Book review: Chris Crawford on Interactive Storytelling

13 views
Skip to first unread message

Mike Rozak

unread,
Oct 29, 2004, 2:07:50 AM10/29/04
to
I just read through "Chris Crawford on Interactive Storytelling". You can
read many lengthy reviews on Amazon.com. Here's my synopsis for how the book
is relevent to virtual worlds and adventure games:

1) Chris Crawford spends a fair amount of the book explaining why the
emperor has no clothes. I find this interesting even if I don't agree with
all his points, although I agree with most of them.

2) The rest of the book is devoted to his vision of Interactive
Storytelling. Basically, this amounts to making a world that is divided into
"stages" (kind of like nodal MUD/adventure-game rooms, but they're all
directly connected to one another without standard NSEW connections) and
filling the world with actors. (As opposed to a MUD or adventure game that
often emphasizes objects above AI.)

3) The actors are all controlled by AI. Rather than a combat or movement
oriented AI, this AI is socially oriented. Basically, characters all have
personalities (values) and opinions about other characters. When they see
another character doing something, or talk to the other character, their
opinion changes. Plus, characters gossip with one another and pass stories
around, causing opinions to change and further action.

4) The world is controlled by a "director" AI that makes sure the story
advances, characters don't get stuck, or whatever else it wishes.

5) Throw a player into the mix.

What players DO in the game/entertainment is interact with NPC API's.
Players do NOT solve puzzles (unless you classify maniuplating AIs as puzzle
solving), do NOT engage in combat, and do NOT carry around or use objects.

If you are interested in AI and/or a different approach to MUDs/adventure
games, then I recommend the book. I have been looking for a non-academic NPC
AI book for awhile, and this is the best one I've found so far. (Note that
pathfinding and other standard-game AI issues are NOT dealt with, which is
fine with me.)


Mike Rozak
http://www.mxac.com.au


Cubist

unread,
Oct 30, 2004, 7:00:06 AM10/30/04
to
"Mike Rozak" <Mike...@bigpond.com> wrote in message news:<W6lgd.2577$K7....@news-server.bigpond.net.au>...
> I just read through "Chris Crawford on Interactive Storytelling"...
[much snippage]

> What players DO in the game/entertainment is interact with NPC API's.
So far, so good...

> Players do NOT solve puzzles (unless you classify maniuplating AIs as puzzle
> solving)

Hmmm. Questionable, but defensible.

>do NOT engage in combat
Okay, this is silly. Even under Crawford's "NPCs interact with
(N)PCs" paradigm, with AIs forming/affecting the opinions of other
NPCs, it should be possible for NPC #1 to decide that NPC #2 needs
killin' -- and that might very well mean *combat*. And who the hell is
Crawford to decide that the *PC* must *never* engage in combat,
anyway?

>and do NOT carry around or use objects.

Absolute rubbish. I guess Crawford himself carries nothing in his
pockets, or *never* uses any such objects that he *does* carry...

Apparently, Crawford has noticed that certain things are
overemphasized in Interactive Fiction, and his response is that these
overemphasized things must be completely eradicated, *not* returned to
an appropriate level of importance. Silly.

RootShell

unread,
Oct 30, 2004, 8:29:20 AM10/30/04
to
"Mike Rozak" <Mike...@bigpond.com> escreveu na mensagem
news:W6lgd.2577$K7....@news-server.bigpond.net.au...

> I just read through "Chris Crawford on Interactive Storytelling". You can
> read many lengthy reviews on Amazon.com. Here's my synopsis for how the
book
> is relevent to virtual worlds and adventure games:

[snip cut]

> If you are interested in AI and/or a different approach to MUDs/adventure
> games, then I recommend the book. I have been looking for a non-academic
NPC
> AI book for awhile, and this is the best one I've found so far. (Note that
> pathfinding and other standard-game AI issues are NOT dealt with, which is
> fine with me.)
>
> Mike Rozak
> http://www.mxac.com.au

Mike... But in the end? is it still worthy to buy it or not? Since im trying
to get my hands on every good books that exist about IF!

Any there any particular books (specially Inform related) that the members
of this newsgroup recommend?

I already own "Twisty Little Passages, An Approach To Interactive Fiction"
by Nick Montfort (and loved it, I read it in 2 days)

I'm also waiting for Amazon to send me the following two books (which i
already read in PDF from the ifarchive, but decided to purchase the printed
edition):

"The Inform Beginner's Guide (3rd Edition)" by Roger Firth and Sonja
Kesserich
"Inform Designer's Manual (4th Edition)" by Graham Nelson

I'm also keeping my eye out on "IFTheory" by Emily Short, which looks like
it's going to be a huge success, but unfortunately it also seems that is
going to take ages to be released :( (or so it seems)

But somehow, im still eager for knowledge about interactive fiction, and, as
such, would love to know other good books about this matter.

Please advise (i know sometimes you may think of me as a "rude person" but
im not, so please give me some help on this).

