> these, the solution to a completely gratuitous puzzle, and I liked that
> a lot. I'd not have thought any worse of the game if I'd never found
> it, because I had no idea it was there until I had actually solved
> the puzzle.
>
> Mary Kuhner mkku...@eskimo.com
>

Which puzzle is that?

### Mary K. Kuhner

Dec 17, 2002, 6:50:16 PM12/17/02
to

Mark J. Tilford <til...@ugcs.caltech.edu> wrote:
>On 17 Dec 2002 20:23:28 GMT, Mary K. Kuhner <mkku...@kingman.gs.washington.edu> wrote:

>> If you want to reward a clever solution, how about putting in something
>> nifty but nonessential that is reached only via cleverness, like an
>> Easter egg or a cutscene or a nifty object? _Zork II_ had one of
>> these, the solution to a completely gratuitous puzzle, and I liked that
>> a lot. I'd not have thought any worse of the game if I'd never found
>> it, because I had no idea it was there until I had actually solved
>> the puzzle.

>Which puzzle is that?

Trying to think how not to spoil it, since finding it was one of the
high points of my teenage IF experience....

It's the puzzle that obtains the can of grue repellant. I woke up
in the middle of the night, literally, to say "Eight! Eight!" and
stagger down to the computer room and try out my solution. You
don't ever need to solve that or even notice it, and maybe many
people find it immediately and don't see that it's a puzzle, but
it was a big one for me.

Mary Kuhner mkku...@eskimo.com