The kind where Alice says that Bob is lying, and Bob says that Alice is a murderer, and Claire says that Bob is telling the truth, and so on.
One of my IFComp entries is going to feature these sorts of puzzles, and I’ve built a generator that creates them automatically at runtime. But since I know how the generator works, I can solve the puzzles trivially—and that means I can’t really gauge how hard they are for others.
So I’m looking for people to test a sample puzzle and say how it feels. I think I could just post the puzzles publicly without breaking any rules, since they will be heavily, heavily modified before appearing in IFComp, but I don’t want to tip my hand too much.
(The final version of the puzzle will be presented entirely in the ancient Hittite language, which is likely to ratchet up the difficulty by several notches. So I’m aiming for “not trivial, but not difficult either”. Something anyone can solve with pencil and paper, and experienced logicians can solve in their head.)
I like but am not especially sharp at those kind of logic puzzles (I was great at them when I was studying for the LSAT… 23 years ago), and I’m confident that I have an above-median knowledge of Ancient Hittite (inasmuch as I know it exists), so glad to take a look!
When I was first building the generator for these puzzles, I was guided by a blog post based on the Smullyan book (specifically trying to write algorithms to imitate his various puzzle styles). Sadly that post is gone now and doesn’t seem to have been backed up anywhere. But that’s where I got the inspiration to do it all with graph theory.
Also, I’m glad (and a bit surprised!) that so many people have shown interest! But I’m going to take a few days to incorporate the feedback from the first round of testers before I send out any more puzzles—that way I can know if the changes accomplished what I wanted them to accomplish or not.
Well, playing IF is a type of logic game. I usually don’t like it when puzzles like Towers of Hanoi, Lights Out and stuff like that interrupts my IF experience, but I happen to like three things enough that I’ll always enjoy them: nonograms, basic substitution cyphers, and logic grid puzzles.
I mean, they are still taking me out of the game, so I would usually say that an author has to really consider whether they really want them. A logic puzzle appears in Broken Age, and I felt it was a bit unwelcome. But honestly, and subjectively, it seems to be a better, less disruptive fit to IF.
Anyway, this is to say that I also surprised myself by being interested. I guess the overlap between IF and logic grid really is non-trivial.
I’m a big fan of these puzzles, too. Like two gatekeepers, one lying, the other telling the truth…You probably know this puzzle.
I hope I’m not derailing the topic:
What do you people think about puzzles that have a root in the world/story? Like a safe to crack for example? Do you think even these puzzles distract from the story?
There’s a whole series of logic puzzle books called the Montague Island Mysteries, that feature puzzles with a common narrative thread. So for example, a group of people arrive on the island for a retreat, then a series of events occurs involving all of the persistent characters. They’re IF, of a sort, in the same way that Type Help is.
Wandering off on a tangent slightly, but I find that the loose narrative framing you frequently find around “canonical” math problems often has an appealing aesthetic quality. I’ve made an IF game based on the Bridges of Konigsberg before, and there’s an entire loose pseudo-genre of “prisoner problems” that has untapped game potential that I haven’t been able to figure out yet.
I usually don’t like it when puzzles like Towers of Hanoi, Lights Out and stuff like that interrupts my IF experience, but I happen to like three things enough that I’ll always enjoy them: nonograms, basic substitution cyphers, and logic grid puzzles.