Thanks,
RootShell


Adam Thornton

unread,
Oct 30, 2004, 11:08:35 AM10/30/04
to
In article <116f9bab.04103...@posting.google.com>,

Cubist <cub...@aol.com> wrote:
>
> Apparently, Crawford has noticed that certain things are
>overemphasized in Interactive Fiction, and his response is that these
>overemphasized things must be completely eradicated, *not* returned to
>an appropriate level of importance. Silly.

But he uses FORTY-ONE state variables to track the emotional states of
his actors! FORTY-ONE!

For a thorough and well-considered deconstruction of Crawford's
arguments, I suggest you play the Holodeck scene of _Stiffy Makane: The
Undiscovered Country_.

Well, except for the "thorough" part. But, actually, yes,
"well-considered"; it's only a few sentences, but it sums up my
fundamental objection to the Erasmatron (and things of its ilk)
concisely but accurately.

Adam

samwyse

unread,
Oct 30, 2004, 3:42:54 PM10/30/04
to
On or about 10/30/2004 7:29 AM, RootShell did proclaim:

> Mike... But in the end? is it still worthy to buy it or not? Since im trying
> to get my hands on every good books that exist about IF!

Chris Crawford is the gaming equivalent of, oh, Kurt Vonnegut. Both did
some absolutely brillant work early on, both probably have their best
days behind them, both still have more interesting ideas than anyone
else in their fields, and both are doing things that are nowhere near
where the mainstream currently lies. So, you buy their books despite
fears that they are irrelevant to where you are and what you're doing,
because they just might be mapping out where you're going to be a few
years from now.

Mike Rozak

unread,
Oct 30, 2004, 6:07:21 PM10/30/04
to
Cubist wrote:
>>do NOT engage in combat
> Okay, this is silly. Even under Crawford's "NPCs interact with
> (N)PCs" paradigm, with AIs forming/affecting the opinions of other
> NPCs, it should be possible for NPC #1 to decide that NPC #2 needs
> killin' -- and that might very well mean *combat*. And who the hell is
> Crawford to decide that the *PC* must *never* engage in combat,
> anyway?

His system allows for any sort of verb, which could allow for "combat", but
not in the RPG sense of lots of math, hit points, etc. It would be combat on
a much simpler form, where either the actor is killed or not.


>>and do NOT carry around or use objects.
> Absolute rubbish. I guess Crawford himself carries nothing in his
> pockets, or *never* uses any such objects that he *does* carry...

His book briefly mentions possession of objects, but I think his system
treats possession using a different mechanism than adventure games. I
suspect they're more of a state variable (although I could be wrong), and
object possession is significantly less important than interaction with
NPCs.


> Apparently, Crawford has noticed that certain things are
> overemphasized in Interactive Fiction, and his response is that these
> overemphasized things must be completely eradicated, *not* returned to
> an appropriate level of importance. Silly.

I think he's trying to boil an "interactive story" to do down to his
essence. For him, the most important parts are characters and their
relationships with one another. I felt that he boiled it down a bit too
much, but that's all a matter of opinion, and it doesn't detract from the
social AI ideas he puts forth. However, as soon as you add objects, combat,
and connected rooms AI becomes more difficult to write.

--

Mike Rozak
http://www.mxac.com.au


Mike Rozak

unread,
Oct 30, 2004, 6:21:55 PM10/30/04
to
Rootshell wrote:

> Mike... But in the end? is it still worthy to buy it or not? Since im
> trying
> to get my hands on every good books that exist about IF!
>
> Any there any particular books (specially Inform related) that the members
> of this newsgroup recommend?
>
> I already own "Twisty Little Passages, An Approach To Interactive Fiction"
> by Nick Montfort (and loved it, I read it in 2 days)

Some random points:

1) If you write your adventures in Inform, TADs, or some other existing IF
creator, I don't think you'll be able to impliment Crawford's ideas. (For
example: He describes a database of the history of what all NPCs have done
during the game, keeping indecies into it for gossip points. Inform/TADs may
not handle this.) If you're using a language like C, C++, Java, Python,
etc., you'll have no problem.

2) Chris Crawford probably wouldn't call his system IF, but you can think of
it as IF without many rooms, objects, or puzzles, but with a lot of NPCs.
(Maybe a bit more like Infocom's Deadline?) If you want to get a different
viewpoint and potential ideas for IF, it's a good book to read.

3) It is nothing like Twisty Little Passages. TWP is mostly an overview of
what's out there. Crawford's book discusses technique, as well as a lot of
"the emperor has no clothes" points that will make you think. Some people
may react negatively to his observations though.

4) There aren't many books on IF. The Inform designers guide is very good.
TLP is a good overview of what's out there and why it works. Emily Short was
working on a book, but I haven't heard about it lately. I would add
Crawford's book, even though it's not strictly about IF. I also read Hamlet
on the Holodeck, but found it less useful. You might also look into AIML
(for creating chatterbots); there is a book, but I haven't read it. If
you're interested in multiuser works then read Richard Bartle's book,
"Designing virtual worlds", which is really about MUDs and MMORPGs. The
specific code for creating a multiuser dungeon/IF can be gotten from Ron
Pendelton's "MUD game programming".

