Matt Chat 119: Josh Mandel on remaking King's Quest

warning: Creating default object from empty value in /home/buckman/public_html/neo/modules/advanced_forum/advanced_forum.module on line 492.
Matt Barton's picture

In this first segment of my interview with Sierra designer Josh Mandel, we chat about how Josh got his start, remaking King's Quest, Zeliard, Leisure Suit Larry, Laffer Utilities, and much more. We also talk about the difference between creating puzzles for parsers vs. point-and-clicks. Josh thinks the latter was a dumbing down with grave consequences for the genre.

Download the MP4.

Comments

Chip Hageman
Chip Hageman's picture
Offline
Joined: 10/06/2010
Josh Mandel
Matt Barton wrote:

In this first segment of my interview with Sierra designer Josh Mandel, we chat about how Josh got his start, remaking King's Quest, Zeliard, Leisure Suit Larry, Laffer Utilities, and much more. We also talk about the difference between creating puzzles for parsers vs. point-and-clicks. Josh thinks the latter was a dumbing down with grave consequences for the genre.

Great episode Matt. It's a shame Josh suffered flooding from the recent hurricane... it just causes a host of problems on multiple levels.. black mold, drywall and carpet damage (if the basement is finished).. Losing a Vectrex would pretty much bring tears to my eyes... losing a Virtual Boy.. eh. I'd probably place the Vectrex on a pile of Virtual Boys to keep it above the water level.

All kidding aside; as to the design docs: I wish more designers would look to these as historic documents. It would be great if an effort could be organized to have designers (or interested third parties) scan these pages to preserve them for posterity.

Anyway, I'm looking forward hearing about his experiences with Freddy Pharkas, Frontier Pharmacist. It's one of Al Lowe's greatest under-rated classics... in fact, I have it slated for another play through fairly soon.

As to parser vs point and click.. I agree with Josh. The parser system just gave you a better feeling of control over the game... that being said, I think Lucas Arts had a better implementation of the point and click engine.

n/a
Matt Barton
Matt Barton's picture
Offline
Joined: 01/16/2006
Thanks, Chip. I have to

Thanks, Chip. I have to admit, I hate parsers. That's the main reason I never really got into the Sierra games. There was a LOT to like, and they oozed charm, but man, that parser just made things to difficult, and not in a good way. Infocom did an admirable job trying to presuppose inputs, but even those often left you stuck trying to figure out how to word something (and that's assuming you even knew how the designers intended you to solve a particular puzzle).

As Josh says, though, the point-and-click solution has problems, too. I think it makes some designers lazy, since they know you can always try everything to get past something (even if it makes NO sense).

For adventure games to really move forward, you'd need artificial intelligence, really. It'd have to be smart enough to allow for lots of different solutions, especially those the designers themselves never anticipated. I think tech like that is a good 50 years off at best, though. It also starts to affect the designers' ability to tell a coherent story.

n/a
Chip Hageman
Chip Hageman's picture
Offline
Joined: 10/06/2010
Z-Machine Parser

Dunno, the Infocom parser was a thing of beauty.

"Wikipedia" wrote:

The Infocom parser was widely regarded as the best of its era. It accepted complex, complete sentence commands like "put the blue book on the writing desk" at a time when most of its competitors parsers were restricted to simple two word verb-noun combinations such as "put book". The parser was actively upgraded with new features like undo and error correction, and later games would 'understand' multiple sentence input: 'pick up the gem and put it in my bag. take the newspaper clipping out of my bag then burn it with the book of matches'.

In a non-technical sense, Infocom was responsible for developing the interactive style that would be emulated by many later interpreters.

As to AI in games.. this is normally a very difficult thing to do. The dirty secret is that all adventure games are scripted no matter what anyone tells you.. i.e. you need to complete certain quests for a given result. Letting the game "wing it" means it would likely need to design new goals on the fly. I'm certain we will get there.. but for all of the open-ended gameplay from games like Ultima, E.S. Oblivion and the like.. we aren't there yet.

n/a
Matt Barton
Matt Barton's picture
Offline
Joined: 01/16/2006
I NEED to do a podcast on

I NEED to do a podcast on this topic. It's been stuck in my craw for weeks, but the thing is, it's so complex I can't quite wrap my head around it. I guess I could at least point out why it is so difficult.

