The Sphinx's Lair

« first day (1785 days earlier) ← previous day next day → last day (2076 days later) »

North

00:44

Can someone explain the difference between what makes a puzzle open-ended and what makes it too broad?

2 hours later…

Deusovi

02:23

@North "Open-endedness" is a quality of a puzzle itself: it means the puzzle intentionally has multiple possible answers, or possibly none. (I would argue that this is always bad, but that's beside the point.) "Too broad" is a quality of a puzzle as it relates to Puzzling Stack Exchange: it means that the puzzle is not a good fit for this site because it has no clear best answer, either because there are too many answers or multiple components that can be individually answered.

Most open-ended puzzles are too broad; for an open-ended puzzle not to be too broad, it must have a way of choosing a single best answer out of many (for instance, the longest word or set of words satisfying certain criteria). Even then, it still may be too broad: for instance, if solutions can be arbitrarily long, or if there's no clear rule for what counts as a word.

(I don't think open-ended "puzzles" should be on the site at all, to be completely honest. They're fun games, but IMO they're not puzzles.)

Rubio

I don't mind open-ended puzzles. At least, not if they're any good, and are well constrained.

I'll let you know if I spot one. :)

(Ok I think we actually do have a couple, but - the bad ones are way more prevalent)

Don Thousand

It's just, most open-ended puzzles aren't usually designed to be so. That's a good sign it's bad.

Deusovi

02:41

I'm not sure I've every seen an open-ended puzzle that I'd consider both "well-constrained" and "good". All the ones that I've seen here rely on words in some way, and then you run into one of these two:
- what counts as a word isn't well-defined, so it's too broad
- what counts as a word is based on some arbitrary source, so it's unsatisfying

Maybe a logic-based one (like "what's the smallest number of clues you need to make a uniquely solvable Slitherlink whose solution touches all four edges of the grid") would be nice. (For what it's worth: you can get 2n-5 clues.)

But I'm still not sure that anything open-ended should even count as a puzzle. Open-ended things can be interesting questions (and in fact, that one would be on-topic for this site, because it's about puzzles). But I think puzzles should have only one solution, and anything more than that is either a broken puzzle or not a puzzle at all.

Don Thousand

Unless, the puzzle itself somehow revolves around being open ended? The possible solutions are somehow aggregated in the puzzle?

Deusovi

I can't think of a way for that to be solvable as soon as it's posted and years in the future, which would be a pretty significant problem

Don Thousand

I can't think of one either. Just trying to play devil's advocate lol

Honestly, the OEIS has ruined so many good sequence puzzles I've come up with. I think, there's no way this is in the OEIS, and then there is always is

And I know I could hash a sequence to make it undiscoverable, but that kinda ruins the point of the puzzle

Deusovi

That's one of many reasons I don't think sequence puzzles are good puzzles.

Rubio

03:08

but.. but... A "simple" number sequence puzzle

Sphinx

Q: No shovels, please!

$ \color{red}E $ Second time skimming through Beatles residence Sandwich way What do these clues point to? Looking for a single word answer. This puzzle is a hint for Ah, the old SE puzzlearoo!

Sphinx