Mike Rozak

unread,
Oct 30, 2004, 6:30:21 PM10/30/04
to
Samwyse wrote:
> Chris Crawford is the gaming equivalent of, oh, Kurt Vonnegut. Both did
> some absolutely brillant work early on, both probably have their best days
> behind them, both still have more interesting ideas than anyone else in
> their fields, and both are doing things that are nowhere near where the
> mainstream currently lies. So, you buy their books despite fears that
> they are irrelevant to where you are and what you're doing, because they
> just might be mapping out where you're going to be a few years from now.

I view it as a matter of design riskyness. If a game designer makes another
Myst or another Everyquest they know they'll sell copies of their software,
but they'll never have a spectacular success. Chris Crawford (as well as a
few others) try to do something completely different. Doing something
completely different usually results in failure, but once in awhile it
produces an amazing success.

Furthermore, Chris Crawford is rabidly against eye candy common to popular
games, just as people in this newsgroup tend to be. (I'm someplace
in-between.) It's difficult to make a mass-market game (and hence be
"successful") without it. I'm currently enjoying Myst IV, but it's 99% eye
candy, with fundamental logic someplace between "Hunt the Wumpus" and
Crowther's "Adventure".

Cubist

unread,
Oct 31, 2004, 10:24:17 AM10/31/04
to
"Mike Rozak" <Mike...@bigpond.com> wrote in message news:<7uUgd.4254$K7....@news-server.bigpond.net.au>...

> 4) There aren't many books on IF. The Inform designers guide is very good.
> TLP is a good overview of what's out there and why it works. Emily Short was
> working on a book, but I haven't heard about it lately. I would add
> Crawford's book, even though it's not strictly about IF. I also read Hamlet
> on the Holodeck, but found it less useful. You might also look into AIML
> (for creating chatterbots); there is a book, but I haven't read it. If
> you're interested in multiuser works then read Richard Bartle's book,
> "Designing virtual worlds", which is really about MUDs and MMORPGs. The
> specific code for creating a multiuser dungeon/IF can be gotten from Ron
> Pendelton's "MUD game programming".
For those who would like to follow up on "MUD game programming",
it's worth noting that Google thinks the last name of that book's
author is "Penton".

David Jones

unread,
Oct 31, 2004, 1:32:52 PM10/31/04
to
"RootShell" <root...@netcabo.pt> wrote in message news:<41838928$0$8364$a729...@news.telepac.pt>...

>
> Any there any particular books (specially Inform related) that the members
> of this newsgroup recommend?

Strunk and White, Elements of Style.

Kernighan and Pike, The Practice of Programming.

Neither have much to do with IF but are excellent for honing ones
writing and programming.

David Jones

Adam Thornton

unread,
Nov 1, 2004, 12:52:57 AM11/1/04
to
In article <tgUgd.4242$K7....@news-server.bigpond.net.au>,

Mike Rozak <Mike...@bigpond.com> wrote:
>I think he's trying to boil an "interactive story" to do down to his
>essence. For him, the most important parts are characters and their
>relationships with one another. I felt that he boiled it down a bit too
>much, but that's all a matter of opinion, and it doesn't detract from the
>social AI ideas he puts forth. However, as soon as you add objects, combat,
>and connected rooms AI becomes more difficult to write.

Crawford makes the mistake of believing in Aristotle.

Plot that emerges dialectically from character makes for good Greek
tragedy, but there's a whole lot of worthwhile storytelling that is not
Greek tragedy. One can argue--pretty successfully, I think--that
stories about someone consciously choosing to violate the behavior his
nature dictates can be just as, or more interesting than, stories in
which someone's nature dictates his behavior.

Adam

samwyse

unread,
Nov 1, 2004, 6:30:47 AM11/1/04
to
On or about 10/31/2004 11:52 PM, Adam Thornton did proclaim:

Crawford is working on "interactive drama" more so than "interactive
fiction", so it isn't surprising that he's spending all of his efforts
on what we'd call NPC AI. Nor that, for now at least, he's modeling
relatively "simple" behavior.

While not as concise as Holodeck scene of _Stiffy Makane: The
Undiscovered Country_, here's a couple of excerpts from an idrama
mailing list that discuss what Crawford's doing:

http://www.flutterby.com/idrama/archives/msg00195.html
>
> But it's inconvenient (though not technically impossible) to
> encode into character state variables such things as characters'
> goals, plans, and conflicts. For example, Joe hates John, so he
> plans to be nice to John until he gains John's confidence, and
> then betray him. [...] Trying to build a whole story world this
> way, we would quickly run out of available variables in the
> Erasmatron. More importantly, we would find it impossible to
> keep everything straight in the author's mind.

http://www.flutterby.com/idrama/archives/msg00196.html
>
>A storyworld is going to be "messier" than a shooter or puzzle
> game, because human interaction is messier.

Reply all
Reply to author
Forward
0 new messages