Good puzzles and bad (and worse)

[tahnan]

Some day, I'll get around to actually putting up a website devoted to the theory and practice of puzzle-writing, in particular Mystery-Hunt-style "open-ended" puzzles, as opposed to crosswords or sudoku or the like (there're plenty of guides to writing crosswords done by people far more qualified than I am). Among other things, I'd like to work towards an understanding of what makes a puzzle bad vs. good, and what makes a puzzle good vs. great; for as much as can be determined independently of taste.

Here's a start. The following is, objectively, without question, a bad puzzle. Indeed, perhaps the worst puzzle I've seen. Worse than my standard example of a bad puzzle (namely, "How much money do I have in my pocket?"--at least there you have a chance). Worse...oh, here, just go look: https://secure.tanga.com/puzzles/265-2-28-2007.

Meditate on it. Consider it. Then click on the solution link below it. Meditate some more. And reflect that, no matter what else may happen to you, you will never be called upon to solve this puzzle, and your life is therefore just slightly more blessed than it otherwise might have been.

[As an interesting side note: testsolving is often considered a crucial step. In this case, the author explained: "...my test solver did solve it with the couple of hints I expected to give out during the first hour...." To which I replied, Here's the thing. If the hints are necessary to solve the puzzle, then they're no longer "hints"; they're just part of the puzzle. You might as well put them into the puzzle, if people can't solve without them. (Imagine a crossword grid, where after an hour of staring you get "hints"--i.e., the clues.) This is part of the aforementioned understanding of good vs. bad, and it underscores the fact that "It was testsolved" is clearly not an actual justification for a puzzle.]

Comments

[(Anonymous)]

The horror. The horror.

[thedan.livejournal.com]

This was me, btw.

[jydog1.livejournal.com]

but if I have more than one colon then I have 'colons,' no? I find the puzzle very 'colonic.'

[bourbon-cowboy.livejournal.com]

WOW! That's so awful it almost makes me angry. Jesus. And my question is, so why are the beer cans red? With the colons and the redness, I was sort of going towards "bilirubin."

[bourbon-cowboy.livejournal.com]

(Note, by the way, that I'm slightly colorblind, so my guess may be way off.)

[tahnan.livejournal.com]

They're actually Coke cans; that circle at the top of each says "Original Formula". Which is why they're red. That's also, mind you, completely irrelevant--you thought they were beer cans, and therefore that the red aspect mattered; I said 'Oh, they're cola cans' and thought the cola aspect mattered (especially with the "cola/colon" connection).

[toonhead-npl.livejournal.com]

I guessed "Tupac" so that the leftovers would be the other four cans.

[hahathor.livejournal.com]

Actually in this year's Mystery Hunt there were several puzzles (seven to be exact) where you got the puzzle and then, some time later, you got the hints you needed to solve it. Worked for me . . .

[mrmorse.livejournal.com]

That wasn't really being given the puzzle and then getting hints. That was more like getting the puzzle in two halves. Which is a generally reasonable thing to do.

[cazique.livejournal.com]

not to mention that the initial half of these were meant to be puzzles you'd receive in hell - the thinking was that an unsolvable puzzle would not be at all out of place in hell.

[mrmorse.livejournal.com]

That's not a colon, it's two dots. And those aren't aluminum cans, they're Coke cans. And that's not how you spell "colonel." And...okay, I'll stop.

[lapak.livejournal.com]

The spelling issue was the first thing that really jumped off the page and hit me in the head, yes.

"This word is pronounced 'kernal'" made me want to shout "No! It isn't! It isn't a word at all!"

So, tahnan - did you just post this to make us all groan, then? We need an example of a good puzzle, now, just to take the bad taste out of my mouth.

[jofish22.livejournal.com]

Wow, that really is hideous. Can you hunt down whoever did it and kill them? That would be a good puzzle.

Having said that, I would be very interested in you writing about what makes a good puzzle -- with the added question of "& how do you know that's the case?" It's something I've put a bunch of thought into over time, and it's very much related to my dissertation research about evaluating computer systems that are about experience and not about tasks. Obviously, a puzzle is not better the faster you solve it, unlike, say, an ATM transaction or getting a license renewed at the DMV; similarly, the opposite is also not true. I think there are similarities in Csikszentmihalyi's notion of flow, in some of the work on what makes computer games fun, but I'd be interested in hearing your take on the question.

