Mattcrampy
Level: Smitemaster

Rank Points: 2388
Registered: 05-29-2003
IP: Logged

An interesting concept I've seen is the idea of architectural patterns - that is, stock solutions that have been proven as the most effective way to solve a problem. Obviously, no architect worth his salt will just drop a boring building somewhere, and the patterns are vague enough to leave room for architectual flair - just think of all the ways you can design a courtyard.

This idea has been adapted with some success to programming - the coders among us will appreciate the idea of having a set of 30 or so 'stock solutions' to common tasks in programming, such as having lots of different ways to display graphics and lots of different things to display, for example, without having to make a whole load of classes with basically the same code in them and therefore a whole lot of bugs.

I doubt this could be adapted to DROD successfully - after all, the idea of creating a puzzle in DROD depends much on doing something unexpected, whereas patterns are built around doing something neat and expected. DROD levels thrive on complexity.

Thoughts?

Matt

____________________________
What do you call an elephant at the North Pole?

agaricus5
Level: Smitemaster
Rank Points: 1838
Registered: 02-04-2003
IP: Logged

That's the whole point of many of the new puzzle designs and some of the old ones - to design a complex and intriguing puzzle while making the room look as original and aesthetically pleasing as possible, often by including structural patterns and unusual formations.

Well, that's what I think, anyway.

____________________________
Resident Medic/Mycologist

bibelot
Level: Smiter
Rank Points: 322
Registered: 07-11-2003
IP: Logged

I've found that repeating single components of rooms in other rooms can be helpful at times. For example, in my dungeon I have a series of rooms that require the player to ensure that exactly a certain number of monsters enter a given chamber. Then there's a whole section connected to that that is designed to check that this exact number is there. If the player has seen this component of the room before, then they don't have to spend time figuring out what it does, which is probably the boring part of the room anyway. Instead, they can focus on whatever the real puzzle is.

But I've definitely seen what I would consider patterns on a smaller scale. For example, having a roach in a chamber that cannot be accessed from the outside enforces that a mimic be placed there. Or, to force that certain orbs are hit in a certain order, just have a passageway with arrows and trapdoors and such. And that doesn't make the puzzle worse, it just makes less unnecessary stuff to think about.

That's why I think that developing these types of patterns would help both architects and players. The main downside I can see is that making these patterns common knowledge would allow the architect to be less clever, as he/she could just copy components from other people. This could lead to puzzles that seem unoriginal and reused, but I would like to think that an architect would want to design a room that was somehow new anyway.

ErikH2000
Level: Legendary Smitemaster

Rank Points: 2810
Registered: 02-04-2003
IP: Logged

Mattcrampy wrote:
An interesting concept I've seen is the idea of architectural patterns - that is, stock solutions that have been proven as the most effective way to solve a problem. Obviously, no architect worth his salt will just drop a boring building somewhere, and the patterns are vague enough to leave room for architectual flair - just think of all the ways you can design a courtyard.

Well, there are definitely some patterns that are emerging. All of these I've seen come up more than once:

PROBLEM: Require tar mother to live past the time where the player is able to kill her. Player must make a counter-intuitive decision to not destroy the tar mother until later.
SOLUTION: Create a fork with tar blocking it such that accessing one path will close off the other. Tar growth is necessary before the other path will become accessible.

PROBLEM: Require player must complete some task in a limited number of moves or the room will become insolvable.
SOLUTION: Create a serpent of a length matching the number of turns you wish to allow the player. The serpent's head will point into a dead end. While the serpent is still alive, he will obstruct a roach from moving to a location that would create an impassable guantlet. A second roach is already positioned on the other side of the guantlet before the serpent dies.

As you point out, as a player it is better to find unique situations in DROD than tried-and-true engineering solutions, so probably it would not be so valuable to organize a collection of patterns.

Still, it's interesting you hit on this pattern notion.

-Erik

____________________________
The Godkiller - Chapter 1 available now on Steam. It's a DROD-like puzzle adventure game.
Decent Apps - open source tools for writing and deploying privacy-focused LLM-based web apps.

Tscott
Level: Smiter

Rank Points: 382
Registered: 02-10-2003
IP: Logged

In my first level, Impossible Cubes http://www.drod.net/forum/viewtopic.php?TopicID=886 I tried to keep with a theme or pattern. The original idea was an 'impossible objects' level which would have a variety of impossible objects like the impossible tribar triangle, or the impossible tuning fork translated somehow into DROD. But the cube was the only thing I could manage to recreate (at the moment, I still might work out a way for the others) so cubes became the basis of the level. Of course, I then simplified it further to 'squares with tricks' for many of the rooms as the actual cube design only seemed to lend itself to tar puzzles.

I also like the idea of a map (the big picture) being part of the pattern, like the serpentine look of Level 7, the introduction of the snakes, in the original DROD levels. In my level, while the main 4X5 block of rooms isn't a square it looks square-ish on the map, and allowed me to make an area outside it which broke the pattern on purpose for a couple rooms.

I hope that I can be creative enough in any future releases and stick with a theme of some sort, rather than random puzzles from room to room (not that there's anything wrong with that).

____________________________
And I can recall our caravel: a little wicker beetle shell with four fine maste and lateen sails,
its bearings on Cair Paravel. O my love, O it was a funny little thing to be the ones to've seen.
-Joanna Newsom "Bridges and Balloons"

Mattcrampy
Level: Smitemaster

Rank Points: 2388
Registered: 05-29-2003
IP: Logged

I'm building a hold where a lot of the style features combine with other parts of other rooms to make intersting patterns. I may try doing large scale art like the "Ketchup and Puppies" or the "Tar Exploit" holds, but over multiple rooms so you can't tell.

The biggest trouble I've had so far is building rooms with non-trivial puzzles that are still interesting. Most of the target audience have finished DROD and are looking for the puzzles Erik never dared to do, and I can't do it! I JUST CAN'T DO IT!!!

Ahem.

Matt

[Edited by Mattcrampy on 07-29-2003 at 09:12 AM GMT: Puppies, Puppies, so good I named it twice]

____________________________
What do you call an elephant at the North Pole?

ErikH2000
Level: Legendary Smitemaster

Rank Points: 2810
Registered: 02-04-2003
IP: Logged

Mattcrampy wrote:
The biggest trouble I've had so far is building rooms with non-trivial puzzles that are still interesting. Most of the target audience have finished DROD and are looking for the puzzles Erik never dared to do, and I can't do it! I JUST CAN'T DO IT!!!

Well, I for one, like the simpler puzzles better. So far the most enjoyable hold for me has been TScott's Impossible Cubes, which is only moderately difficult. I also appreciate when someone can make a difficult puzzle without using a lot of complexity. Ricky's Dungeon had several clever rooms like this.

-Erik

____________________________
The Godkiller - Chapter 1 available now on Steam. It's a DROD-like puzzle adventure game.
Decent Apps - open source tools for writing and deploying privacy-focused LLM-based web apps.

zex20913
Level: Smitemaster

Rank Points: 1723
Registered: 02-04-2003
IP: Logged

It was Ketchup and Puppies, not Kittens!

____________________________