Thread by @filiphracek

Ok, this is going to be a long thread, mostly about an obscure methodology in urbanism but also about gaming and an idea for an interactive fiction (IF) entry to @ifcomp.

So, my parents are architects and my mom used to be involved in urban planning. She once told me about an innovative* way of public park planning.

*) It was innovative at the time my mom learned about it, so probably in the 1970s-1980s. It may have been surpassed since.

Imagine you're an urban planner architecting a public park. Let's say it's a rectangular area in a walkable city (e.g. a regular European city with lots of pedestrians), but otherwise there are no other requirements.

A major part of your work is designing the paved path across the park. You want to maximize the green parts but also want to let people to cross the park efficiently.

You could just guess where most people will want to be walking, but there is a high price of guessing wrong.

No enough paved paths mean that people will be walking over the grass, making footpaths in the process. This leads to bare ground in many places, which in turn leads to mud. In winter, these paths can be hazardous (they won't be maintained).

On the other hand, too many paved paths mean is a waste of money and waste of park space. You're building a park so that there's more green in the city, not more pavement.

Ok, so here's the inovative approach. At first, you don't build paved paths. You take a month when there's not likely to be mud or ice, and you leave the soon-to-be-park unfinished.

People will start walking over it. This will create many natural footpaths all over the place. In a week or so, some footpaths will start "winning" over others, until almost everyone just uses the major footpaths.

That's when you, the architect, come back, and measure the footpaths. You design the rest of the park around these natural paths, and put pavement in the place of the current footways.

Well done. You just designed a public park that's as green as possible but also practical for the people living in the area.

What does this have to do this anything? Recently, I thought: How can we use this same approach for UI design? I have some ideas, but that's outside the scope of this thread.

A tangent idea is about games. Of course, games already do something similar. There are playtests and game analytics, and these can provide the feedback that the game designer uses for further changes to their design.

But what if we want to transfer the public park paths methodology to games in a little more literal way? Game analytics and playtests only help _after_ you have a game. No playtested game is as blank of a medium as a rectangle of ground.

Imagine a 3D shooter, for example. You can't let people wander around an empty space, expecting them to give you anything useful for your level design.

Same with 2D platformers, graphic adventures, racing games — basically any other type of videogame. None of these work for our purpose.

But then there is parser interactive fiction (IF). That's the kind of IF where you have to write each of your actions. "get lamp", "attack goblin", and so on. (A more complete definition by @iftfoundation is here: iftechfoundation.org/frequently-ask….)

Parser IF is not my neck of the woods (egamebook.com in choice-based simulation IF, a different type) but I still know enough about it that this excites me.

How would you implement the public park pathway methodology to parser IF? Here's my proposal.

First, you create a "blank" game. This game is only a succession of puzzle sets (get out of a prison cell, overcome a guard, get transportation), but the puzzles themselves aren't implemented. They are only introduced.

So, the player of the "blank" game just gets a room/problem description. Something like "You are a beast master (you can talk to animals) imprisoned in a dark cell. You need to get out.".

Now, every time the player does something ("examine wall", "talk to ant", "make ant go south"), we just tell the player to assume that whatever she just did succeeded in the way she expected.

We also ask the player to type "finished" after they solved each puzzle in this way. That gives them the following puzzle.

Once the player is finished with this "blank" game, we ask them to send the transcript of the game to us.

We gather many transcripts, and then try to find the most common paths and ideas.

The finished game is the equivalent of the public park with paved pathways over the natural footways. It's the most often used ideas, fleshed out with a story.

We can use the less popular ideas to implement fail scenarios ("You examine the ceiling but there's nothing there. Not even a fly. Bummer.").

I hypothesize that such a game would be fun to play. It would probably be easier than the average parser IF game (which is a good thing, from my perspective), and more predictable, but with good writing and ideas, that would not matter.

There's a possibility that the main paths would be too straightforward or boring ("open cell door, exit"). But I have strong faith that most of the players of the "blank" version would rather come up with interesting ideas.

The finished product would obviously raise questions about authorship. The players of the "blank" version would be co-authors, in my opinion, even though they only provided the commands, and those commands were aggregated.

Now, this is something I'd be very interested in exploring, as an @ifcomp entry. But I'm pretty sure I won't do it. I promised myself that I first have to finish the next iteration of egamebook.com before more experiments.

But I had to get it out of my system somehow. I'd love to discuss it, and learn about similar ideas from the past. Also, if you're into parser IF and this captured your imagination, feel free to run with it!

The end.