Clear to me!

[530nm330hz.livejournal.com]

Gee, I can see how it would be "popcorn" but I don't buy their explanation:

Those are clearly cans of soda, or, as it is called here in New England, "pop". And I agree that the symbol is a "colon". So you start off with "popcolon" --- except that in Hebrew, the "o" sound can be written with a "vav" which has a Gematria of "6". So if you take the bigram "lo" and intepret it as the letter l, shifted forward 6 positions in the alphabet, it becomes "r", which yields "popcorn".

So why is this a bad puzzle, exactly? :-)

[36 hours to Purim and counting....]

Re: Clear to me!

[dr_whom]

Soda is called "soda" in New England, or occasionally "tonic". "Pop" belongs to the Midwest and Northwest.

Re: Clear to me!

[530nm330hz.livejournal.com]

Details.

Re: Clear to me!

[dr_whom]

Details.

Wow.

[lilisonna.livejournal.com]

Okay. I am not a puzzle person, but even I can recognize that as bloody awful.

[kirbyk.livejournal.com]

It'd just been simpler for them to say that it was pop, and the dots were corns, like people get on their feet, and be done with it.

This isn't a puzzle, it's 'guess my stream of consciousness'.

[tinhorn2.livejournal.com]

Tahnan (and any others who want another unfortunate puzzling experience)--

Recall:

https://secure.tanga.com/puzzles/215-1-23-2007

In some ways, I seethe at that one more than the one with the cans, because after a minute or so, it was easy to conclude that the answer to the cans puzzle was probably going to suck. OK, not easy to conclude that it was going to suck quite as much as it actually did, but still. In the puzzle I linked above, there were enough things that looked interesting, but none of them worked, and the ultimate solution seemed leapy in several ways, particularly ways that could have been fixed with a little care (and testsolving -- oh, where is the testsolving?).

If you'll permit me an analogy, to me, the cans puzzle was like a bad singer nowhere near the notes, at whom you just shake your head, whereas (to me) the mirror puzzle above was like an off-key singer that is affirmatively painful to listen to.

I've said this before about Tanga -- letting so many bad puzzles through is a real problem beyond just the bad puzzles. It makes it hard to stick with any puzzle, even a good one, for more than about 5 minutes, lest you waste time on something that's hopelessly broken. I probably stayed with the mirror puzzle for about 1/2 hour before punting; nowadays I'll punt much quicker. If that means missing a chance at solving a particularly elegant puzzle, well, my loss, but in the long run I'll come out with a net plus.

And yet, Tanga sits on a puzzle that I submitted in December, which I liked more than the one of mine they ran last month, lowering the incentive to create and submit more.

[tahnan.livejournal.com]

Oh, yes, the mirror puzzle was awful, to the point that I actually saved a copy of the image into a directory on my computer for future use in a "look how bad this puzzle is" essay. But at the same time, there was the--pardon the word--kernel of a good idea in the smoke-and-mirrors puzzle; there was, to be honest, nothing even resembling a good idea in this one.

[http://users.livejournal.com/merle_/]

Could you recommend a good text/guide/site for creating some sort of simpler puzzles, like crosswords, crosssums, sudoku, logic puzzles, or the like? I'm mildly interested in the subject, and know I should just figure it out on my own, but have little time to do so...

[tahnan.livejournal.com]

Sadly...no. I mean, I can try, but since I tend to write, er, more baroque things or straight word puzzles, I've never really had much cause to go looking for a guide to the other sort. (Indeed, when I need a grid made, I invariably ask someone else to do it for me.)

saxikath suggests The Random House Puzzlemaker's Handbook for crosswords; there's also The Compleat Cruciverbalist. Cox and Rathvon's Random House Guide to Cryptic Crosswords is probably the gold standard for that sort of puzzle. I don't know much at all about sudoku or cross-sums construction, though.

[http://users.livejournal.com/merle_/]

Cool, thanks. All out of print, of course, but maybe the library has copies...