But basically my contention is that the problems faced by Infocom and Sierra and any adventure game is that they're whole quest is somewhat misguided. It's not that you CAN'T tell stories with games, but you have to tell the kind of stories that work WITH the format rather than AGAINST it. A linear story, where the hero must do X, Y, and Z (in that order) to advance to chapter 2, etc....You're force fitting a conventional narrative structure onto something that resists it by nature.

Elder Scrolls...I don't necessarily love those games, but I think they may be on the right track with algorithmically generated content. The only problem is, it's still in a laughably crude stage. And people need to be convinced that they're in "good hands" narratively speaking, not just playing out a random scenario. So you need a sort of meta-storyteller who is good at telling stories within a dynamic structure. The brute force approach of just having forked paths could work, but you'd need incredible time and resources to manage all that. Maybe it's possible. But probably more practical is to have certain fixed events that will happen regardless of what the player does, such as an earthquake or alien invasion, but let the player work out the details of the plot, reacting in real-time to his own decisions. This is basically what all adventure games claim to do, but they fail when you MUST do X, Y, and Z to get from point A to point B, etc. The structure is just too rigid.

They way I'm trying to respond to this in my game design is to give the appearance of freedom. I think that's enough and players will respond accordingly once they feel out the limitations of the system. People are probably naturally inclined to suspend their disbelief and want to do so; all you need to do is make it convincing enough so that you're not breaking it. I don't think artificial intelligence is really the goal; all you need to do is fill in the picture enough for the player to do the rest, which they're happy to do. The more common problem (IMO) is that the designers fill in TOO MUCH of the picture, which they think makes it more realistic, but in actuality inhibits immersion by forcing the player's imagination into predefined channels. That's why I think a lot of people find the Lara Croft of the first Tomb Raider game a lot sexier than in the later "more realistic" sequels, or even Angelina Jolie in the movie.

n/a
Chip Hageman
Chip Hageman's picture
Offline
Joined: 10/06/2010
The Great Underground Empire...
Matt Barton wrote:

It's not that you CAN'T tell stories with games, but you have to tell the kind of stories that work WITH the format rather than AGAINST it.

That was the conclusion Richard Garriott came to.

Matt Barton wrote:

A linear story, where the hero must do X, Y, and Z (in that order) to advance to chapter 2, etc....You're force fitting a conventional narrative structure onto something that resists it by nature.

This is why I hate Japanese RPGs.. well, that and the hair. ;)

Matt Barton wrote:

Elder Scrolls...I don't necessarily love those games, but I think they may be on the right track with algorithmically generated content. The only problem is, it's still in a laughably crude stage.

Agreed.. Fable did this too. Depending on your characters actions certain options would open and others would close. That type of low level game terraforming is easy, but ultimately not very engaging. I think what you are looking to do is have the game morph objectives (plot points) based on the characters decisions. The only way I see of accomplishing that is to build every choice into the game and let the actions open and close branches.. but there will still be a finite number of choices and outcomes.

Matt Barton wrote:

The more common problem (IMO) is that the designers fill in TOO MUCH of the picture, which they think makes it more realistic, but in actuality inhibits immersion by forcing the player's imagination into predefined channels.

That is exactly the reason why I loved text adventures... specifically Infocom text adventures. They painted a scene with words, dropped you to a text parser and let you (and your imagination) work it out. It left all of the imagination and wonder in the game.. and it gave you a great sense of accomplishment when you solved the puzzles.

Matt Barton wrote:

That's why I think a lot of people find the Lara Croft of the first Tomb Raider game a lot sexier than in the later "more realistic" sequels, or even Angelina Jolie in the movie.

Mmmm, Angelina... What were you saying Matt?

n/a
Matt Barton
Matt Barton's picture
Offline
Joined: 01/16/2006
ha
Chip Hageman wrote:

Mmmm, Angelina... What were you saying Matt?

LOL.

n/a
Loggins (not verified)
Nice

Great episode - very interesting. I loved the Sierra games in the 80s and 90s, and it was fun to hear some inside views from someone that worked on the games i loved at the time.

Im now going to watch the next two installments of this interview - just wanted to leave a little comment for you before i continue watching.

Regards
Loggins.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.