One of the central game mechanics in the Legend of Grimrock -series (can we already call it a series?) are puzzles. They’re a very natural pairing for an oldschool dungeon crawler like ours and it provides a nice counterweight for the exploration and combat you’ll be doing.

After we finished Grimrock 1 I thought that we might have squeezed out all the possible grid based puzzles with levers, pits, doors and traps that we could possibly muster. And I felt that way for a long time too but strangely, after wrapping up the release of Dungeon Editor for Grimrock 1 with Petri, I realized there might still be a few puzzle ideas hiding deep somewhere in the back of my head, just waiting to be digged out. Figuratively speaking of course, not literally… :)

Of course, a lot has happened that has fed the puzzle creation process somewhat. The Dungeon Editor makes it a lot easier to prototype and build puzzle ideas and we’ve also come up with some new mechanisms that can be used for entirely new styles of puzzle or combined with the old elements for a fresh take on the older puzzles.

All in all, it’s impossible to predict exactly where the next idea comes from but no matter what their origins are, they’ll end up in the puzzle_ideas.txt in my dropbox. At this moment there’s about 40 unused ideas, many of which are very abstract and probably won’t end up as actual working puzzles, but others are much closer to reality and more ripe for use. Let’s take a look how one of the ideas ended up in the text file and how I took the idea to completion (as seen on the screenshot above) and how it changed along the way.

Here’s the original notes I wrote down about the puzzle:

- "16 steps"
  - you need to travel from A to B taking exactly N steps along the way
  - multiple routes are offered

But before I had even that, I had been playing some Slitherlink recently and I was thinking that could I somehow transform the classic puzzle into a Grimrock puzzle. In Slitherlink, the player draws a continuous line on a graph paper so I started thinking if the player party could in essence draw a similar continuous line by moving the player party in the grid of the game world. A difficult numbers problem, like that in the original Slitherlink, would be too cumbersome to grasp in the first person view of our game so I came up with a simpler idea: player would just need to cross the room reserved for the puzzle in a certain number of steps without crossing his previous path, thus drawing a squiggly Slitherlink-like continuous line through the room. At this point I stopped to write down the core idea for the “16 steps” puzzle in the puzzle_ideas.txt so that I could return to work on the idea later.

I’ve discovered that usually puzzles where the “rules” are simple but the solution is not, work the best so it was only natural that to improve this puzzle, I ended up making the rules simpler. After drawing a few possible room layouts for the puzzle on paper, I realized that counting the steps was an unnecessary rule since all I need is just a room with a funky layout where the player needs to step on all the squares. If the player makes it across the room but he hasn’t gone through all the squares, the puzzle won’t be complete and the exit door won’t open. This felt like a good starting point so I moved on into the Dungeon Editor to start hammering out a working prototype.

How to translate the still sort of abstract idea into the actual game using common puzzle parts was solved quite easily by using a room covered with closed trapdoors which would open after player moved off of a trapdoor. This way the player leaves a line of opened trapdoors in his wake, making it so that the player can’t cross the previous path and it also provides us a neat way to reset the puzzle: if the player gets stuck, he needs to drop down a pit and when he comes back, the trapdoors reset into the original state and the player can retry.

At this point I already knew that I had a good puzzle in my hands and that all it needed was a right size and shape for the room. Sometimes just adding or taking away a tile or two can transform this kind of puzzle from almost impossible to too easy so it often takes a lot of experimenting to come up with the right layout. This part of puzzle creation might sound like a boring grind but I actually find it quite enjoyable since it totally feels like a puzzle in itself and it’s always very satisfying when you end up with a layout that seems to be balanced just right. So, how did the room end up looking like in the end? Well, you’ll just have to wait and see until you get to play the game yourself since I’m not going to hand out any more spoilers here! :)

  10 Responses to “Puzzle Creation in Grimrock”

  1. This was a fun read!
    I really loved the puzzles in Grimrock, due to the sheer variety they offered.
    You had some that focused on logic, some on player agility, and some on both.

    You said that everyone in the devteam helped think them up, didn’t you? I think that really helped.

  2. That’s a great puzzle format, I’m going to steal it, I think.

    Way excited for more new and interesting puzzles. The home made dungeons have been a great source, but with a new game comes new mechanics. I am very eager to see what you guys cook up.

    Also, beautiful environments. I have the ogre bridge as my desktop and people ask me where I took it if they only get a glance. Much prettier than dank dungeons.

  3. Had to make an account just to congratulate you guys on the graphics style. It’s marvelously crisp and clear yet still rugged and dirty. Textures look absolutely amazing. Can’t wait to play LoG 2.

  4. Puzzles on grimrock was one of the most valuables things for me, now seems better enough on grimrock 2, a very good news!!. Congratulations to antti for all the good ideas about it.
    First picture with ogre was morning, now night falls… interesting, very interesting ^_^
    I hope minimun requirements not too higher respect first part. My poor pc have more years than me!! :P

  5. That new tileset, trees, and sky are all looking pretty awesome there.

    I remember actually creating that type of puzzle in the level editor. I think the only layout I had though was this:


    Can’t wait for more new puzzle types, especially ones like this with many layouts available!

  6. Must be pretty tough to keep coming up with whole new puzzle concepts instead of re-iterating on existing ideas with a slightly different setup – sounds like you’re doing some great work. I guess for a puzzle like this, it might be nice to provide a map of the area before hand so that you can plan your path without having to fail the first time to see where all of the potential move spaces are. The component system must make it nice and easy to create new objects/items for use in puzzles too – for putting cogs together in the right sort of alignment on the wall etc.

    Anyway, looks and sounds excellent. Look forward to seeing some of these puzzles in action!

  7. I remember playing through a lot of these kinds of puzzles in the Legend of Zelda: Link’s Awakening DX. If you want some inspiration for puzzles play through a bunch of the Zelda games, there are tons of great ones in them.

  8. Fantastic looking graphics there!

    It’s also great to read about the puzzle design. I hope that you’ll also keep a fair few of the “narrative” puzzles in there, where there are written clues you put together (such as notes with “historical” information that you later use as a number or pictographic combination, although I appreciate that would require being able to file all that information away as the paper would clog up your inventory), or riddles you have to figure out, or just plain hidden passwords etched into surfaces in odd locations* or just on scrolls hidden under junk objects. It’s those sort of puzzles which really breath life into a world more than the strictly mechanical ones for me and I really enjoyed some of the more abstract puzzles in the first Grimrock game (the one with the two Dragon statues being a stand-out as it required you to make notes).

    * Maybe actually make use of the free-look to look up at the ceiling or down at the floor!

  9. Please include less timing puzzles this time around. I couldn’t complete most of them on my laptop, so I had to wait until I was on my desktop before progressing further when I came to a timing based puzzle. Also, if you guys could include less puzzles that upon failure dropped me to the next dungen level down that would be great too, because I really got tired of falling to the next dungeon level down just to find a teleporter that returned me to the level I fell from just to try again, and again, and again.

  10. I agree with GrahamGoring about the need of “narrative” puzzles.
    Mechanical “trial and error” puzzles are not the best stuff that GR/DM kind of exploring game can offer. Trying to just find the right combination is just plain work that gets soon boring.
    Instead in a really good puzzle there should be a story that keeps the interest up and the solution is more rewarding in the end.

© 2013 Almost Human Ltd. Suffusion theme by Sayontan Sinha